id
int64 393k
2.82B
| repo
stringclasses 68
values | title
stringlengths 1
936
| body
stringlengths 0
256k
β | labels
stringlengths 2
508
| priority
stringclasses 3
values | severity
stringclasses 3
values |
---|---|---|---|---|---|---|
555,904,048 | go | x/crypto/acme/autocert: add hooks for transaction locks | The `Manager.createCert` method has a lock around the ACME transaction so that multiple goroutines can safely call this method concurrently. This does not protect against multiple processes (possibly on different machines) doing this concurrently.
If transaction hooks were added, then a caller could provide a custom `Cache` which uses remote storage, allowing multiple servers to safely request certificate generation/renewal and share the certificates.
If this approach is acceptable, I can start work on a PR. | NeedsDecision,FeatureRequest | low | Minor |
555,928,830 | go | proposal: runtime/pprof: add PMU-based profiles | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13.4 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes; tested on `go version devel +74d366f484 Mon Jan 27 20:45:39 2020 +0000 linux/amd64`
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/home/xxx/.cache/go-build"
GOENV="/home/xxx/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOINSECURE=""
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/home/xxx/go"
GOPRIVATE=""
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/home/xxx/go"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/home/xxx/go/pkg/tool/linux_amd64"
GCCGO="gccgo"
AR="ar"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD=""
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build696469159=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
The following experiments demonstrate that pprof CPU profiles lack accuracy (closeness to the ground truth) and precision (repeatability across different runs). The tests are highly predictive in nature and involve little runtime overheads (allocation, GC, system call, etc.). They are carefully designed so that we can compare pprof CPU profiles against our expectations. The evaluation shows a wide difference from expectation. One should not confuse this to be a runtime issue; the issue is with the use of OS timers used for sampling; OS timers are coarse-grained and have a high skid. In a proposal/design document that will follow, I will propose a design to extend CPU profiling by sampling CPU Performance Monitoring Unit (PMU) aka hardware performance counters. PMU event-based sampling is a mature technology available on almost all modern CPUs. PMU as a sampling agent offers many benefits (in no particular priority order):
* Enhance sampling granularity from 100s of milli-second to 100s of microseconds.
* Avoid precision and accuracy problems of OS timers.
* Enable collecting rich hardware profiles such as cache misses, branch misses, inter-socket cacheline transfers, to name a new, which help root cause performance problems more easily.
There are two test cases `goroutine.go` and `serial.go` in this issue.
## test 1 (parallel)
Download the first test case, `goroutine.go` program, from https://github.com/chabbimilind/GoPprofDemo/blob/master/goroutine.go
It has ten exactly the same goroutines: `f1-f10` and I use pprof to collect the CPU profiles.
Run the following command several times and notice that each time, pprof reports a different amount of time spent in each of the ten routines.
`go run goroutine.go && go tool pprof -top goroutine_prof`
## test 2 (serial)
Download the second test case, `serial.go` go program, from https://github.com/chabbimilind/GoPprofDemo/blob/master/serial.go:
It has ten functions (`A_expect_1_82` - `J_expect_18_18`). The function `A_expect_1_82` is expected to consume 1.82% of the total execution time and `J_expect_18_18` is expected to consume 18.18% of the execution time, and so on. The code is serial and there is complete data dependence between each function and each iteration of the loop in the functions to avoid any hardware-level optimizations.
Run the following command several times.
`go run serial.go && go tool pprof -top serial_prof`
<!--
If possible, provide a recipe for reproducing the error.
A complete runnable program is good.
A link on play.golang.org is best.
-->
### What did you expect to see?
For `goroutine.go`, each function (`f1-10`) should be attributed with exactly (or almost exactly) 10% of the execution time on each run.
For `serial.go` the time attribution should roughly follow the distribution shown below in each run.
```
FUNCTION NAME EXPECTED RELATIVE TIME
A_expect_1_82 1.82%
B_expect_3_64 3.64%
C_expect_5_46 5.46%
D_expect_7_27 7.27%
E_expect_9_09 9.09%
F_expect_10_91 10.91%
G_expect_12_73 12.73%
H_expect_14_546 14.546%
I_expect_16_36 16.36%
J_expect_18_18 18.18%
```
### What did you see instead?
#### test 1 (parallel)
Run 1, 2, and 3, shown below, respectively show the `pprof -top` output of `goroutine.go`.
You will notice in a single run (say Run 1), `f1-f10` have a wide variance in the time attributed to them; the expectation is that each of them gets 10% execution time. There is up to `6x` difference in time attributed to the function with the highest amount if attribution (`main.f7`, `4210ms`, in `Run 1`) vs. the function with the lowest a amount of attribution (`main.f9`, `700ms`, in `Run 1`). This shows a poor accuracy (deviation from the ground truth) of pprof timer-based profiles.
Furthermore, the time attributed to a function widely varies from run to run. Notice how the top-10 ordering changes. In `Run 1`, `main.f7` is shown to run for `4210ms`, whereas in `Run 2` it is shown to run for only `520ms`. The expectation is that the measurements remain the same from run to run. This shows a poor precision (unpredictability of measurement) of pprof timer-based profiles.
#### goroutine.go/Run 1:
```
File: goroutine
Type: cpu
Time: Jan 27, 2020 at 3:45pm (PST)
Duration: 6.70s, Total samples = 18060ms (269.37%)
Showing nodes accounting for 18060ms, 100% of 18060ms total
flat flat% sum% cum cum%
4210ms 23.31% 23.31% 4210ms 23.31% main.f7
2610ms 14.45% 37.76% 2610ms 14.45% main.f2
2010ms 11.13% 48.89% 2010ms 11.13% main.f6
1810ms 10.02% 58.91% 1810ms 10.02% main.f10
1780ms 9.86% 68.77% 1780ms 9.86% main.f3
1410ms 7.81% 76.58% 1410ms 7.81% main.f1
1310ms 7.25% 83.83% 1310ms 7.25% main.f4
1110ms 6.15% 89.98% 1110ms 6.15% main.f5
1110ms 6.15% 96.12% 1110ms 6.15% main.f8
700ms 3.88% 100% 700ms 3.88% main.f9
```
#### goroutine.go/Run 2:
```
File: goroutine
Type: cpu
Time: Jan 27, 2020 at 3:45pm (PST)
Duration: 6.71s, Total samples = 17400ms (259.39%)
Showing nodes accounting for 17400ms, 100% of 17400ms total
flat flat% sum% cum cum%
3250ms 18.68% 18.68% 3250ms 18.68% main.f2
2180ms 12.53% 31.21% 2180ms 12.53% main.f9
2100ms 12.07% 43.28% 2100ms 12.07% main.f1
1770ms 10.17% 53.45% 1770ms 10.17% main.f6
1700ms 9.77% 63.22% 1700ms 9.77% main.f5
1550ms 8.91% 72.13% 1550ms 8.91% main.f4
1500ms 8.62% 80.75% 1500ms 8.62% main.f8
1440ms 8.28% 89.02% 1440ms 8.28% main.f3
1390ms 7.99% 97.01% 1390ms 7.99% main.f10
520ms 2.99% 100% 520ms 2.99% main.f7
```
#### goroutine.go/Run 3:
```
File: goroutine
Type: cpu
Time: Jan 27, 2020 at 3:48pm (PST)
Duration: 6.71s, Total samples = 17.73s (264.31%)
Showing nodes accounting for 17.73s, 100% of 17.73s total
flat flat% sum% cum cum%
3.74s 21.09% 21.09% 3.74s 21.09% main.f7
2.08s 11.73% 32.83% 2.08s 11.73% main.f9
2.05s 11.56% 44.39% 2.05s 11.56% main.f2
1.85s 10.43% 54.82% 1.85s 10.43% main.f10
1.78s 10.04% 64.86% 1.78s 10.04% main.f1
1.43s 8.07% 72.93% 1.43s 8.07% main.f3
1.42s 8.01% 80.94% 1.42s 8.01% main.f8
1.18s 6.66% 87.59% 1.18s 6.66% main.f6
1.17s 6.60% 94.19% 1.17s 6.60% main.f5
1.03s 5.81% 100% 1.03s 5.81% main.f4
```
#### test 2 (serial)
The output for `go run serial.go && go tool pprof -top serial_prof` for three runs is shown below.
Comparing the flat% (or cum%) against the expected percentage for each function shows a large difference. For example, `main.H_expect_14_546`, which is expected to have 14.546% execution time, is attributed 25% of the execution time in Run 1. Furthermore, run to run, there is a lack of precision, for example, `main.I_expect_16_36` is attributed 6.25% (20ms) execution time in Run 1, whereas it is attributed 21.88% (70ms) execution time in Run 2.
#### serial.go/Run 1:
```
File: serial
Type: cpu
Time: Jan 27, 2020 at 1:42pm (PST)
Duration: 501.51ms, Total samples = 320ms (63.81%)
Showing nodes accounting for 320ms, 100% of 320ms total
flat flat% sum% cum cum%
80ms 25.00% 25.00% 80ms 25.00% main.H_expect_14_546
80ms 25.00% 50.00% 80ms 25.00% main.J_expect_18_18
60ms 18.75% 68.75% 60ms 18.75% main.G_expect_12_73
20ms 6.25% 75.00% 20ms 6.25% main.B_expect_3_64
20ms 6.25% 81.25% 20ms 6.25% main.D_expect_7_27
20ms 6.25% 87.50% 20ms 6.25% main.F_expect_10_91
20ms 6.25% 93.75% 20ms 6.25% main.I_expect_16_36
10ms 3.12% 96.88% 10ms 3.12% main.A_expect_1_82
10ms 3.12% 100% 10ms 3.12% main.C_expect_5_46
0 0% 100% 320ms 100% main.main
0 0% 100% 320ms 100% runtime.main
```
#### serial.go/Run 2:
```
File: serial
Type: cpu
Time: Jan 27, 2020 at 1:44pm (PST)
Duration: 501.31ms, Total samples = 320ms (63.83%)
Showing nodes accounting for 320ms, 100% of 320ms total
flat flat% sum% cum cum%
70ms 21.88% 21.88% 70ms 21.88% main.I_expect_16_36
50ms 15.62% 37.50% 50ms 15.62% main.J_expect_18_18
40ms 12.50% 50.00% 40ms 12.50% main.E_expect_9_09
40ms 12.50% 62.50% 40ms 12.50% main.F_expect_10_91
40ms 12.50% 75.00% 40ms 12.50% main.H_expect_14_546
30ms 9.38% 84.38% 30ms 9.38% main.D_expect_7_27
20ms 6.25% 90.62% 20ms 6.25% main.B_expect_3_64
20ms 6.25% 96.88% 20ms 6.25% main.G_expect_12_73
10ms 3.12% 100% 10ms 3.12% main.C_expect_5_46
0 0% 100% 320ms 100% main.main
0 0% 100% 320ms 100% runtime.main
```
#### serial.go/Run 3:
```
File: serial
Type: cpu
Time: Jan 27, 2020 at 1:45pm (PST)
Duration: 501.39ms, Total samples = 310ms (61.83%)
Showing nodes accounting for 310ms, 100% of 310ms total
flat flat% sum% cum cum%
110ms 35.48% 35.48% 110ms 35.48% main.J_expect_18_18
70ms 22.58% 58.06% 70ms 22.58% main.G_expect_12_73
60ms 19.35% 77.42% 60ms 19.35% main.F_expect_10_91
30ms 9.68% 87.10% 30ms 9.68% main.I_expect_16_36
20ms 6.45% 93.55% 20ms 6.45% main.H_expect_14_546
10ms 3.23% 96.77% 10ms 3.23% main.B_expect_3_64
10ms 3.23% 100% 10ms 3.23% main.C_expect_5_46
0 0% 100% 310ms 100% main.main
0 0% 100% 310ms 100% runtime.main
```
### Improved results with PMU-based profiling.
In a prototype PMU-based profiling implementation, below are the pprof profiles for CPU cycles hardware performance counter for the same `goroutine.go` program. Notice that each functions gets the same (or almost the same) CPU cycles attribution within a single run and across runs.
#### goroutine.go/Run 1:
```
File: goroutine
Type: cycles
Time: Jan 27, 2020 at 4:49pm (PST)
Showing nodes accounting for 234000000000, 100% of 234000000000 total
flat flat% sum% cum cum%
23400000000 10.00% 10.00% 23400000000 10.00% main.f1
23400000000 10.00% 20.00% 23400000000 10.00% main.f10
23400000000 10.00% 30.00% 23400000000 10.00% main.f2
23400000000 10.00% 40.00% 23400000000 10.00% main.f3
23400000000 10.00% 50.00% 23400000000 10.00% main.f4
23400000000 10.00% 60.00% 23400000000 10.00% main.f5
23400000000 10.00% 70.00% 23400000000 10.00% main.f6
23400000000 10.00% 80.00% 23400000000 10.00% main.f7
23400000000 10.00% 90.00% 23400000000 10.00% main.f8
23400000000 10.00% 100% 23400000000 10.00% main.f9
```
#### goroutine.go/Run 2:
```
File: goroutine
Type: cycles
Time: Jan 27, 2020 at 4:51pm (PST)
Showing nodes accounting for 234000000000, 100% of 234000000000 total
flat flat% sum% cum cum%
23800000000 10.17% 10.17% 23800000000 10.17% main.f1
23500000000 10.04% 20.21% 23500000000 10.04% main.f7
23500000000 10.04% 30.26% 23500000000 10.04% main.f9
23400000000 10.00% 40.26% 23400000000 10.00% main.f10
23400000000 10.00% 50.26% 23400000000 10.00% main.f2
23400000000 10.00% 60.26% 23400000000 10.00% main.f4
23400000000 10.00% 70.26% 23400000000 10.00% main.f6
23400000000 10.00% 80.26% 23400000000 10.00% main.f8
23300000000 9.96% 90.21% 23300000000 9.96% main.f3
22900000000 9.79% 100% 22900000000 9.79% main.f5
```
Below are the pprof profiles for CPU cycles hardware performance counter for the same `serial.go` program. Notice that each function gets close to expected CPU cycles attribution within a single run and across runs.
#### serial.go/Run 1:
```
File: serial
Type: cycles
Time: Jan 27, 2020 at 4:54pm (PST)
Showing nodes accounting for 1105000000, 100% of 1105000000 total
flat flat% sum% cum cum%
200000000 18.10% 18.10% 200000000 18.10% main.J_expect_18_18
183000000 16.56% 34.66% 183000000 16.56% main.I_expect_16_36
165000000 14.93% 49.59% 165000000 14.93% main.H_expect_14_546
137000000 12.40% 61.99% 137000000 12.40% main.G_expect_12_73
120000000 10.86% 72.85% 120000000 10.86% main.F_expect_10_91
100000000 9.05% 81.90% 100000000 9.05% main.E_expect_9_09
82000000 7.42% 89.32% 82000000 7.42% main.D_expect_7_27
63000000 5.70% 95.02% 63000000 5.70% main.C_expect_5_46
37000000 3.35% 98.37% 37000000 3.35% main.B_expect_3_64
18000000 1.63% 100% 18000000 1.63% main.A_expect_1_82
0 0% 100% 1105000000 100% main.main
0 0% 100% 1105000000 100% runtime.main
```
#### serial.go/Run 2:
```
File: serial
Type: cycles
Time: Jan 27, 2020 at 4:54pm (PST)
Showing nodes accounting for 1105000000, 100% of 1105000000 total
flat flat% sum% cum cum%
200000000 18.10% 18.10% 200000000 18.10% main.J_expect_18_18
183000000 16.56% 34.66% 183000000 16.56% main.I_expect_16_36
159000000 14.39% 49.05% 159000000 14.39% main.H_expect_14_546
142000000 12.85% 61.90% 142000000 12.85% main.G_expect_12_73
119000000 10.77% 72.67% 119000000 10.77% main.F_expect_10_91
100000000 9.05% 81.72% 100000000 9.05% main.E_expect_9_09
82000000 7.42% 89.14% 82000000 7.42% main.D_expect_7_27
61000000 5.52% 94.66% 61000000 5.52% main.C_expect_5_46
40000000 3.62% 98.28% 40000000 3.62% main.B_expect_3_64
18000000 1.63% 99.91% 18000000 1.63% main.A_expect_1_82
1000000 0.09% 100% 1105000000 100% main.main
0 0% 100% 1105000000 100% runtime.main
```
### Dependence on the number of cores and length of test execution:
The results of `goroutine.go` test depend on the number of CPU cores available. On a multi-core CPU, if you set `GOMAXPROCS=1`, `goroutine.go` will not show a huge variation, since each goroutine runs for several seconds. However, if you set `GOMAXPROCS` to a larger value, say 4, you will notice a significant measurement attribution problem. One reason for this problem is that the itimer samples on Linux are not guaranteed to be delivered to the thread whose timer expired.
The results of `serial.go` can change based on the time of execution of each function. By passing the `-m=<int>` to the program, you can make the program run for longer or shorter. By making it run longer, the profiles can be made more accurate, but when a function runs for less than 100ms, the accuracy is often low. | Proposal,Proposal-Hold | high | Critical |
555,952,490 | TypeScript | Auto-format incorrectly formats multi-line arrays | *TS Template added by @mjbvz*
**TypeScript Version**: 3.8.0-dev.20200125
**Search Terms**
- format
- array
- multiline
---
Version: 1.41.0 (system setup)
Commit: 9579eda04fdb3a9bba2750f15193e5fafe16b959
Date: 2019-12-11T18:37:42.077Z
Electron: 6.1.5
Chrome: 76.0.3809.146
Node.js: 12.4.0
V8: 7.6.303.31-electron.0
OS: Windows_NT x64 10.0.17763
Steps to Reproduce:
1. Paste the following code into a TypeScript file:
```
const array0 = ["banana",
"apple",
"pear",
"peach"
];
const array1 = [String("banana"),
String("apple"),
String("pear"),
String("peach")
];
const array2 = [
String("banana"),
String("apple"),
String("pear"),
String("peach")
];
const peach = String("peach");
const array3 = [String("banana"),
String("apple"),
"pear",
peach
];
```
2. Save with auto format on save enabled.
3. The code array entries will be inconsistently indented as shown below:
```
const array0 = ["banana",
"apple",
"pear",
"peach"
];
const array1 = [String("banana"),
String("apple"),
String("pear"),
String("peach")
];
const array2 = [
String("banana"),
String("apple"),
String("pear"),
String("peach")
];
const peach = String("peach");
const array3 = [String("banana"),
String("apple"),
"pear",
peach
];
```
Primitives and variables are indented but method calls and constructors are not indented. This only happens if the first entry in the array is on the same line as the opening bracket. I would expect the text to be formatted as shown in step 1.
| Bug,Domain: Formatter | low | Minor |
555,964,661 | rust | casting trait methods to FnOnce(&mut dyn Trait) is not ergonomic | In the following code, I'm trying to cast `Trait::f` to `FnOnce(&mut dyn Trait)` which should be possible, but the obvious method doesn't work. None of the combinations I tried got it to work (I could just use a lambda that called `f`, but that defeats the purpose).
```rust
trait Trait {
fn f(&mut self);
}
fn g<F: FnOnce(&mut dyn Trait)>(f: F) {}
fn f() {
g(Trait::f)
}
```
([Playground](https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=def622e8ed38388cb543e450467bdc00))
Errors:
```
Compiling playground v0.0.1 (/playground)
error[E0631]: type mismatch in function arguments
--> src/lib.rs:8:7
|
2 | fn f(&mut self);
| ---------------- found signature of `for<'r> fn(&'r mut _) -> _`
...
5 | fn g<F: FnOnce(&mut dyn Trait)>(f: F) {}
| - ---------------------- required by this bound in `g`
...
8 | g(Trait::f)
| ^^^^^^^^ expected signature of `for<'r> fn(&'r mut (dyn Trait + 'r)) -> _`
error[E0271]: type mismatch resolving `for<'r> <for<'s> fn(&'s mut _) {<_ as Trait>::f} as std::ops::FnOnce<(&'r mut (dyn Trait + 'r),)>>::Output == ()`
--> src/lib.rs:8:5
|
5 | fn g<F: FnOnce(&mut dyn Trait)>(f: F) {}
| - ---------------------- required by this bound in `g`
...
8 | g(Trait::f)
| ^ expected bound lifetime parameter, found concrete lifetime
error: aborting due to 2 previous errors
For more information about this error, try `rustc --explain E0271`.
error: could not compile `playground`.
To learn more, run the command again with --verbose.
``` | C-enhancement,A-diagnostics,T-compiler | low | Critical |
555,981,973 | pytorch | Building PyTorch ignores my current Python version | ## π Bug
<!-- A clear and concise description of what the bug is. -->
I am trying to build PyTorch on my Raspberry Pi 4(Raspbian Buster) with Python 3.6.9(virtual env), but it keeps using Python 3.7
## To Reproduce
Steps to reproduce the behavior:
1. git clone https://github.com/pytorch/pytorch.git
2. git checkout v1.3.1
3. sudo python setup.py build
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->

## Expected behavior
<!-- A clear and concise description of what you expected to happen. -->
I am expecting it copys files from python 3.6.9 virtual environment.
## Environment
> Collecting environment information...
> PyTorch version: N/A
> Is debug build: N/A
> CUDA used to build PyTorch: N/A
>
> OS: Raspbian GNU/Linux 10 (buster)
> GCC version: (Raspbian 8.3.0-6+rpi1) 8.3.0
> CMake version: version 3.13.4
>
> Python version: 3.6
> Is CUDA available: N/A
> CUDA runtime version: Could not collect
> GPU models and configuration: Could not collect
> Nvidia driver version: Could not collect
> cuDNN version: Could not collect
>
> Versions of relevant libraries:
> [pip] Could not collect
> [conda] Could not collect
>
## Additional context
<!-- Add any other context about the problem here. -->
| module: build,triaged | low | Critical |
556,098,557 | vscode | Rename refactoring context actions work on the entire input | Testing https://github.com/microsoft/vscode/issues/89331

1. Right click a refactor row and choose `Apply Refactoring`
All the checked changes will be applied and the view will disappear. I would expect only the refactoring I've right clicked on would be applied.
| ux,under-discussion,workspace-edit | low | Minor |
556,105,430 | pytorch | How to customize build torchscript model to be used in end devices codebase | ## π Feature
I want to compile my model to be executed in the Python/C script running on our customers computers/end devices, without the need to load the entire torch/libtorch package, but only what is needed based on the model operations.
## Motivation
Currently, the size of my ResNet model (for example) is ~100MB but it needs torch/libtorch, which requires ~1.5GB of space.
End devices (smart cameras, robots, etc.) are low in resources. R&D efforts for deployment on end devices includes a large efforts to optimize the model and reduce its size to minimum. Having my model accompanied by torch/libtorch is a difficult restriction. I am aware that the mobile community is leading the attention for similar features. However, considering modern smartphones resources, there is even a greater need for such a solution for other end devices.
## Current status
Currently i am doing this series of commands:
`model = torchvision.models.resnet50(pretrained=True)`
`model.eval()`
`example = torch.ones(1, 3, 224, 224)`
`traced_model = torch.jit.trace(model, example)`
`ops = torch.jit.export_opnames(model)`
`traced_model.save('traced_model.pt')`
`with open('model_ops.yaml', 'w') as output:`
` yaml.dump(ops, output)`
The request is to enable building a model i can use in another python/c script without the need to load the entire torch or libtorch packages, but only what is needed based on the model operations.
## Alternatives
I am not aware of such alternatives. Will be happy to hear about them, if there are any.
cc @suo | oncall: jit,triaged,oncall: mobile | low | Major |
556,180,231 | pytorch | ninja: build stopped: subcommand failed. | ## π Bug
Does anyone have a recipe that is known to work?
<!-- A clear and concise description of what the bug is. -->
[426/3392] Building CXX object third_party/ideep/m...nn/src/CMakeFiles/mkldnn.dir/cpu/cpu_reorder.cpp.o
ninja: build stopped: subcommand failed.
Traceback (most recent call last):
File "setup.py", line 737, in <module>
build_deps()
File "setup.py", line 316, in build_deps
cmake=cmake)
File "/home/developer/pytorch/tools/build_pytorch_libs.py", line 62, in build_caffe2
cmake.build(my_env)
File "/home/developer/pytorch/tools/setup_helpers/cmake.py", line 339, in build
self.run(build_args, my_env)
File "/home/developer/pytorch/tools/setup_helpers/cmake.py", line 141, in run
check_call(command, cwd=self.build_dir, env=env)
File "/home/developer/anaconda3/envs/pytorch/lib/python3.7/subprocess.py", line 363, in check_call
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['cmake', '--build', '.', '--target', 'install', '--config', 'Release', '--', '-j', '32']' returned non-zero exit status 1.
## To Reproduce
Steps to reproduce the behavior:
python setup.py clean
python setup..py install
1.
1.
1.
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
## Expected behavior
builds successfully
<!-- A clear and concise description of what you expected to happen. -->
## Environment
Ubuntu 14.04 with Xenial kernel 4.4.0 generic
Anaconda, Python 3.7
Tried C++ 4.9, 5.5, 7, 9
CUDA 10.1, cdnn 7.5.6
Please copy and paste the output from our
[environment collection script](https://raw.githubusercontent.com/pytorch/pytorch/master/torch/utils/collect_env.py)
(or fill out the checklist below manually).
You can get the script and run it with:
```
wget https://raw.githubusercontent.com/pytorch/pytorch/master/torch/utils/collect_env.py
# For security purposes, please check the contents of collect_env.py before running it.
python collect_env.py
```
- PyTorch Version (e.g., 1.0):
- OS (e.g., Linux):
- How you installed PyTorch (`conda`, `pip`, source):
- Build command you used (if compiling from source):
- Python version:
- CUDA/cuDNN version:
- GPU models and configuration:
- Any other relevant information:
## Additional context
<!-- Add any other context about the problem here. -->
| module: build,triaged | low | Critical |
556,184,980 | pytorch | torch.tensordot has inconsistent signature with torch script | ## π Bug
<!-- A clear and concise description of what the bug is. -->
Inconsistent API across torch script and no script
## To Reproduce
Steps to reproduce the behavior:
Working samples
```python
def my_tensordot(a, b):
return torch.tensordot(a, b, [[1], [1]])
```
```python
@torch.jit.script
def my_tensordot(a, b):
return torch.tensordot(a, b, [1], [1])
```
Broken one (using same API as in python)
```python
@torch.jit.script
def my_tensordot(a, b):
return torch.tensordot(a, b, [[1], [1]])
```
I get
```
aten::tensordot(Tensor self, Tensor other, int[] dims_self, int[] dims_other) -> (Tensor):
Expected a value of type 'List[int]' for argument 'dims_self' but instead found type 'List[List[int]]'.
```
Removing jit decorator from working jit example does not help
```python
def my_tensordot(a, b):
return torch.tensordot(a, b, [1], [1])
```
```
TypeError: tensordot() takes from 2 to 3 positional arguments but 4 were given
```
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
## Expected behavior
Same API should work in jit and non jit mode, like here
```python
@torch.jit.script
def my_tensordot(a, b):
return torch.tensordot(a, b, [[1], [1]])
```
## Environment
Please copy and paste the output from our
[environment collection script](https://raw.githubusercontent.com/pytorch/pytorch/master/torch/utils/collect_env.py)
(or fill out the checklist below manually).
You can get the script and run it with:
```
PyTorch version: 1.3.1
Is debug build: No
CUDA used to build PyTorch: 10.1.243
OS: Ubuntu 16.04.6 LTS
GCC version: (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609
CMake version: Could not collect
Python version: 3.6
Is CUDA available: No
CUDA runtime version: Could not collect
GPU models and configuration: GPU 0: GeForce GTX 950M
Nvidia driver version: 384.130
cuDNN version: Could not collect
Versions of relevant libraries:
[pip3] numpy==1.17.1
[pip3] torch==1.3.1
[pip3] torch-cluster==1.4.5
[pip3] torch-geometric==1.3.2
[pip3] torch-scatter==1.4.0
[pip3] torch-sparse==0.4.3
[pip3] torch-spline-conv==1.1.1
[pip3] torchvision==0.4.2
[conda] Could not collect
```
## Additional context
I found this running python with `PYTORCH_JIT=false`
## Possible Solutions
1. Overload both functions to handle both cases. No user code change required, errors dissapear
2. Jit supports python API variant: few users are affected. pytorch version should be checked at import time
2. Python function supports jit API variant: a lot of users are affected. pytorch version should be checked at import time
cc @suo | oncall: jit,triaged | low | Critical |
556,196,607 | pytorch | Add Julia Bindings to Torch backend | ## π Feature
Usage of Julia to access Torch backend apiβs.
## Motivation
Array Programming is a paradigm, just like saying βObject Orientated Programmingβ or βFunctional Programmingβ, along w/ a compiler that can handle Dispatch and mathematical types makes it a language perfect for deep learning
## Pitch
Create Julia bindings to access torch backend while leveraging the capabilities of the Julia language itself.
## Additional context
Include GPU support for NVIDIA/CUDA. (Julia has this)
| feature,low priority,triaged,module: language binding | low | Major |
556,228,648 | TypeScript | Catch when callbacks are missing a return. | <!-- π¨ STOP π¨ π¦π§π’π£ π¨ πΊπ»πΆπ· π¨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker.
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ, especially the "Common Feature Requests" section: https://github.com/Microsoft/TypeScript/wiki/FAQ
-->
## Search Terms
<!-- List of keywords you searched for before creating this issue. Write them down here so that others can find this suggestion more easily -->
- array callback return
- eslint array-callback-return
- catch missing return for callbacks
## Suggestion
TypeScript does not recognise this as an error even though it clearly is - `findIndex` takes a callback which should return a boolean, returning an (implicit) `undefined` is IMO in 99% of cases a mistake.
```ts
[1,2,3].findIndex(e => {
e === 1
})
```
I also think that if there is an *explicit return* of a falsey value such as `null` or `undefined`, then that should be fine. It's the lack of any return at all which should trigger an error, in my eyes.
I'm aware that ESLint can do this but not everyone uses ESLint, also ESLint can be a huge pain to configure sometimes, especially when Prettier, TypeScript and other things are involved.
## Use Cases
It will definitely catch errors, IMHO, for people who don't use ESLint or do not have the rule `array-callback-return`.
## Examples
No error:
```ts
[1,2,3].findIndex(e =>
e === 1
)
```
No error:
```ts
[1,2,3].findIndex(e => e === 1)
```
A warning/errror: "Did you forget a return?"
```ts
[1,2,3].findIndex(e => {
e === 1
})
```
## Checklist
My suggestion meets these guidelines:
* [x] This wouldn't be a breaking change in existing TypeScript/JavaScript code
* [x] This wouldn't change the runtime behavior of existing JavaScript code
* [x] This could be implemented without emitting different JS based on the types of the expressions
* [x] This isn't a runtime feature (e.g. library functionality, non-ECMAScript syntax with JavaScript output, etc.)
* [x] This feature would agree with the rest of [TypeScript's Design Goals](https://github.com/Microsoft/TypeScript/wiki/TypeScript-Design-Goals).
| Suggestion,Needs Proposal | low | Critical |
556,248,161 | material-ui | [Modal] aria-hidden should not be applied on non-portal modals | <!-- Provide a general summary of the issue in the Title above -->
`aria-hidden=true` gets applied on top level container when a modal is opened. But if `disablePortal` is used, `aria-hidden` should not be set on that top level container as the modal will be created inline and most likely within that top level container
Running axe on the opened modal yields the following error:
> "ARIA hidden element must not contain focusable elements (aria-hidden-focus)"
>
> Fix all of the following:
Focusable content should have tabindex='-1' or be removed from the DOM
You can find more information on this issue here:
https://dequeuniversity.com/rules/axe/3.3/aria-hidden-focus?application=axeAPI
https://www.w3.org/TR/wai-aria-practices/examples/dialog-modal/dialog.html
> The dialog element is not a descendant of any element that has aria-hidden set to true.
<!--
Thank you very much for contributing to Material-UI by creating an issue! β€οΈ
To avoid duplicate issues we ask you to check off the following list.
-->
<!-- Checked checkbox should look like this: [x] -->
- [x] The issue is present in the latest release.
- [x] I have searched the [issues](https://github.com/mui-org/material-ui/issues) of this repository and believe that this is not a duplicate.
## Current Behavior π―
1. Modal with `disablePortal` prop is opened and `aria-hidden` is set to true on top level html node
## Expected Behavior π€
1. Modal with `disablePortal` prop is opened and `aria-hidden` is _not_ set to true on top level html node
2. Modal _without_ `disablePortal` prop is opened and `aria-hidden` is set to true on top level html node
## Steps to Reproduce πΉ
https://codesandbox.io/s/material-mui-demo-9oddt?fontsize=14&hidenavigation=1&theme=dark
## Context π¦
<!--
What are you trying to accomplish? How has this issue affected you?
Providing context helps us come up with a solution that is most useful in the real world.
-->
## Your Environment π
<!--
Include as many relevant details about the environment with which you experienced the bug.
If you encounter issues with typescript please include version and tsconfig.
-->
| Tech | Version |
| ----------- | ------- |
| Material-UI | v4.9.0 |
| React | v16.12.0 |
| Browser | |
| TypeScript | v3.8.0-dev.20200125 |
| etc. | |
| accessibility,component: modal | medium | Critical |
556,253,588 | godot | is_rel_path returns true for an empty string | **Godot version:** 505fee0b6db806c9fc83acd3a074f40ac57a83e2
**Issue description:**
`"".is_rel_path()` returns `True`
I don't believe "" should be considered a relative path...
**Minimal reproduction project:**
[test.zip](https://github.com/godotengine/godot/files/4122887/test.zip)
| discussion,topic:core | low | Minor |
556,268,291 | vscode | CodeIcon: in documentation hover could get some polish | Refs: https://github.com/microsoft/vscode/issues/89328
I think once a codicon is in the text, we need to adjust the lineheight or some other styling to make it look nicer:

| ux,polish,suggest,workbench-hover | low | Minor |
556,312,439 | rust | E0700 does not point to where the captured lifetime is | The error message
```rust
error[E0700]: hidden type for `impl Trait` captures lifetime that does not appear in bounds
```
does not point to where in the block the captured lifetime is (which makes debugging it difficult).
It does, however, point to the scope:
```rust
note: hidden type `impl futures::Future` captures the scope of call-site for function at 48:40
--> rs/agent-updater/src/fetcher.rs:48:40
|
48 | ) -> Result<PathBuf, FetcherError> {
| ________________________________________^
49 | | let checksum = m.get_valid_checksum()?;
51 | |
... |
72 | | Ok(folder)
73 | | }
| |_____^
``` | C-enhancement,A-diagnostics,T-compiler,A-async-await,AsyncAwait-Triaged,D-papercut,S-has-mcve | low | Critical |
556,384,027 | pytorch | Torch serialization does not restore tensors properly when custom __reduce__ is defined | ## π Bug
If a `nn.Module` define a custom `__reduce__` method to customize serialization, with the `__reduce__` function returning some tensors (like a state dict), then upon deserialization, the tensors will not contain the right values.
## To Reproduce
```python
import pickle
import torch as th
def stable_instantiate(klass, args, kwargs, state):
self = klass(*args, **kwargs)
print("loading", state)
self.load_state_dict(state)
return self
def stable_pickle(klass):
old_init = klass.__init__
def __init__(self, *args, **kwargs):
self._init_args_kwargs = (args, kwargs)
old_init(self, *args, **kwargs)
def __reduce__(self):
print("reducing", self.state_dict())
return (stable_instantiate,
tuple([klass] + list(self._init_args_kwargs) + [self.state_dict()]))
klass.__init__ = __init__
klass.__reduce__ = __reduce__
return klass
@stable_pickle
class MyModule(th.nn.Module):
def __init__(self, dim):
super().__init__()
self.linear = th.nn.Linear(dim, 1, bias=False)
def forward(self, x):
return self.linear(x)
dim = 4
model = MyModule(dim)
model.linear.weight.data[0] = th.arange(dim).float()
print("model weights", model.linear.weight.data)
print("Trying to serialize with torch.save/load")
th.save(model, "tmp.th")
model2 = th.load("tmp.th")
print("loaded model weights", model2.linear.weight.data)
print("Trying to serialize with pickle")
pickle.dump(model, open("tmp.pckl", "wb"))
model2 = pickle.load(open("tmp.pckl", "rb"))
print("unpickled model weights", model2.linear.weight.data)
```
Steps to reproduce the behavior:
1. Save the above code as `bug.py`
1. Run `python3 bug.py`
1. Observe how when calling `torch.load`, the value of the state dict is just random and not what was set before saving.
```bash
model weights tensor([[0., 1., 2., 3.]])
Trying to serialize with torch.save/load
reducing OrderedDict([('linear.weight', tensor([[0., 1., 2., 3.]]))])
loading OrderedDict([('linear.weight', tensor([[-5.6200e-26, 3.0820e-41, -5.5415e-26, 3.0820e-41]]))])
loaded model weights tensor([[-5.6200e-26, 3.0820e-41, -5.5415e-26, 3.0820e-41]])
Trying to serialize with pickle
reducing OrderedDict([('linear.weight', tensor([[0., 1., 2., 3.]]))])
loading OrderedDict([('linear.weight', tensor([[0., 1., 2., 3.]]))])
unpickled model weights tensor([[0., 1., 2., 3.]])
```
## Expected behavior
Expected `torch.save` and `torch.load` to serialize the state dict returned by `__reduce__`. I'm not sure if the problem is at saving or loading but at some point, all the information is lost and a uninitialized tensor is used instead. `pickle` works like expected.
## Environment
Please copy and paste the output from our
[environment collection script](https://raw.githubusercontent.com/pytorch/pytorch/master/torch/utils/collect_env.py)
(or fill out the checklist below manually).
You can get the script and run it with:
```
wget https://raw.githubusercontent.com/pytorch/pytorch/master/torch/utils/collect_env.py
# For security purposes, please check the contents of collect_env.py before running it.
python collect_env.py
```
- PyTorch Version: 1.4.0
- OS (e.g., Linux): Ubuntu 18.04.3 LTS
- How you installed PyTorch (`conda`, `pip`, source): conda
- Build command you used (if compiling from source): N/A
- Python version: 3.8.1
- CUDA/cuDNN version: N/A
- GPU models and configuration: N/A
- Any other relevant information: Nope
| module: serialization,triaged | low | Critical |
556,386,059 | go | net/http: Dial I/O Timeout even when request context is not canceled | Seems related to https://github.com/golang/go/commit/869e576517f825aecdc8730b0d22f8d6b59bd749 (c.f. https://github.com/kubernetes-sigs/aws-encryption-provider/issues/61).
Maybe this is an expected behavior... Is it possible to see dial timeouts even when the request context has not been canceled? I am using https://golang.org/pkg/net/http/#RoundTripper with dial timeout 30 seconds and request context is 300Β΅s. If this is possible, can you help understand why this happens?
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13.6 darwin/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes.
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/Users/leegyuho/Library/Caches/go-build"
GOENV="/Users/leegyuho/Library/Application Support/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="darwin"
GONOPROXY="*"
GONOSUMDB="*"
GOOS="darwin"
GOPATH="/Users/leegyuho/go"
GOPRIVATE="*"
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/usr/local/go"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/usr/local/go/pkg/tool/darwin_amd64"
GCCGO="gccgo"
AR="ar"
CC="clang"
CXX="clang++"
CGO_ENABLED="1"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=/var/folders/y_/_dn293xd5kn7xlg6jvp7jpmxs99pm9/T/go-build128444364=/tmp/go-build -gno-record-gcc-switches -fno-common"
</pre></details>
### What did you do?
```go
package main
import (
"context"
"fmt"
"net/http"
"net/http/httptest"
"strings"
"time"
)
func main() {
// this happens about 50% of the time, so try it a few times.
for i := 0; i < 10; i++ {
fmt.Println("trying", i)
try()
}
}
func try() {
mux := http.NewServeMux()
mux.HandleFunc("/test", func(w http.ResponseWriter, req *http.Request) {
switch req.Method {
case "GET":
fmt.Fprint(w, `test`)
default:
http.Error(w, "Method Not Allowed", 405)
}
})
ts := httptest.NewServer(mux)
defer ts.Close()
u := ts.URL + "/test"
// default dial timeout is 30-second
// https://golang.org/pkg/net/http/#RoundTripper
cli := http.DefaultClient
timeout := 300 * time.Microsecond
reqs := 20
errc := make(chan error, reqs)
for i := 0; i < reqs; i++ {
go func() {
ctx, cancel := context.WithTimeout(context.TODO(), timeout)
defer cancel()
req, err := http.NewRequest(http.MethodGet, u, nil)
if err != nil {
errc <- err
return
}
_, err = cli.Do(req.WithContext(ctx))
if err != nil {
// can be: ctx.Err() == nil && err == "i/o timeout"
// Q. but how is that possible?
fmt.Println("Do failed with", err, "/ context error:", ctx.Err())
}
errc <- err
}()
}
// "context deadline exceeded" for requests after timeout
exp := `context deadline`
for i := 0; i < reqs; i++ {
err := <-errc
if err == nil {
continue
}
fmt.Println("error:", err)
if !strings.Contains(err.Error(), exp) {
panic(fmt.Sprintf("#%d: expected %q, got %v", i, exp, err))
}
}
}
```
### What did you expect to see?
No panic.
### What did you see instead?
```
trying 0
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
error: Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
Do failed with Get http://127.0.0.1:54098/test: context deadline exceeded / context error: context deadline exceeded
Do failed with Get http://127.0.0.1:54098/test: context deadline exceeded / context error: context deadline exceeded
Do failed with Get http://127.0.0.1:54098/test: context deadline exceeded / context error: context deadline exceeded
Do failed with Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout / context error: <nil>
panic: #0: expected "context deadline", got Get http://127.0.0.1:54098/test: dial tcp 127.0.0.1:54098: i/o timeout
```
| NeedsInvestigation | medium | Critical |
556,395,845 | TypeScript | @template is lost when importing method | <!-- π¨ STOP π¨ π¦π§π’π£ π¨ πΊπ»πΆπ· π¨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker. Even if you think you've found a *bug*, please read the FAQ first, especially the Common "Bugs" That Aren't Bugs section!
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ: https://github.com/Microsoft/TypeScript/wiki/FAQ
Please fill in the *entire* template below.
-->
<!--
Please try to reproduce the issue with the latest published version. It may have already been fixed.
For npm: `typescript@next`
This is also the 'Nightly' version in the playground: http://www.typescriptlang.org/play/?ts=Nightly
-->
**TypeScript Version:** 3.6.3
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
template
**Code**
```ts
const $indicatrix = require('./indicatrix')
/**
* Will print the loading text and refresh the CLI line to show the ellipsis while the promise is loading.
* @param {string} text The text to display in the CLI.
* @param {!Promise<T>|function(...*): !Promise<T>} promise The promise or an async function that returns the promise.
* @return {Promise<T>}
* @template T
*/
function indicatrix(text, promise, options) {
return $indicatrix(text, promise, options)
}
module.exports = indicatrix
```
import this from another file
```js
/**
* @type {_indicatrix.indicatrix}
*/
async function $indicatrix(text, promise, options = {}) {
// implementation ...
}
/**
* @suppress {nonStandardJsDocs}
* @typedef {import('..')} _indicatrix.indicatrix
*/
```
**Expected behavior:**
supposed to be like in the original location

**Actual behavior:**
the template bit is lost

| Needs Investigation | low | Critical |
556,421,488 | rust | UnixDatagram: please support vectored send | The `UnixDatagram` type provides `send` and `send_to` methods, but does not provide a vectored send operation. Such an operation is particularly helpful for datagrams, where separate send operations result in separate datagrams; a vectored send allows sending data from multiple buffers in a single datagram.
Given that `UnixDatagram` only runs on UNIX, where `writev` works just fine on a socket, such an operation could use `writev`; alternatively, this could use `sendmsg`, which supports supplying an iovec. `sendmsg` would also allow a vectored `send_to` operation. | T-libs-api,C-feature-request | low | Minor |
556,430,617 | go | cmd/compile: error line number reported is incorrect if it appears after line 0xFFFFF | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13.7 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes.
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/home/icza/.cache/go-build"
GOENV="/home/icza/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GONOPROXY="xxx"
GONOSUMDB="xxx"
GOOS="linux"
GOPATH="/home/icza/gows"
GOPRIVATE="xxx"
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/usr/local/go"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/usr/local/go/pkg/tool/linux_amd64"
GCCGO="gccgo"
AR="ar"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD=""
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build104015454=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
Create a valid `.go` source file that has more than `1048574` lines (`0xFFFFF`). And add some "garbage" to the end of it to make it invalid, to make compilation fail. Attempting to build it (`go build big.go`) or run it (`go run big.go`) the go tool detects the error and displays an error message where the line number is capped at `1048574` improperly, e.g.:
./big.go:1048574:2: syntax error: unexpected invalid after top level declaration
Here's a simple Go app that creates a `big.go` file that has approximately twice as many rows (it has a raw string literal that has 2*0xFFFFF empty lines):
```
package main
import (
"bytes"
"io/ioutil"
"strings"
)
func main() {
src := &bytes.Buffer{}
src.WriteString(`package main
func main() { println(len(s)) }
const s=`)
src.WriteString("`")
src.WriteString(strings.Repeat("\n", 2*0xFFFFF))
src.WriteString("`")
// This is the garbage to make compilation fail:
src.WriteString("invalid")
if err := ioutil.WriteFile("big.go", src.Bytes(), 0666); err != nil {
panic(err)
}
}
```
If we decrease the generated number of lines, e.g.:
src.WriteString(strings.Repeat("\n", 10_000))
We get the proper line number (`10003`):
./big.go:10003:2: syntax error: unexpected invalid after top level declaration
### What did you expect to see?
I expected the proper line number to be printed where the error is detected (`2*0xFFFFFF+3 = 2097153`):
./big.go:2097153:2: syntax error: unexpected invalid after top level declaration
### What did you see instead?
The displayed error line number is improperly capped at `1048574`:
./big.go:1048574:2: syntax error: unexpected invalid after top level declaration
**Note:** This issue was reported on StackOverflow: https://stackoverflow.com/questions/59951140/go-compiler-error-line-number-is-incorrect | NeedsInvestigation,compiler/runtime | low | Critical |
556,475,404 | rust | UdpSocket: please support vectored send | The `UdpSocket` type provides `send` and `send_to` methods, but does not provide a vectored send operation. Such an operation is particularly helpful for datagrams, where separate send operations result in separate datagrams; a vectored send allows sending data from multiple buffers in a single datagram.
This would require calling `sendmsg`, which supports supplying an iovec. (This should work on Windows as well.) | T-libs-api,C-feature-request | low | Minor |
556,514,751 | godot | In an HBoxContainer a Viewport with size 0 in a stretch-enabled VIewportcontainer does not render until first resize | **Godot version:**
3.1.2
**OS/device including version:**
OpenSuse Tumbleweed
**Issue description:**
is related to https://github.com/godotengine/godot/issues/26187 and can be easily fixed setting the viewport size to something good in the first place but maybe this helps someone figure out https://github.com/godotengine/godot/issues/34805
**Steps to reproduce:**
make a 3d scene and a 2d scene, put a viewportcontainer in a hboxcontainer and make them resize and enable stretch,
put the viewport in the viewportcontainer and put the viewports size to 0,0, notice how it doesn't render on first frames,
resize the window to resize all the containers
notice how it now works :D
**Minimal reproduction project:**
[testproject.zip](https://github.com/godotengine/godot/files/4124967/testproject.zip) | bug,confirmed,topic:gui | low | Minor |
556,527,194 | TypeScript | Documentation returned from constant type instead of from the constant itself | <!-- π¨ STOP π¨ π¦π§π’π£ π¨ πΊπ»πΆπ· π¨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker. Even if you think you've found a *bug*, please read the FAQ first, especially the Common "Bugs" That Aren't Bugs section!
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ: https://github.com/Microsoft/TypeScript/wiki/FAQ
Please fill in the *entire* template below.
-->
From https://github.com/microsoft/vscode/issues/89508
<!--
Please try to reproduce the issue with the latest published version. It may have already been fixed.
For npm: `typescript@next`
This is also the 'Nightly' version in the playground: http://www.typescriptlang.org/play/?ts=Nightly
-->
**TypeScript Version:** 3.8.0-dev.20200128
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
- jsdoc
- quickinfo
- signatureHelp
**Code**
For the code:
```ts
/**
* Interface!
*/
interface Handler<T> {
/**
* Call!
*/
(): void;
}
/**
* Instance!
*/
declare const catHandler: Handler<string>;
catHandler()
```
1. Hover over `catHandler()`
**Expected behavior:**
Documentation `Instance!` is returned
**Actual behavior:**
Documentation for `Call!` returned:
<img width="429" alt="Screen Shot 2020-01-28 at 2 47 01 PM" src="https://user-images.githubusercontent.com/12821956/73312221-0ffc6b00-41dd-11ea-93fc-39b212e0afa8.png">
/cc @Tyriar | Suggestion,Awaiting More Feedback | low | Critical |
556,536,384 | rust | StorageLive (and even StorageDead) may be unnecessary in MIR. | A while back I was discussing `Storage{Live,Dead}` and dominators, with @tmandry (in the context of generator layout optimizations), and came to the conclusion that `StorageLive` pretty much *has to* dominate all uses (I doubt we ever added a check that it does so, though).
More recently, I was trying to figure out what the simplest "`StorageLive` sinking" (i.e. moving the statement "later" in the CFG) optimization we could do was.
The conclusion I came to was that we might not need `StorageLive` at all, because there might be a deterministic "best placement" we could compute (assuming we need *exactly* one `llvm.lifetime.start` per `alloca`).
<hr/>
That best placement would be the *least (common) dominator* of all mentions of a MIR local.
Even indirect accesses require a direct borrow beforehand, so this should cover everything.
(Assuming that, given CFG points `x`, `y`, `z`, "`x` is a common dominator of `y` and `z`" means "`x` dominates both `y` and `z`", i.e. "to reach either `y` or `z` you must go through `x` first", and the "least" such `x` is the one not dominating other common dominators of `y` and `z`, i.e. it's "the closest to `y` and `z`")
This could be:
* just before the single assignment of that local
* `let x = x + y;`
* at the end of a block branching into paths which all assign that local
* `let x = if c { a } else { b };`
* `let x; if c { x = a; } else { x = b; }` (roughly equivalent)
I am not sure about interactions with loops, though.
But this doesn't have to remain theoretical, we could compute this "ideal `StorageLive` position" and then compare it with the existing one (presumably one would dominate the other? not sure this would catch any loop issues though).
<hr/>
`StorageDead` could also be similar ("least (common) post-dominator"?).
However, it also has the effect of invalidating borrows, so we would need to keep an `InvalidateBorrows(x)` statement around, and consider it one of the mentions of `x`.
Then "`Storage{Live,Dead}` range shrinking" would simply boil down to hoisting `InvalidateBorrows(x)` up past statements which couldn't indirectly access `x`.
<hr/>
cc @nikomatsakis @ecstatic-morse @rust-lang/wg-mir-opt | C-cleanup,T-compiler,A-MIR | medium | Critical |
556,536,719 | terminal | Alternate console buffers of different sizes confuse ConPTY (and therefore Windows Terminal) | <!--
This bug tracker is monitored by Windows Terminal development team and other technical folks.
**Important: When reporting BSODs or security issues, DO NOT attach memory dumps, logs, or traces to Github issues**.
Instead, send dumps/traces to [email protected], referencing this GitHub issue.
If this is an application crash, please also provide a Feedback Hub submission link so we can find your diagnostic data on the backend. Use the category "Apps > Windows Terminal (Preview)" and choose "Share My Feedback" after submission to get the link.
Please use this form and describe your issue, concisely but precisely, with as much detail as possible.
-->
# Environment
```
Microsoft Windows [Version 10.0.18362.535]
Version: 0.8.10261.0
```
# Steps to reproduce
<!-- A description of how to trigger this bug. -->
1. Open a TUI application in Windows Terminal.
2. Resize the Terminal Window
3. Close the TUI application
The resulting terminal window acts like it is the original size before resizing, but the resized area remains visible.
For reference, the TUI application I was testing is a command line 2048 game: https://github.com/jwlodek/py_cui_2048
It can be installed with pip for python3:
```
pip install py_cui_2048
py2048
```
# Expected behavior
<!-- A description of what you're expecting, possibly containing screenshots or reference material. -->
When performing the same steps in `cmd` the terminal is resized back to its original size after resize and TUI application closed. This is also the case for a standalone `powershell` window.
# Actual behavior
<!-- What's actually happening? -->
The window does seem to internally resize, but the extra space is still drawn. Please see video below for visual example:

| Help Wanted,Area-Rendering,Product-Conpty,Area-Output,Issue-Bug,Priority-2 | low | Critical |
556,540,251 | material-ui | Slide recreates style object property | * [x] The issue is present in the latest release.
* [x] I have searched the [issues](https://github.com/mui-org/material-ui/issues) of this repository and believe that this is not a duplicate.
## Current Behavior
Slide component always recreates children `style` object and sets its `visibility` field to `'hidden'` / `undefined`.
In such situation we cannot use neither `React.memo` (without compare function) nor `React.PureComponent`, cause function `shallowEquals` always returns `false`.
https://github.com/mui-org/material-ui/blob/5c84559ce012e403d01773112774de56d7e74723/packages/material-ui/src/Slide/Slide.js#L212-L222
## Expected Behavior
Slide should try to use original style objects whenever it can.
Something like that, should be enough:
```js
// Trying to get children style without recreating style object
const getChildrenStyle = (state) => {
const forwardedStyle = (() => {
if (!style) {
return children.props.style;
}
if (!children.props.style) {
return style;
}
return {...style, ...children.props.style};
})();
if (state === 'exited' && !inProp) {
return {visibility: "hidden", ...forwardedStyle}
}
return forwardedStyle;
};
```
## Steps to Reproduce
https://codesandbox.io/s/serene-easley-ll4ut
## Your Environment
Tech | Version
--- | ---
Material-UI | v4.9.0
React | latest
Browser | chrome
| component: transitions | low | Minor |
556,549,622 | TypeScript | Private Fields: Code action for marking a field private | Testing https://github.com/microsoft/vscode/issues/89387.
It would make sense to provide a code action / refactoring for turning a public field into a private field that also generates a getter if needed. | Suggestion,Awaiting More Feedback | low | Minor |
556,550,117 | nvm | nvm changes user profile - must be "opt-in" rather than "opt-out" | <!-- Thank you for being interested in nvm! Please help us by filling out the following form if youβre having trouble. If you have a feature request, or some other question, please feel free to clear out the form. Thanks! -->
#### Operating system and version:
Linux.
#### `nvm debug` output:
Not relevant to this request.
#### `nvm ls` output:
<details>
$ nvm ls
-> v12.14.1
v13.6.0
v13.7.0
system
default -> lts/* (-> v12.14.1)
node -> stable (-> v13.7.0) (default)
stable -> 13.7 (-> v13.7.0) (default)
iojs -> N/A (default)
unstable -> N/A (default)
lts/* -> lts/erbium (-> v12.14.1)
lts/argon -> v4.9.1 (-> N/A)
lts/boron -> v6.17.1 (-> N/A)
lts/carbon -> v8.17.0 (-> N/A)
lts/dubnium -> v10.18.1 (-> N/A)
lts/erbium -> v12.14.1
```sh
```
</details>
#### How did you install `nvm`?
Install script.
#### What steps did you perform?
```
wget -qO- https://raw.githubusercontent.com/nvm-sh/nvm/v0.35.2/install.sh | XDG_CONFIG_HOME=`pwd` bash
:
=> Appending nvm source string to /home/bryce/muggeridge/.bashrc
=> Appending bash_completion source string to /home/bryce/muggeridge/.bashrc
:
```
Modifying the user's profile is altogether unfriendly (and un-Linux). It has resulted in some of our users' login session breaking, since it cannot find the NVM_DIR.
This *has* to become an opt-in feature.
I know you can opt-out, by setting the environment variable PROFILE=/dev/null. However, most users are unaware of that, and nor should they have to become aware.
#### What happened?
The user's profile is updated by default, without them requesting that to be updated. Some users' login sessions were subsequently broken. Dare I say, the current behaviour is virus-like, albeit well-intentioned as I understand what you were trying to achieve.
#### What did you expect to happen?
A user ought to opt-in to have their profile updated. Currently, the default behaviour requires the user to opt-out.
#### Is there anything in any of your profile files that modifies the `PATH`?
<!-- (e.g. `.bashrc`, `.bash_profile`, `.zshrc`, etc) -->
<!-- Please remove the following section if it does not apply to you -->
#### If you are having installation issues, or getting "N/A", what does `curl -I --compressed -v https://nodejs.org/dist/` print out?
<details>
<!-- do not delete the following blank line -->
```sh
```
</details>
| installing nvm: profile detection | low | Critical |
556,580,596 | pytorch | PyTorch freezes on second call to scripted densenet model from torchvision | ## π Bug
PyTorch freezes on second call to scripted densenet model from torchvision.
## To Reproduce
Steps to reproduce the behavior:
The script freezes after printing the first result.
```python
import torch
import torchvision
"""
Freezes on the second call to run the script model.
"""
model = torchvision.models.densenet121(pretrained=True)
x = torch.randn(2, 3, 224, 224, requires_grad=True)
script_model = torch.jit.script(model)
print('resuit:', script_model(x))
print('second result:', script_model(x))
```
## Environment
Both PyTorch and torchvision are nightly, installed by `conda install pytorch torchvision cpuonly -c pytorch-nightly`
PyTorch version: 1.5.0.dev20200128
Is debug build: No
CUDA used to build PyTorch: Could not collect
OS: Ubuntu 16.04.5 LTS
GCC version: (Ubuntu 5.4.0-6ubuntu1~16.04.12) 5.4.0 20160609
CMake version: version 3.5.1
Python version: 3.7
Is CUDA available: No
CUDA runtime version: Could not collect
GPU models and configuration: GPU 0: GeForce GTX 1070
Nvidia driver version: 410.78
cuDNN version: Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.7.4.1
/usr/local/cuda-10.0/lib64/libcudnn.so.7
cc @suo @fmassa | oncall: jit,triaged,module: vision | low | Critical |
556,585,505 | pytorch | RPC mock mode for unit tests. | ## π Feature
I am building some distributed features that rely on RPC. For my unit test, once I need to test multiple workers, RPC will bring up a process group. As a developer that uses RPC api, launching communication layer feels too heavy for unit tests.
It would be great to have a mock version of RPC that bypasses the communication layer (process group / gloo / etc ) and enables lightweight RPC API testing.
cc @pietern @mrshenli @pritamdamania87 @zhaojuanmao @satgera @gqchen @aazzolini @rohan-varma @xush6528 @jjlilley @osalpekar | triaged,module: rpc | low | Minor |
556,591,155 | pytorch | JIT performance discrepancies | PyTorch Nightly (today)
Fedora 31
Anaconda Python 3.7
GPU: 970 (staying under 3.5 GB limit, last 0.5GB is lower quality memory)
Not completely confident it's a bug, if it isn't there should certainly be something mentioned in the documentation.
All result variance is ~0.01
```
import numpy as np
import torch
import time
data_type = torch.torch.cuda.DoubleTensor
cuda = torch.device('cuda')
points = torch.rand((2**22,3), dtype=torch.double, device=cuda, requires_grad=False).type(data_type)
points[:, 2] = 0.0
@torch.jit.script
def attractor_func(v):
x, y, z = torch.chunk(v, 3, dim=1)
#lorenz
a = 10.0
b = 28.0
c = 8.0 / 3.0
dx = a * (y - x)
dy = x * (b - z) - y
dz = x * y - c * z
return torch.cat([dx, dy, dz], 1)
@torch.jit.script
def rk4(v):
h = 0.001
k1 = attractor_func(v)
k2 = attractor_func(v + (h / 2.0) * k1)
k3 = attractor_func(v + (h / 2.0) * k2)
k4 = attractor_func(v + h * k3)
return h * (k1 / 6.0 + k2 / 3.0 + k3 / 3.0 + k4 / 6.0)
rk4 = torch.jit.trace(rk4, example_inputs=points)
start = time.time()
for _ in range(1024):
points += rk4(points)
end = time.time()
print(end - start)
```
output:
18.279905796051025
comment out the trace, output:
18.649290084838867
Why is the trace adding performance to JIT scripts?
Uncomment the trace and comment the script tags on both functions, output:
28.479955911636353
Why does the trace benefit greatly from JIT scripts?
Another example that may have more to do with wrong function choice or some minor inefficiency somewhere:
```
import numpy as np
import torch
import time
data_type = torch.torch.cuda.DoubleTensor
cuda = torch.device('cuda')
points = torch.rand((2**22,3), dtype=torch.double, device=cuda, requires_grad=False).type(data_type)
points[:, 2] = 0.0
def attractor_func(v):
x, y, z = torch.chunk(v, 3, dim=1)
#duffing
a = 0.25
b = 0.3
c = 1.0
dx = y
dy = x - x**3 - a * y + b * torch.cos(c * z)
dz = x - x + 1.0
return torch.cat([dx, dy, dz], 1)
def rk4(v):
h = 0.001
k1 = attractor_func(v)
k2 = attractor_func(v + (h / 2.0) * k1)
k3 = attractor_func(v + (h / 2.0) * k2)
k4 = attractor_func(v + h * k3)
return h * (k1 / 6.0 + k2 / 3.0 + k3 / 3.0 + k4 / 6.0)
rk4 = torch.jit.trace(rk4, example_inputs=points)
start = time.time()
for _ in range(1024):
points += rk4(points)
end = time.time()
print(end - start)
```
output:
46.12030529975891
change to `dz = x.clone().fill_(1.0)`
output:
65.20094799995422
change to `dz = torch.ones((2**22,1), dtype=torch.double, device=cuda, requires_grad=False).type(data_type)`
output:
46.92164850234985
change to `dz = x.new_ones((2**22,1))`
output:
46.9119393825531
Only the last two make sense. Though it seems like itβs precisely the last two that make the use of a script tag for the function impossible. In this case, unlike the previous one the script tags don't appear to do much.
cc @suo @VitalyFedyunin @ngimel | module: performance,oncall: jit,triaged | low | Critical |
556,599,458 | flutter | Add test to verify that loading code using dart:mirrors via IsolateMirror.loadURI does not regress. | This is not supported in Flutter but an internal user depends on this functionality. Test must be added to ensure that this does not regress till they migrate away to an alternative. The entirety of the filesystem and package map handling code in Tonic is written to support this use case. Since this use case does not at first glance seem supported, attempts may be made by engineers to get rid of the same.
xref https://github.com/flutter/engine/pull/16157 | a: tests,engine,e: embedder,P2,team-engine,triaged-engine | low | Minor |
556,639,251 | scrcpy | Lagging on my Mobile(Poco f1) | hi first of all thanks for making this software.
now lets come to the problem I am using Xiaomi Poco f1 device and I am casting my screen to my pc(windows 10) for recording I am recording with OBS till now no problem everything is working fine but when I start to record with OBS somehow my device starts to lag the game especially PUBG Mobile I don't know why I think Poco f1 has good hardware(SD 845) and should not lag in this case.
if you have any setting which I can play around then please tell me. it would be appreciated.
Thanks. | performance,obs | low | Minor |
556,654,116 | angular | A unit test for a component with an animation does not behave as expected, failing the test | # π bug report
Apologises if I have not filled this out correctly, or if I missed a similar issue.
### Affected Package
Unsure, the issue is present when using either the BrowserAnimationsModule or the NoopAnimationsModule, in unit tests.
### Description
A component I created, is a slide in panel from the right. The slide in is animated using the Angular animations module. When using the application as normal, the slider appears and disappears as expected, animating as expected.
When testing the component using karma/jasmine the component does not behave as expected. One of two things will happen, either the panel is hidden and will not show, or the panel is visible initially and will not hide. When inspecting the styles of the slide in panel, there are some mangled styles that are not present when using the app as normal (style="width:50vw;opacity:1;0:width;1:opacity;opacity:1;webkit-opacity:1;width:50vw;").
The Component is supposed to be hidden initially, and when triggered, an animation with change from 0 width to 100% of its contents, and 0 opacity to 1 opacity. The test is checking the components opacity style to confirm that the component is visible. It is checked before and after the components activation. When there is a mouse click, that is not on the component, the component is supposed to hide itself again.
## π¬ Minimal Reproduction
https://stackblitz.com/edit/angular-xblhso
As part of the reproduction, in the application window, there is a normal app version at the top of the window, which you can use to see how it should work. Then a jasmine section below that, with a unit test that is failing, as per description above.
Also this is my first time using stackblitz, hope it works the way it seems like it is supposed to...
<!--
If StackBlitz is not suitable for reproduction of your issue, please create a minimal GitHub repository with the reproduction of the issue.
A good way to make a minimal reproduction is to create a new app via `ng new repro-app` and add the minimum possible code to show the problem.
Share the link to the repo below along with step-by-step instructions to reproduce the problem, as well as expected and actual behavior.
Issues that don't have enough info and can't be reproduced will be closed.
You can read more about issue submission guidelines here: https://github.com/angular/angular/blob/master/CONTRIBUTING.md#-submitting-an-issue
-->
## π Your Environment
**Angular Version:**
<pre><code>
$ ng version
_ _ ____ _ ___
/ \ _ __ __ _ _ _| | __ _ _ __ / ___| | |_ _|
/ β³ \ | '_ \ / _` | | | | |/ _` | '__| | | | | | |
/ ___ \| | | | (_| | |_| | | (_| | | | |___| |___ | |
/_/ \_\_| |_|\__, |\__,_|_|\__,_|_| \____|_____|___|
|___/
Angular CLI: 8.0.4
Node: 10.15.3
OS: win32 x64
Angular: 8.0.2
... animations, common, compiler, compiler-cli, core, forms
... platform-browser, platform-browser-dynamic, platform-server
... router, upgrade
Package Version
-----------------------------------------------------------
@angular-devkit/architect 0.800.4
@angular-devkit/build-angular 0.800.4
@angular-devkit/build-optimizer 0.800.4
@angular-devkit/build-webpack 0.800.4
@angular-devkit/core 0.8.6
@angular-devkit/schematics 8.0.4
@angular/cli 8.0.4
@ngtools/webpack 8.0.4
@schematics/angular 8.0.4
@schematics/update 0.800.4
rxjs 6.5.2
typescript 3.4.5
webpack 4.30.0
</code></pre>
| type: bug/fix,area: testing,area: animations,freq2: medium,P3 | low | Critical |
556,696,642 | flutter | Make the Cupertino Date Picker easier to driver with Flutter driver | Similar to #49100 if would be great to make this widget easier to driver using Flutter driver by adding well known keys to the days, months, years, hours, minutes, seconds CupertinoPicker widgets within this control. These picker could then be scrolled and tapped and thus any date and/ or time could be picked.
At the moment this widget is impossible to driver using the driver.
I'm happy to do a PR for this so let me know any thoughts or if I have missed something simple with driving this element with the driver. | framework,f: date/time picker,f: cupertino,c: proposal,P3,team-design,triaged-design | low | Major |
556,714,626 | godot | Bug that caused project to lose all files except for the default files | **Godot version:**
v3.1.stable.official
**OS/device including version:**
Windows 10
**Issue description:**
The project that lost the files will be in like when it was first created with the default files
**Steps to reproduce:**
I encountered this bug when in a project then quit to project list then enter another project and quit again, then the project list will be empty, import your projects back and a project would be affected with the bug
| bug,topic:editor | low | Critical |
556,716,938 | flutter | DropDownButton looses selection when disabled | A DropDownButtonhas no `enable` property. Instead, the `onChanged` property must be set to null. In this case, any selection will be replaced by `disabledHint` or `hint` which is inconsistent with other input widgets.
It would be much better to have
1. a property `enable` to enable / disable the widget (more clear to handle)
2. display the selected value (its widget) independent of enabled state
3. move hints to `InputDecoration` which should become a property `decoration` for the dropdown button.
| framework,f: material design,c: proposal,P3,team-design,triaged-design | low | Minor |
556,722,162 | TypeScript | prettify complex type display (for generic-heavy types) | <!-- π¨ STOP π¨ π¦π§π’π£ π¨ πΊπ»πΆπ· π¨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker.
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ, especially the "Common Feature Requests" section: https://github.com/Microsoft/TypeScript/wiki/FAQ
-->
## Search Terms
https://github.com/microsoft/TypeScript/labels/Domain%3A%20Type%20Display
<!-- List of keywords you searched for before creating this issue. Write them down here so that others can find this suggestion more easily -->
## Suggestion
give indentations + line-changes on type representation. User config for this might be an option.
## Examples
suppose I have some generic-heavy function `someFunc` that returns
```
const a = someFunc();
// typeof a = RootGeneric<GenericA, GenericB<number, SubGenericB_A<number> ...
// full type not shown π
```
A better output can be:
```
// typeof a =
RootGeneric<
GenericA,
GenericB<
number,
SubGenericB_A<number> ...
```
more things:
+ editors can implement "expand(+)" buttons ? \
(collapsed-tree-view)
<!-- Show how this would be used and what the behavior would be -->
## Checklist
My suggestion meets these guidelines:
* [x] This wouldn't be a breaking change in existing TypeScript/JavaScript code
* [x] This wouldn't change the runtime behavior of existing JavaScript code
* [x] This could be implemented without emitting different JS based on the types of the expressions
* [x] This isn't a runtime feature (e.g. library functionality, non-ECMAScript syntax with JavaScript output, etc.)
* [x] This feature would agree with the rest of [TypeScript's Design Goals](https://github.com/Microsoft/TypeScript/wiki/TypeScript-Design-Goals).
| Suggestion,Awaiting More Feedback | low | Critical |
556,770,453 | material-ui | [material-ui][docs] Document the correct way to extend a component | We are using Material-UI as the base of a project, but are running in to issues extending components
```
import { Link, LinkProps } from "@material-ui/core";
export const HnLink = Link;
```
Works fine
```
import { Link, LinkProps } from "@material-ui/core";
export const HnLink: React.FC<LinkProps> = props => {
// Some extra functionality
return <Link {...props} />
}
```
Does not accept `component` and `to` props, resulting in TypeScript errors.
```
Type '{ to: string; component: typeof Link; }' is not assignable to type 'IntrinsicAttributes & AnchorHTMLAttributes<HTMLAnchorElement> & Pick<OverrideProps<TypographyTypeMap<{}, "span">, "span">, "ref" | ... 262 more ... | "variantMapping"> & { ...; } & CommonProps<...> & Pick<...> & { ...; }'.
Property 'to' does not exist on type 'IntrinsicAttributes & AnchorHTMLAttributes<HTMLAnchorElement> & Pick<OverrideProps<TypographyTypeMap<{}, "span">, "span">, "ref" | ... 262 more ... | "variantMapping"> & { ...; } & CommonProps<...> & Pick<...> & { ...; }'.ts(2322)
```
The same is true of Tabs, wherein wrapping Tabs in our own component causes onChange to not be accepted
```
Type '(event: ChangeEvent<{}>, value: number) => void' is not assignable to type '((event: ChangeEvent<{}>, value: any) => void) & ((event: FormEvent<HTMLButtonElement>) => void)'.
Type '(event: ChangeEvent<{}>, value: number) => void' is not assignable to type '(event: FormEvent<HTMLButtonElement>) => void'.ts(2322)
```
As also seen here: https://github.com/mui-org/material-ui/issues/17454
Is there a 'correct' way to wrap a Material-UI component that can alleviate typing issues? | docs,typescript | low | Critical |
556,777,028 | rust | false positive "trait already implemented" (E0371) | While investigating #68564, I found a real bug.
https://play.rust-lang.org/?version=nightly&mode=debug&edition=2018&gist=c959b9e4ee176c4430586425307787ec
```rust
pub trait Fold<T: ?Sized> {
fn fold(&mut self, node: T) -> T;
}
struct A;
struct B;
impl Fold<A> for dyn Fold<B> {}
```
It results in an error even on stable.
`Fold<A>` and `Fold<B>` is treated as a separate trait (#68564) while resolving, but treated as a same trait while verifying impl blocks.
```
Compiling playground v0.0.1 (/playground)
error[E0371]: the object type `(dyn Fold<B> + 'static)` automatically implements the trait `Fold`
--> src/lib.rs:11:1
|
11 | impl Fold<A> for dyn Fold<B> {}
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ `(dyn Fold<B> + 'static)` automatically implements trait `Fold`
error: aborting due to previous error
For more information about this error, try `rustc --explain E0371`.
error: could not compile `playground`.
To learn more, run the command again with --verbose.
``` | A-trait-system,T-compiler,C-bug | low | Critical |
556,807,713 | pytorch | Documentation is not loaded by IDEs | ## π Bug
The documentation of torch functions (e.g. `torch.sort`, `torch.max`) and the class `torch.Tensor` is not loaded by common IDEs as PyCharm and Visual Studio Code.
The documentation of other modules (e.g. `torch.nn`, `torch.optim`) is loaded correctly.
The problem seems to be related to dynamic documentation which is not supported by IDEs (static docs are not inserted in `torch.__init__.pyi`). The (dynamic) documentation is loaded correctly in python consoles only.
Please see the related PyCharm issue.
## To Reproduce
Steps to reproduce the behavior:
1. Create a new Python project with PyCharm or Visual Studio Code
1. Import torch
1. Start typing a torch function or a tensor method (e.g. `torch.sort`, `torch.ones', 'torch.ones().max`). An empty documentation popup will be opened.
The same behavior occurs with keyboard shortcuts (PyCharm: `Ctrl+Q`, Visual Studio Code: `Ctrl+K, Ctrl+I`)
## Expected behavior
Static documentation should be available for every torch module to support documentation loading by IDEs.
## Environment
- PyTorch Version (e.g., 1.0): 1.2.0, 1.3.1, 1.4.0
- OS (e.g., Linux): Windows and Linux
- How you installed PyTorch (`conda`, `pip`, source): pip
- Build command you used (if compiling from source):
- Python version: 3.6
- CUDA/cuDNN version: 10
- GPU models and configuration: 1080Ti
- Any other relevant information: PyCharm 2019.3.2, Visual Studio Code 1.41.1
## Additional context
Related PyCharm issue: [https://youtrack.jetbrains.com/issue/PY-40262](https://youtrack.jetbrains.com/issue/PY-40262) | module: docs,triaged | low | Critical |
556,811,476 | angular | DefaultValueAccessor doesn't update the model correctly when entering CJK characters with an IME that has autocorrect | # π bug report
### Affected Package
The issue is caused by package @angular/forms
### Is this a regression?
No, seems to have always been there since it was implemented.
### Description
The model is not updated correctly by `DefaultValueAccessor` when using an IME to enter CJK characters that use composition. `COMPOSITION_BUFFER_MODE` waits until `compositionend` event to update the model, but that event is skipped when the keyboard has autocorrect on and the user focuses another field without confirming the last entered character.
Related information:
* [Similar bug in AngularJS](https://stackoverflow.com/a/49505054). They suggest to use `autocorrect="off"` at the input, but obviously you could want to keep that functionality and still have the model updated.
* [Explanation from Chrome team](https://bugs.chromium.org/p/chromium/issues/detail?id=818881#c9) about the way the keyboard works is by design:
> The composition text in the case in the video recording is "a" with the underline effect. "compositionend" event will appear after the underlined text is committed by, e.g. SPACE key.
## π¬ Minimal Reproduction
https://stackblitz.com/edit/angular-composition-buffer-mode?file=src%2Fapp%2Fcomp-true.component.ts
You can try using a CJK keyboard with autocorrect and then:
* Start a composition
* Focus another field with the mouse without pressing space or enter
* For the fields that use `COMPOSITION_BUFFER_MODE = true` the model is not updated.
## π Your Environment
**Angular Version:**
<pre><code>Angular CLI: 8.3.23
Node: 10.18.0
OS: win32 x64
Angular: 8.2.14
... animations, common, compiler, compiler-cli, core, forms
... language-service, platform-browser, platform-browser-dynamic
... router, service-worker
Package Version
-----------------------------------------------------------
@angular-devkit/architect 0.803.23
@angular-devkit/build-angular 0.803.23
@angular-devkit/build-optimizer 0.803.23
@angular-devkit/build-webpack 0.803.23
@angular-devkit/core 8.3.23
@angular-devkit/schematics 8.3.23
@angular/cdk 8.2.3
@angular/cli 8.3.23
@ngtools/webpack 8.3.23
@schematics/angular 8.3.23
@schematics/update 0.803.23
rxjs 6.5.4
typescript 3.5.3
webpack 4.39.2
</code></pre>
**Anything else relevant?**
This happens with Chrome, Safari, IE... it has to do with the IME used.
The easy workaround for apps is `COMPOSITION_BUFFER_MODE = false`, that solves the issue.
However, this could be easily solved in the framework by updating the model on blur, understanding that it does the same effect as `compositionend` in [`DefaultValueAccessor`](https://github.com/angular/angular/blob/1537791f06df5d85dd27e763836fd99bb1530991/packages/forms/src/directives/default_value_accessor.ts#L145) with a code like:
```ts
/** @internal */
_blur(value: any): void {
this._composing = false;
this._compositionMode && this._composing && this.onChange(value);
this.onTouched();
}
``` | area: forms,forms: ControlValueAccessor,P4 | low | Critical |
556,820,231 | rust | 'Referencing function in another module` error when compiling for target wasm32-unknown-emscripten (debug mode) | I'm trying to compile a Rust project that depends on the [`byte-slice-cast`](https://crates.io/crates/byte-slice-cast) crate, but I'm encountering a strange error. The same error pops up if having [`block-padding`](https://crates.io/crates/block-padding) as a dependency. You can find all the details in this issue: https://github.com/sdroege/byte-slice-cast/issues/13. As mentioned there, this only happens when compiling in debug mode, while release builds succeed.
Some info about the environment:
`rustc 1.40.0 (73528e339 2019-12-16)`
`emcc (Emscripten gcc/clang-like replacement) 1.39.1 ((unknown revision))` | A-LLVM,T-compiler,O-wasm,C-bug,O-emscripten | low | Critical |
556,843,382 | excalidraw | Figure out a way to merge scene serializers | As discussed in https://github.com/excalidraw/excalidraw/pull/583#issuecomment-579734116, currently we use different scene serializers when storing to `localStorage`, and when saving to backend or JSON (file).
This means there is potential for unintended divergence, such as when we decide to persist additional state keys and forget to modify both serializers. Note that there's already a divergence: when storing to `localStorage`, we don't store `version` which we do in the other cases.
At the same time, we cannot simply use the same serializer, because we don't want to store the exact the same state (such as scroll position, etc.) in every situation.
Thus, we should figure out a way to mitigate (or minimize) the potential for unintended divergence ()
The helpers in question are:
- [`saveToLocalStorage()`](https://github.com/excalidraw/excalidraw/blob/82717bcedad3f9aa5860513dcf3994372a9073fc/src/scene/data.ts#L301) - when storing to `localStorage`.
- [`serializeAsJSON()`](https://github.com/excalidraw/excalidraw/blob/82717bcedad3f9aa5860513dcf3994372a9073fc/src/scene/data.ts#L29) - for everything else.
One way is to use a single API, but supply a list of keys to *omit* (i.e. using a blacklist, which will minimize the likelihood of omitting keys added in the future). | refactor | low | Major |
556,857,098 | TypeScript | SkipLibCheck fails to work in multi-project set-up | **TypeScript Version:** 3.8. Beta
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
skiplibcheck
**Steps to reproduce**
1. Extract the attached zip file to a folder and open `repro.sln` in Visual Studio (I'm using 2019)
2. Expand the `npm` folder on project B and install the missing `@azure/storage-blob` package.
3. Rebuild the solution.
**Expected behavior:**
No build errors.
**Actual behavior:**
2 errors:
> error TS2740: Build:Type 'Buffer' is missing the following properties...
> error TS2345: Build:Argument of type 'ErrnoException' is not assignable...
**Explanation**
We want have a custom definition for NodeJS. The npm package imported in Project B also has its own (incomplete) definition. So, we've added `skiplibcheck` to all the projects.
Project B builds fine by itself.
Project C references some NodeJS types. Project C also builds fine by itself.
Project A references types from both B and C.
When attempting to build A, we get the error above. **But** the error is for project C.
Basically, when building Project A, the `skiplibcheck` option is being ignored and the incomplete NodeJS definition from the imported package is being used.
**Further Observations**
Curiously, if file `A/foo/bar.ts` is moved to `A/bar.ts` (and the imports amended) the build compiles fine.
**Attachment**
[Repro.zip](https://github.com/microsoft/TypeScript/files/4128129/Repro.zip)
| Needs Investigation | low | Critical |
556,871,668 | material-ui | [system] Support new props with variants | Hi,
is this possible extend `PaperClassKey` and `PaperProps` with custom values?
I would like to get something like this:
```jsx
export const theme = createMuiTheme({
overrides: {
MuiPaper: {
dark: {
backgroundColor: COLORS.BLACK,
},
light: {
backgroundColor: COLORS.WHITE,
},
},
},
});
```
And using the component would look like this:
```
<Paper filled="dark" />
``` | new feature,waiting for π,package: system | low | Major |
556,903,054 | pytorch | Error while trying to build pytorch from source in conda environment | Hi,
I am trying to build pytorch from source in a conda environment, following the instructions from https://github.com/pytorch/pytorch#from-source. The CPU installation works fine (until this step: `[2521/3544] Building CXX object caffe2/CMakeFiles/torch_cpu.dir/__/torch/csrc/distributed/autograd/engine/dist_engine.cpp.o
`) but as soon as the installation process moves to cuda, it fails (output below)
Any help would be much appreciated. Here is some information about my environment:
<details>
```
$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Wed_Oct_23_19:24:38_PDT_2019
Cuda compilation tools, release 10.2, V10.2.89
$ conda env export
conda env export -n fairseq
name: fairseq
channels:
- pytorch
- conda-forge
- defaults
dependencies:
- _libgcc_mutex=0.1=conda_forge
- _openmp_mutex=4.5=0_gnu
- bzip2=1.0.8=h516909a_2
- ca-certificates=2019.11.28=hecc5488_0
- certifi=2019.11.28=py38_0
- cffi=1.13.2=py38h8022711_0
- cmake=3.16.3=h28c56e5_0
- expat=2.2.9=he1b5a44_1
- intel-openmp=2019.4=243
- krb5=1.16.4=h2fd8d38_0
- ld_impl_linux-64=2.33.1=h53a641e_8
- libblas=3.8.0=14_openblas
- libcblas=3.8.0=14_openblas
- libcurl=7.65.3=hda55be3_0
- libedit=3.1.20170329=hf8c457e_1001
- libffi=3.2.1=he1b5a44_1006
- libgcc-ng=9.2.0=h24d8f2e_2
- libgfortran-ng=7.3.0=hdf63c60_4
- libgomp=9.2.0=h24d8f2e_2
- liblapack=3.8.0=14_openblas
- libopenblas=0.3.7=h5ec1e0e_6
- libssh2=1.8.2=h22169c7_2
- libstdcxx-ng=9.2.0=hdf63c60_2
- libuv=1.34.0=h516909a_0
- magma-cuda101=2.5.1=1
- mkl=2019.4=243
- mkl-include=2019.5=281
- ncurses=6.1=hf484d3e_1002
- ninja=1.10.0=hc9558a2_0
- numpy=1.17.5=py38h95a1406_0
- openssl=1.1.1d=h516909a_0
- pip=20.0.2=py38_0
- pycparser=2.19=py38_1
- python=3.8.1=h357f687_1
- pyyaml=5.3=py38h516909a_0
- readline=8.0=hf8c457e_0
- rhash=1.3.6=h14c3975_1001
- setuptools=45.1.0=py38_0
- sqlite=3.30.1=hcee41ef_0
- tk=8.6.10=hed695b0_0
- wheel=0.34.1=py38_0
- xz=5.2.4=h14c3975_1001
- yaml=0.2.2=h516909a_1
- zlib=1.2.11=h516909a_1006
prefix: /home/quent/.local/bin/anaconda3/envs/fairseq
```
```
# Steps that I did
$ conda install -n fairseq conda install numpy ninja pyyaml mkl mkl-include setuptools cmake cffi
$ conda install -c pytorch magma-cuda101 -n fairseq
$ git clone --recursive https://github.com/pytorch/pytorch && cd pytorch
$ export CMAKE_PREFIX_PATH=$HOME/.local/bin/anaconda3/envs/fairseq
$ optirun python setup.py install
Building wheel torch-1.5.0a0+5e23110
-- Building version 1.5.0a0+5e23110
cmake -GNinja -DBUILD_PYTHON=True -DBUILD_TEST=True -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/home/quent/.local/bin/pytorch/torch -DCMAKE_PREFIX_PATH=/home/quent/.local/bin/anaconda3/envs/fairseq -DNUMPY_INCLUDE_DIR=/home/quent/.local/bin/anaconda3/envs/fairseq/lib/python3.8/site-packages/numpy/core/include -DPYTHON_EXECUTABLE=/home/quent/.local/bin/anaconda3/envs/fairseq/bin/python -DPYTHON_INCLUDE_DIR=/home/quent/.local/bin/anaconda3/envs/fairseq/include/python3.8 -DPYTHON_LIBRARY=/home/quent/.local/bin/anaconda3/envs/fairseq/lib/libpython3.8.so.1.0 -DTORCH_BUILD_VERSION=1.5.0a0+5e23110 -DUSE_NUMPY=True /home/quent/.local/bin/pytorch
-- The CXX compiler identification is GNU 9.2.0
-- The C compiler identification is GNU 9.2.0
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Not forcing any particular BLAS to be found
-- Performing Test COMPILER_WORKS
-- Performing Test COMPILER_WORKS - Success
-- Performing Test SUPPORT_GLIBCXX_USE_C99
-- Performing Test SUPPORT_GLIBCXX_USE_C99 - Success
-- Performing Test CAFFE2_EXCEPTION_PTR_SUPPORTED
-- Performing Test CAFFE2_EXCEPTION_PTR_SUPPORTED - Success
-- std::exception_ptr is supported.
-- Performing Test CAFFE2_NEED_TO_TURN_OFF_DEPRECATION_WARNING
-- Performing Test CAFFE2_NEED_TO_TURN_OFF_DEPRECATION_WARNING - Failed
-- Turning off deprecation warning due to glog.
-- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX2_EXTENSIONS
-- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX2_EXTENSIONS - Success
-- Current compiler supports avx2 extension. Will build perfkernels.
-- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX512_EXTENSIONS
-- Performing Test CAFFE2_COMPILER_SUPPORTS_AVX512_EXTENSIONS - Success
-- Current compiler supports avx512f extension. Will build fbgemm.
-- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY
-- Performing Test COMPILER_SUPPORTS_HIDDEN_VISIBILITY - Success
-- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY
-- Performing Test COMPILER_SUPPORTS_HIDDEN_INLINE_VISIBILITY - Success
-- Performing Test COMPILER_SUPPORTS_RDYNAMIC
-- Performing Test COMPILER_SUPPORTS_RDYNAMIC - Success
-- Building using own protobuf under third_party per request.
-- Use custom protobuf build.
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Caffe2 protobuf include directory: $<BUILD_INTERFACE:/home/quent/.local/bin/pytorch/third_party/protobuf/src>$<INSTALL_INTERFACE:include>
-- Trying to find preferred BLAS backend of choice: MKL
-- MKL_THREADING = OMP
-- Looking for sys/types.h
-- Looking for sys/types.h - found
-- Looking for stdint.h
-- Looking for stdint.h - found
-- Looking for stddef.h
-- Looking for stddef.h - found
-- Check size of void*
-- Check size of void* - done
-- Looking for cblas_sgemm
-- Looking for cblas_sgemm - found
-- MKL libraries: /home/quent/.local/bin/anaconda3/envs/fairseq/lib/libmkl_intel_lp64.so;/home/quent/.local/bin/anaconda3/envs/fairseq/lib/libmkl_gnu_thread.so;/home/quent/.local/bin/anaconda3/envs/fairseq/lib/libmkl_core.so;-fopenmp;/usr/lib/libpthread.so;/usr/lib/libm.so;/usr/lib/libdl.so
-- MKL include directory: /home/quent/.local/bin/anaconda3/envs/fairseq/include
-- MKL OpenMP type: GNU
-- MKL OpenMP library: -fopenmp
-- The ASM compiler identification is GNU
-- Found assembler: /usr/bin/cc
-- Check if compiler accepts -pthread
-- Check if compiler accepts -pthread - yes
-- Brace yourself, we are building NNPACK
-- Performing Test NNPACK_ARCH_IS_X86_32
-- Performing Test NNPACK_ARCH_IS_X86_32 - Failed
-- Found PythonInterp: /home/quent/.local/bin/anaconda3/envs/fairseq/bin/python (found version "3.8.1")
-- NNPACK backend is x86-64
-- Failed to find LLVM FileCheck
-- Found Git: /usr/bin/git (found version "2.25.0")
-- git Version: v1.4.0-505be96a
-- Version: 1.4.0
-- Performing Test HAVE_CXX_FLAG_STD_CXX11
-- Performing Test HAVE_CXX_FLAG_STD_CXX11 - Success
-- Performing Test HAVE_CXX_FLAG_WALL
-- Performing Test HAVE_CXX_FLAG_WALL - Success
-- Performing Test HAVE_CXX_FLAG_WEXTRA
-- Performing Test HAVE_CXX_FLAG_WEXTRA - Success
-- Performing Test HAVE_CXX_FLAG_WSHADOW
-- Performing Test HAVE_CXX_FLAG_WSHADOW - Success
-- Performing Test HAVE_CXX_FLAG_WERROR
-- Performing Test HAVE_CXX_FLAG_WERROR - Success
-- Performing Test HAVE_CXX_FLAG_PEDANTIC
-- Performing Test HAVE_CXX_FLAG_PEDANTIC - Success
-- Performing Test HAVE_CXX_FLAG_PEDANTIC_ERRORS
-- Performing Test HAVE_CXX_FLAG_PEDANTIC_ERRORS - Success
-- Performing Test HAVE_CXX_FLAG_WSHORTEN_64_TO_32
-- Performing Test HAVE_CXX_FLAG_WSHORTEN_64_TO_32 - Failed
-- Performing Test HAVE_CXX_FLAG_WFLOAT_EQUAL
-- Performing Test HAVE_CXX_FLAG_WFLOAT_EQUAL - Success
-- Performing Test HAVE_CXX_FLAG_FSTRICT_ALIASING
-- Performing Test HAVE_CXX_FLAG_FSTRICT_ALIASING - Success
-- Performing Test HAVE_CXX_FLAG_WNO_DEPRECATED_DECLARATIONS
-- Performing Test HAVE_CXX_FLAG_WNO_DEPRECATED_DECLARATIONS - Success
-- Performing Test HAVE_CXX_FLAG_WSTRICT_ALIASING
-- Performing Test HAVE_CXX_FLAG_WSTRICT_ALIASING - Success
-- Performing Test HAVE_CXX_FLAG_WD654
-- Performing Test HAVE_CXX_FLAG_WD654 - Failed
-- Performing Test HAVE_CXX_FLAG_WTHREAD_SAFETY
-- Performing Test HAVE_CXX_FLAG_WTHREAD_SAFETY - Failed
-- Performing Test HAVE_CXX_FLAG_COVERAGE
-- Performing Test HAVE_CXX_FLAG_COVERAGE - Success
-- Performing Test HAVE_STD_REGEX
-- Performing Test HAVE_STD_REGEX
-- Performing Test HAVE_STD_REGEX -- success
-- Performing Test HAVE_GNU_POSIX_REGEX
-- Performing Test HAVE_GNU_POSIX_REGEX
-- Performing Test HAVE_GNU_POSIX_REGEX -- failed to compile
-- Performing Test HAVE_POSIX_REGEX
-- Performing Test HAVE_POSIX_REGEX
-- Performing Test HAVE_POSIX_REGEX -- success
-- Performing Test HAVE_STEADY_CLOCK
-- Performing Test HAVE_STEADY_CLOCK
-- Performing Test HAVE_STEADY_CLOCK -- success
-- Performing Test COMPILER_SUPPORTS_AVX512
-- Performing Test COMPILER_SUPPORTS_AVX512 - Success
-- Found OpenMP_C: -fopenmp (found version "4.5")
-- Found OpenMP_CXX: -fopenmp (found version "4.5")
-- Found OpenMP: TRUE (found version "4.5")
CMake Warning at third_party/fbgemm/CMakeLists.txt:80 (message):
OpenMP found! OpenMP_C_INCLUDE_DIRS =
-- Performing Test __CxxFlag__fmerge_all_constants
-- Performing Test __CxxFlag__fmerge_all_constants - Success
** AsmJit Summary **
ASMJIT_DIR=/home/quent/.local/bin/pytorch/third_party/fbgemm/third_party/asmjit
ASMJIT_TEST=FALSE
ASMJIT_TARGET_TYPE=STATIC
ASMJIT_DEPS=pthread;rt
ASMJIT_LIBS=asmjit;pthread;rt
ASMJIT_CFLAGS=-DASMJIT_STATIC
ASMJIT_PRIVATE_CFLAGS=-Wall;-Wextra;-fno-math-errno;-fno-threadsafe-statics;-DASMJIT_STATIC
ASMJIT_PRIVATE_CFLAGS_DBG=
ASMJIT_PRIVATE_CFLAGS_REL=-O2;-fmerge-all-constants
-- Could NOT find Numa (missing: Numa_INCLUDE_DIR Numa_LIBRARIES)
CMake Warning at cmake/Dependencies.cmake:602 (message):
Not compiling with NUMA. Suppress this warning with -DUSE_NUMA=OFF
Call Stack (most recent call first):
CMakeLists.txt:397 (include)
-- Using third party subdirectory Eigen.
Python 3.8.1
-- Found PythonInterp: /home/quent/.local/bin/anaconda3/envs/fairseq/bin/python (found suitable version "3.8.1", minimum required is "2.7")
-- Found PythonLibs: /home/quent/.local/bin/anaconda3/envs/fairseq/lib/libpython3.8.so.1.0 (found suitable version "3.8.1", minimum required is "2.7")
-- Could NOT find pybind11 (missing: pybind11_DIR)
-- Could NOT find pybind11 (missing: pybind11_INCLUDE_DIR)
-- Using third_party/pybind11.
-- pybind11 include dirs: /home/quent/.local/bin/pytorch/cmake/../third_party/pybind11/include
-- Found MPI_C: /usr/lib/openmpi/libmpi.so (found version "3.1")
-- Found MPI_CXX: /usr/lib/openmpi/libmpi_cxx.so (found version "3.1")
-- Found MPI: TRUE (found version "3.1")
-- MPI support found
-- MPI compile flags: -pthread
-- MPI include path: /usr/include
-- MPI LINK flags path: -Wl,-rpath -Wl,/usr/lib/openmpi -Wl,--enable-new-dtags -pthread
-- MPI libraries: /usr/lib/openmpi/libmpi_cxx.so/usr/lib/openmpi/libmpi.so
CMake Warning at cmake/Dependencies.cmake:833 (message):
OpenMPI found, but it is not built with CUDA support.
Call Stack (most recent call first):
CMakeLists.txt:397 (include)
-- Adding OpenMP CXX_FLAGS: -fopenmp
-- Will link against OpenMP libraries: /usr/lib/libgomp.so;/usr/lib/libpthread.so
-- Found CUDA: /opt/cuda (found version "10.2")
-- Caffe2: CUDA detected: 10.2
-- Caffe2: CUDA nvcc is: /opt/cuda/bin/nvcc
-- Caffe2: CUDA toolkit directory: /opt/cuda
-- Caffe2: Header version is: 10.2
-- Found CUDNN: /usr/lib/libcudnn.so
-- Found cuDNN: v7.6.5 (include: /usr/include, library: /usr/lib/libcudnn.so)
-- Automatic GPU detection failed. Building for common architectures.
-- Autodetected CUDA architecture(s): 3.5;5.0;5.2;6.0;6.1;7.0;7.0+PTX;7.5;7.5+PTX
-- Added CUDA NVCC flags for: -gencode;arch=compute_35,code=sm_35;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_52,code=sm_52;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_70,code=compute_70;-gencode;arch=compute_75,code=compute_75
-- Automatic GPU detection failed. Building for common architectures.
-- Autodetected CUDA architecture(s): 3.5;5.0;5.2;6.0;6.1;7.0;7.0+PTX;7.5;7.5+PTX
-- Could NOT find CUB (missing: CUB_INCLUDE_DIR)
CMake Warning (dev) at third_party/gloo/CMakeLists.txt:21 (option):
Policy CMP0077 is not set: option() honors normal variables. Run "cmake
--help-policy CMP0077" for policy details. Use the cmake_policy command to
set the policy and suppress this warning.
For compatibility with older versions of CMake, option is clearing the
normal variable 'BUILD_BENCHMARK'.
This warning is for project developers. Use -Wno-dev to suppress it.
-- MPI include path: /usr/include
-- MPI libraries: /usr/lib/openmpi/libmpi_cxx.so/usr/lib/openmpi/libmpi.so
-- Found CUDA: /opt/cuda (found suitable version "10.2", minimum required is "7.0")
-- CUDA detected: 10.2
-- Could NOT find NCCL (missing: NCCL_INCLUDE_DIR NCCL_LIBRARY)
CMake Warning at third_party/gloo/cmake/Dependencies.cmake:96 (message):
Not compiling with NCCL support. Suppress this warning with
-DUSE_NCCL=OFF.
Call Stack (most recent call first):
third_party/gloo/CMakeLists.txt:56 (include)
CMake Warning at cmake/Dependencies.cmake:1106 (message):
Metal is only used in ios builds.
Call Stack (most recent call first):
CMakeLists.txt:397 (include)
Generated: /home/quent/.local/bin/pytorch/build/third_party/onnx/onnx/onnx_onnx_torch-ml.proto
Generated: /home/quent/.local/bin/pytorch/build/third_party/onnx/onnx/onnx-operators_onnx_torch-ml.proto
--
-- ******** Summary ********
-- CMake version : 3.16.3
-- CMake command : /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake
-- System : Linux
-- C++ compiler : /usr/bin/c++
-- C++ compiler version : 9.2.0
-- CXX flags : -Wno-deprecated -fvisibility-inlines-hidden -fopenmp -Wnon-virtual-dtor
-- Build type : Release
-- Compile definitions : TH_BLAS_MKL;ONNX_ML=1
-- CMAKE_PREFIX_PATH : /home/quent/.local/bin/anaconda3/envs/fairseq;/opt/cuda
-- CMAKE_INSTALL_PREFIX : /home/quent/.local/bin/pytorch/torch
-- CMAKE_MODULE_PATH : /home/quent/.local/bin/pytorch/cmake/Modules;/home/quent/.local/bin/pytorch/cmake/public/../Modules_CUDA_fix
--
-- ONNX version : 1.6.0
-- ONNX NAMESPACE : onnx_torch
-- ONNX_BUILD_TESTS : OFF
-- ONNX_BUILD_BENCHMARKS : OFF
-- ONNX_USE_LITE_PROTO : OFF
-- ONNXIFI_DUMMY_BACKEND : OFF
-- ONNXIFI_ENABLE_EXT : OFF
--
-- Protobuf compiler :
-- Protobuf includes :
-- Protobuf libraries :
-- BUILD_ONNX_PYTHON : OFF
--
-- ******** Summary ********
-- CMake version : 3.16.3
-- CMake command : /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake
-- System : Linux
-- C++ compiler : /usr/bin/c++
-- C++ compiler version : 9.2.0
-- CXX flags : -Wno-deprecated -fvisibility-inlines-hidden -fopenmp -Wnon-virtual-dtor
-- Build type : Release
-- Compile definitions : TH_BLAS_MKL;ONNX_ML=1
-- CMAKE_PREFIX_PATH : /home/quent/.local/bin/anaconda3/envs/fairseq;/opt/cuda
-- CMAKE_INSTALL_PREFIX : /home/quent/.local/bin/pytorch/torch
-- CMAKE_MODULE_PATH : /home/quent/.local/bin/pytorch/cmake/Modules;/home/quent/.local/bin/pytorch/cmake/public/../Modules_CUDA_fix
--
-- ONNX version : 1.4.1
-- ONNX NAMESPACE : onnx_torch
-- ONNX_BUILD_TESTS : OFF
-- ONNX_BUILD_BENCHMARKS : OFF
-- ONNX_USE_LITE_PROTO : OFF
-- ONNXIFI_DUMMY_BACKEND : OFF
--
-- Protobuf compiler :
-- Protobuf includes :
-- Protobuf libraries :
-- BUILD_ONNX_PYTHON : OFF
-- Found CUDA with FP16 support, compiling with torch.cuda.HalfTensor
-- Removing -DNDEBUG from compile flags
-- Checking prototype magma_get_sgeqrf_nb for MAGMA_V2 - True
-- Compiling with MAGMA support
-- MAGMA INCLUDE DIRECTORIES: /home/quent/.local/bin/anaconda3/envs/fairseq/include
-- MAGMA LIBRARIES: /home/quent/.local/bin/anaconda3/envs/fairseq/lib/libmagma.a
-- MAGMA V2 check: 1
-- Could not find hardware support for NEON on this machine.
-- No OMAP3 processor on this machine.
-- No OMAP4 processor on this machine.
-- Looking for cpuid.h
-- Looking for cpuid.h - found
-- Performing Test HAVE_GCC_GET_CPUID
-- Performing Test HAVE_GCC_GET_CPUID - Success
-- Performing Test NO_GCC_EBX_FPIC_BUG
-- Performing Test NO_GCC_EBX_FPIC_BUG - Success
-- Performing Test C_HAS_AVX_1
-- Performing Test C_HAS_AVX_1 - Failed
-- Performing Test C_HAS_AVX_2
-- Performing Test C_HAS_AVX_2 - Success
-- Performing Test C_HAS_AVX2_1
-- Performing Test C_HAS_AVX2_1 - Failed
-- Performing Test C_HAS_AVX2_2
-- Performing Test C_HAS_AVX2_2 - Success
-- Performing Test CXX_HAS_AVX_1
-- Performing Test CXX_HAS_AVX_1 - Failed
-- Performing Test CXX_HAS_AVX_2
-- Performing Test CXX_HAS_AVX_2 - Success
-- Performing Test CXX_HAS_AVX2_1
-- Performing Test CXX_HAS_AVX2_1 - Failed
-- Performing Test CXX_HAS_AVX2_2
-- Performing Test CXX_HAS_AVX2_2 - Success
-- AVX compiler support found
-- AVX2 compiler support found
-- Performing Test BLAS_F2C_DOUBLE_WORKS
-- Performing Test BLAS_F2C_DOUBLE_WORKS - Failed
-- Performing Test BLAS_F2C_FLOAT_WORKS
-- Performing Test BLAS_F2C_FLOAT_WORKS - Success
-- Performing Test BLAS_USE_CBLAS_DOT
-- Performing Test BLAS_USE_CBLAS_DOT - Success
-- Found a library with BLAS API (mkl).
-- Found a library with LAPACK API (mkl).
disabling ROCM because NOT USE_ROCM is set
-- MIOpen not found. Compiling without MIOpen support
-- MKLDNN_THREADING = OMP:COMP
CMake Warning (dev) at third_party/ideep/mkl-dnn/cmake/options.cmake:33 (option):
Policy CMP0077 is not set: option() honors normal variables. Run "cmake
--help-policy CMP0077" for policy details. Use the cmake_policy command to
set the policy and suppress this warning.
For compatibility with older versions of CMake, option is clearing the
normal variable 'MKLDNN_ENABLE_CONCURRENT_EXEC'.
Call Stack (most recent call first):
third_party/ideep/mkl-dnn/cmake/utils.cmake:24 (include)
third_party/ideep/mkl-dnn/CMakeLists.txt:74 (include)
This warning is for project developers. Use -Wno-dev to suppress it.
-- Found OpenMP_C: -fopenmp (found version "4.5")
-- Found OpenMP_CXX: -fopenmp (found version "4.5")
-- OpenMP lib: provided by compiler
-- Could NOT find Doxygen (missing: DOXYGEN_EXECUTABLE)
-- VTune profiling environment is unset
-- Found MKL-DNN: TRUE
-- Looking for clock_gettime in rt
-- Looking for clock_gettime in rt - found
-- Looking for mmap
-- Looking for mmap - found
-- Looking for shm_open
-- Looking for shm_open - found
-- Looking for shm_unlink
-- Looking for shm_unlink - found
-- Looking for malloc_usable_size
-- Looking for malloc_usable_size - found
-- Performing Test C_HAS_THREAD
-- Performing Test C_HAS_THREAD - Success
-- GCC 9.2.0: Adding gcc and gcc_s libs to link line
-- don't use NUMA
-- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT
-- Performing Test COMPILER_SUPPORTS_NO_AVX256_SPLIT - Success
CMake Deprecation Warning at third_party/sleef/CMakeLists.txt:20 (cmake_policy):
The OLD behavior for policy CMP0066 will be removed from a future version
of CMake.
The cmake-policies(7) manual explains that the OLD behaviors of all
policies are deprecated and that a policy should be set to OLD only under
specific short-term circumstances. Projects should be ported to the NEW
behavior and not rely on setting a policy to OLD.
-- Found OpenSSL: /home/quent/.local/bin/anaconda3/envs/fairseq/lib/libcrypto.so (found version "1.1.1d")
-- Check size of long double
-- Check size of long double - done
-- Performing Test COMPILER_SUPPORTS_LONG_DOUBLE
-- Performing Test COMPILER_SUPPORTS_LONG_DOUBLE - Success
-- Performing Test COMPILER_SUPPORTS_FLOAT128
-- Performing Test COMPILER_SUPPORTS_FLOAT128 - Success
-- Performing Test COMPILER_SUPPORTS_SSE2
-- Performing Test COMPILER_SUPPORTS_SSE2 - Success
-- Performing Test COMPILER_SUPPORTS_SSE4
-- Performing Test COMPILER_SUPPORTS_SSE4 - Success
-- Performing Test COMPILER_SUPPORTS_AVX
-- Performing Test COMPILER_SUPPORTS_AVX - Success
-- Performing Test COMPILER_SUPPORTS_FMA4
-- Performing Test COMPILER_SUPPORTS_FMA4 - Success
-- Performing Test COMPILER_SUPPORTS_AVX2
-- Performing Test COMPILER_SUPPORTS_AVX2 - Success
-- Performing Test COMPILER_SUPPORTS_AVX512F
-- Performing Test COMPILER_SUPPORTS_AVX512F - Success
-- Performing Test COMPILER_SUPPORTS_OPENMP
-- Performing Test COMPILER_SUPPORTS_OPENMP - Success
-- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES
-- Performing Test COMPILER_SUPPORTS_WEAK_ALIASES - Success
-- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH
-- Performing Test COMPILER_SUPPORTS_BUILTIN_MATH - Success
-- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM
-- Performing Test COMPILER_SUPPORTS_SYS_GETRANDOM - Success
-- Configuring build for SLEEF-v3.4.0
Target system: Linux-5.3.18-1-MANJARO
Target processor: x86_64
Host system: Linux-5.3.18-1-MANJARO
Host processor: x86_64
Detected C compiler: GNU @ /usr/bin/cc
-- Using option `-Wall -Wno-unused -Wno-attributes -Wno-unused-result -Wno-psabi -ffp-contract=off -fno-math-errno -fno-trapping-math` to compile libsleef
-- Building shared libs : OFF
-- MPFR : /usr/lib/libmpfr.so
-- MPFR header file in /usr/include
-- GMP : /usr/lib/libgmp.so
-- RT : /usr/lib/librt.so
-- FFTW3 : /usr/lib/libfftw3.so
-- OPENSSL : 1.1.1d
-- SDE : SDE_COMMAND-NOTFOUND
-- RUNNING_ON_TRAVIS : 0
-- COMPILER_SUPPORTS_OPENMP : 1
AT_INSTALL_INCLUDE_DIR include/ATen/core
core header install: /home/quent/.local/bin/pytorch/build/aten/src/ATen/core/TensorBody.h
core header install: /home/quent/.local/bin/pytorch/build/aten/src/ATen/core/TensorMethods.h
-- Include NCCL operators
-- Including IDEEP operators
-- Excluding image processing operators due to no opencv
-- Excluding video processing operators due to no opencv
-- Include Observer library
-- /usr/bin/c++ /home/quent/.local/bin/pytorch/caffe2/../torch/abi-check.cpp -o /home/quent/.local/bin/pytorch/build/abi-check
-- Determined _GLIBCXX_USE_CXX11_ABI=1
-- MPI_INCLUDE_PATH: /usr/include
-- MPI_LIBRARIES: /usr/lib/openmpi/libmpi_cxx.so;/usr/lib/openmpi/libmpi.so
-- MPIEXEC: /usr/bin/mpiexec
-- pytorch is compiling with OpenMP.
OpenMP CXX_FLAGS: -fopenmp.
OpenMP libraries: /usr/lib/libgomp.so;/usr/lib/libpthread.so.
-- Caffe2 is compiling with OpenMP.
OpenMP CXX_FLAGS: -fopenmp.
OpenMP libraries: /usr/lib/libgomp.so;/usr/lib/libpthread.so.
-- Using ATen parallel backend: OMP
-- Using lib/python3.8/site-packages as python relative installation path
CMake Warning at CMakeLists.txt:596 (message):
Generated cmake files are only fully tested if one builds with system glog,
gflags, and protobuf. Other settings may generate files that are not well
tested.
--
-- ******** Summary ********
-- General:
-- CMake version : 3.16.3
-- CMake command : /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake
-- System : Linux
-- C++ compiler : /usr/bin/c++
-- C++ compiler id : GNU
-- C++ compiler version : 9.2.0
-- BLAS : MKL
-- CXX flags : -Wno-deprecated -fvisibility-inlines-hidden -fopenmp -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -O2 -fPIC -Wno-narrowing -Wall -Wextra -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Wno-stringop-overflow
-- Build type : Release
-- Compile definitions : TH_BLAS_MKL;ONNX_ML=1;ONNX_NAMESPACE=onnx_torch;MAGMA_V2;IDEEP_USE_MKL;HAVE_MMAP=1;_FILE_OFFSET_BITS=64;HAVE_SHM_OPEN=1;HAVE_SHM_UNLINK=1;HAVE_MALLOC_USABLE_SIZE=1
-- CMAKE_PREFIX_PATH : /home/quent/.local/bin/anaconda3/envs/fairseq;/opt/cuda
-- CMAKE_INSTALL_PREFIX : /home/quent/.local/bin/pytorch/torch
--
-- TORCH_VERSION : 1.5.0
-- CAFFE2_VERSION : 1.5.0
-- BUILD_CAFFE2_MOBILE : ON
-- USE_STATIC_DISPATCH : OFF
-- BUILD_BINARY : OFF
-- BUILD_CUSTOM_PROTOBUF : ON
-- Link local protobuf : ON
-- BUILD_DOCS : OFF
-- BUILD_PYTHON : True
-- Python version : 3.8.1
-- Python executable : /home/quent/.local/bin/anaconda3/envs/fairseq/bin/python
-- Pythonlibs version : 3.8.1
-- Python library : /home/quent/.local/bin/anaconda3/envs/fairseq/lib/libpython3.8.so.1.0
-- Python includes : /home/quent/.local/bin/anaconda3/envs/fairseq/include/python3.8
-- Python site-packages: lib/python3.8/site-packages
-- BUILD_CAFFE2_OPS : ON
-- BUILD_SHARED_LIBS : ON
-- BUILD_TEST : True
-- BUILD_JNI : OFF
-- INTERN_BUILD_MOBILE :
-- USE_ASAN : OFF
-- USE_CUDA : ON
-- CUDA static link : OFF
-- USE_CUDNN : ON
-- CUDA version : 10.2
-- cuDNN version : 7.6.5
-- CUDA root directory : /opt/cuda
-- CUDA library : /opt/cuda/lib/stubs/libcuda.so
-- cudart library : /opt/cuda/lib64/libcudart.so
-- cublas library : /opt/cuda/lib64/libcublas.so
-- cufft library : /opt/cuda/lib64/libcufft.so
-- curand library : /opt/cuda/lib64/libcurand.so
-- cuDNN library : /usr/lib/libcudnn.so
-- nvrtc : /opt/cuda/lib/libnvrtc.so
-- CUDA include path : /opt/cuda/include
-- NVCC executable : /opt/cuda/bin/nvcc
-- NVCC flags : -DONNX_NAMESPACE=onnx_torch;-gencode;arch=compute_35,code=sm_35;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_52,code=sm_52;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_70,code=compute_70;-gencode;arch=compute_75,code=compute_75;-Xcudafe;--diag_suppress=cc_clobber_ignored;-Xcudafe;--diag_suppress=integer_sign_change;-Xcudafe;--diag_suppress=useless_using_declaration;-Xcudafe;--diag_suppress=set_but_not_used;-std=c++14;-Xcompiler;-fPIC;--expt-relaxed-constexpr;--expt-extended-lambda;-Wno-deprecated-gpu-targets;--expt-extended-lambda;-gencode;arch=compute_35,code=sm_35;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_52,code=sm_52;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_70,code=compute_70;-gencode;arch=compute_75,code=compute_75;-Xcompiler -fPIC;-DCUDA_HAS_FP16=1 -D__CUDA_NO_HALF_OPERATORS__ -D__CUDA_NO_HALF_CONVERSIONS__ -D__CUDA_NO_HALF2_OPERATORS__
-- CUDA host compiler : /usr/bin/cc
-- USE_TENSORRT : OFF
-- USE_ROCM : OFF
-- USE_EIGEN_FOR_BLAS :
-- USE_FBGEMM : ON
-- USE_FFMPEG : OFF
-- USE_GFLAGS : OFF
-- USE_GLOG : OFF
-- USE_LEVELDB : OFF
-- USE_LITE_PROTO : OFF
-- USE_LMDB : OFF
-- USE_METAL : OFF
-- USE_MKL : ON
-- USE_MKLDNN : ON
-- USE_MKLDNN_CBLAS : OFF
-- USE_NCCL : ON
-- USE_SYSTEM_NCCL : OFF
-- USE_NNPACK : ON
-- USE_NUMPY : ON
-- USE_OBSERVERS : ON
-- USE_OPENCL : OFF
-- USE_OPENCV : OFF
-- USE_OPENMP : ON
-- USE_TBB : OFF
-- USE_PROF : OFF
-- USE_QNNPACK : ON
-- USE_REDIS : OFF
-- USE_ROCKSDB : OFF
-- USE_ZMQ : OFF
-- USE_DISTRIBUTED : ON
-- USE_MPI : ON
-- USE_GLOO : ON
-- Public Dependencies : Threads::Threads;caffe2::mkl;caffe2::mkldnn
-- Private Dependencies : qnnpack;pytorch_qnnpack;nnpack;cpuinfo;fbgemm;fp16;/usr/lib/openmpi/libmpi_cxx.so;/usr/lib/openmpi/libmpi.so;gloo;aten_op_header_gen;foxi_loader;rt;gcc_s;gcc;dl
-- Configuring done
CMake Warning at caffe2/CMakeLists.txt:627 (add_library):
Cannot generate a safe runtime search path for target torch_cpu because
files in some directories may conflict with libraries in implicit
directories:
runtime library [libgomp.so.1] in /usr/lib may be hidden by files in:
/home/quent/.local/bin/anaconda3/envs/fairseq/lib
Some of these libraries may not be found correctly.
CMake Warning at cmake/Modules_CUDA_fix/upstream/FindCUDA.cmake:1847 (add_library):
Cannot generate a safe runtime search path for target
caffe2_detectron_ops_gpu because files in some directories may conflict
with libraries in implicit directories:
runtime library [libgomp.so.1] in /usr/lib may be hidden by files in:
/home/quent/.local/bin/anaconda3/envs/fairseq/lib
Some of these libraries may not be found correctly.
Call Stack (most recent call first):
modules/detectron/CMakeLists.txt:13 (CUDA_ADD_LIBRARY)
-- Generating done
-- Build files have been written to: /home/quent/.local/bin/pytorch/build
cmake --build . --target install --config Release -- -j 12
...
[2598/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_sparse_normalize_op_gpu.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_sparse_normalize_op_gpu.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_sparse_normalize_op_gpu.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_sparse_normalize_op_gpu.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_sparse_normalize_op_gpu.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_000060b0_00000000-4_sparse_normalize_op_gpu.cpp4.ii".
CMake Error at torch_cuda_generated_sparse_normalize_op_gpu.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_sparse_normalize_op_gpu.cu.o
[2599/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_stump_func_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_stump_func_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_stump_func_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_stump_func_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_stump_func_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_000060cb_00000000-4_stump_func_op.cpp4.ii".
CMake Error at torch_cuda_generated_stump_func_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_stump_func_op.cu.o
[2600/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_thresholded_relu_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_thresholded_relu_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_thresholded_relu_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_thresholded_relu_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_thresholded_relu_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_000060cd_00000000-4_thresholded_relu_op.cpp4.ii".
CMake Error at torch_cuda_generated_thresholded_relu_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_thresholded_relu_op.cu.o
[2601/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_sparse_to_dense_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_sparse_to_dense_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_sparse_to_dense_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_sparse_to_dense_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_sparse_to_dense_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_00006082_00000000-4_sparse_to_dense_op.cpp4.ii".
CMake Error at torch_cuda_generated_sparse_to_dense_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_sparse_to_dense_op.cu.o
[2602/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_space_batch_op_gpu.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_space_batch_op_gpu.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_space_batch_op_gpu.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_space_batch_op_gpu.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_space_batch_op_gpu.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_00006083_00000000-4_space_batch_op_gpu.cpp4.ii".
CMake Error at torch_cuda_generated_space_batch_op_gpu.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_space_batch_op_gpu.cu.o
[2603/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tile_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tile_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tile_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tile_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tile_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_00006104_00000000-4_tile_op.cpp4.ii".
CMake Error at torch_cuda_generated_tile_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tile_op.cu.o
[2604/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tanh_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tanh_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tanh_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tanh_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tanh_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_000060fd_00000000-4_tanh_op.cpp4.ii".
CMake Error at torch_cuda_generated_tanh_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tanh_op.cu.o
[2605/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_swish_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_swish_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_swish_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_swish_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_swish_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_0000610c_00000000-4_swish_op.cpp4.ii".
CMake Error at torch_cuda_generated_swish_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_swish_op.cu.o
[2606/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_softsign_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_softsign_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_softsign_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_softsign_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_softsign_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_0000611c_00000000-4_softsign_op.cpp4.ii".
CMake Error at torch_cuda_generated_softsign_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_softsign_op.cu.o
[2607/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tan_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tan_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tan_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tan_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_tan_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_00006115_00000000-4_tan_op.cpp4.ii".
CMake Error at torch_cuda_generated_tan_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_tan_op.cu.o
[2608/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_spatial_batch_norm_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_spatial_batch_norm_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_spatial_batch_norm_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_spatial_batch_norm_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_spatial_batch_norm_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/cmake/../third_party/eigen/Eigen/src/Core/arch/GPU/PacketMathHalf.h(149): warning: missing return statement at end of non-void function "Eigen::internal::ptrue(const Packet &) [with Packet=half2]"
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_00006148_00000000-4_spatial_batch_norm_op.cpp4.ii".
CMake Error at torch_cuda_generated_spatial_batch_norm_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_spatial_batch_norm_op.cu.o
[2609/3544] Building NVCC (Device) object caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_summarize_op.cu.o
FAILED: caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_summarize_op.cu.o
cd /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -E make_directory /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/. && /home/quent/.local/bin/anaconda3/envs/fairseq/bin/cmake -D verbose:BOOL=OFF -D build_configuration:STRING=Release -D generated_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_summarize_op.cu.o -D generated_cubin_file:STRING=/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_summarize_op.cu.o.cubin.txt -P /home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/torch_cuda_generated_summarize_op.cu.o.Release.cmake
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/ArrayRef.h(278): warning: attribute does not apply to any entity
/home/quent/.local/bin/pytorch/aten/src/ATen/core/ivalue_inl.h(353): warning: pointless comparison of unsigned integer with zero
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
1 error detected in the compilation of "/tmp/tmpxft_0000615d_00000000-4_summarize_op.cpp4.ii".
CMake Error at torch_cuda_generated_summarize_op.cu.o.Release.cmake:281 (message):
Error generating file
/home/quent/.local/bin/pytorch/build/caffe2/CMakeFiles/torch_cuda.dir/operators/./torch_cuda_generated_summarize_op.cu.o
ninja: build stopped: subcommand failed.
Traceback (most recent call last):
File "setup.py", line 737, in <module>
build_deps()
File "setup.py", line 311, in build_deps
build_caffe2(version=version,
File "/home/quent/.local/bin/pytorch/tools/build_pytorch_libs.py", line 62, in build_caffe2
cmake.build(my_env)
File "/home/quent/.local/bin/pytorch/tools/setup_helpers/cmake.py", line 339, in build
self.run(build_args, my_env)
File "/home/quent/.local/bin/pytorch/tools/setup_helpers/cmake.py", line 141, in run
check_call(command, cwd=self.build_dir, env=env)
File "/home/quent/.local/bin/anaconda3/envs/fairseq/lib/python3.8/subprocess.py", line 364, in check_call
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['cmake', '--build', '.', '--target', 'install', '--config', 'Release', '--', '-j', '12']' returned non-zero exit status 1.
```
</details>
The error:
```
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(166): error: expression must have a constant value
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(90): note: expression cannot be interpreted
/home/quent/.local/bin/pytorch/c10/util/TypeIndex.h(110): note: called from:
detected during:
instantiation of "c10::string_view c10::util::get_fully_qualified_type_name<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/c10/util/typeid.h(415): here
instantiation of "c10::string_view caffe2::TypeMeta::TypeName<T>() [with T=caffe2::Tensor]"
/home/quent/.local/bin/pytorch/caffe2/core/blob.h(88): here
``` | module: build,triaged | low | Critical |
557,010,704 | flutter | Framework classes should be able to take pictures instead of images | For example, https://api.flutter.dev/flutter/painting/BoxDecoration-class.html takes a `DecorationImage`, which expect to get a `ImageProvider` so that it can use `drawImage`.
But a user may want to paint a vector graphic as the decoration, which would be more suitable for `drawPicture` - or, alternatively, something like `drawPath`.
It would be nice if we could find a way to make things like `BoxDecoration` etc. to handle painting vector graphics.
| c: new feature,framework,P3,team-framework,triaged-framework | low | Minor |
557,019,054 | TypeScript | Type bug: using mapped type (Partial) & this type & lookup type | <!-- π¨ STOP π¨ π¦π§π’π£ π¨ πΊπ»πΆπ· π¨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker. Even if you think you've found a *bug*, please read the FAQ first, especially the Common "Bugs" That Aren't Bugs section!
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ: https://github.com/Microsoft/TypeScript/wiki/FAQ
Please fill in the *entire* template below.
-->
<!--
Please try to reproduce the issue with the latest published version. It may have already been fixed.
For npm: `typescript@next`
This is also the 'Nightly' version in the playground: http://www.typescriptlang.org/play/?ts=Nightly
-->
**TypeScript Version:** nightly
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:** mapped type, Partial, this, lookup type
**Code**
```ts
class A {
state = { foo: "foo", bar: 42 };
something(): Partial<A["state"]> {
return { foo: "changed" }; // works
}
}
class B {
state = { foo: "foo", bar: 42 };
something(): Partial<this["state"]> {
return { foo: "changed" }; // fails
}
}
class C {
state = { foo: "foo", bar: 42 };
something(): Partial<this["state"]> {
return { foo: "changed", bar: 123 }; // works
}
}
```
**Expected behavior:**
Should type check.
**Actual behavior:**
```
Type '{ foo: "changed"; }' is not assignable to type 'Partial<this["state"]>'.(2322)
```
**Playground Link:** [Example](http://www.typescriptlang.org/play/?ts=3.8.0-dev.20200128&ssl=15&ssc=2&pln=9&pc=1#code/MYGwhgzhAECC0G8BQ1oQC5nQU2gXkWgDMB7EgLmgCJSSqAaaAIzACdKAWAJmgF8BuJCjQkAttnQALAJYA7AOYAKAJSUACm3TSwIADywA2lQxZsVALoA+RMNSsJAV1azCtSlWCSwC7ABMqfPzQAPTB0ADuJKwA1hDCvEgJSKCQMABCNqgmOPiuZO60DMxsnDwCQlliEjIKKuqa2npS0hBG2WZWmajQ9uhOLgjE+dSe3vJ+AQIhYURg0iBxqAlJKVDQAMJd7bmDbtSFjCzs0NyBFSLizbWq0BqsWjq6za3GmDgW1sjdPY7OeRQjLw+fyHErQACMXAAzIFphEorF4okgA)
**Related Issues:** Tried searching for `Partial this`, did not find anything related.
| Needs Investigation,Rescheduled | low | Critical |
557,021,156 | go | cmd/go: support POSIX jobserver | The automatic parallelism in `cmd/go` is wonderful when the command is used interactively. However, it interacts poorly when another build system is driving `cmd/go`. The driving build system cannot safely start multiple Go builds without overloading the system, and it cannot detect when a Go build is doing very little work so it should start other work.
GNU make has an elegant solution for this, the "jobserver." It is much simpler than the name suggests:
> On POSIX systems the jobserver is implemented as a simple UNIX pipe. The pipe will be pre-loaded with one single-character token for each available job. To obtain an extra slot you must read a single character from the jobserver pipe; to release a slot you must write a single character back into the jobserver pipe.
https://www.gnu.org/software/make/manual/html_node/POSIX-Jobserver.html
I would like to propose that `cmd/go` parses the environment variable `MAKEFLAGS` for `--jobserver-auth=R,W`, and if it finds valid FDs `R` and `W` then it uses tokens from the pipe to control its parallelism. This will make `cmd/go` play nicely with GNU make and other build systems that use the jobserver, like redo. | NeedsInvestigation,FeatureRequest | low | Major |
557,043,318 | opencv | NaN comparison fails on certain image sizes | ##### System information (version)
- OpenCV => 4.1.1
- Operating System / Platform => Ubuntu 18.04
- Compiler => Clang 9
##### Detailed description
Certain sizes of images fail when comparing for NaNs
##### Steps to reproduce
```.cpp
#include <opencv2/core.hpp>
int count_NaNs_in_image(int rows, int cols) {
cv::Mat1f test(rows, cols);
test.setTo(std::numeric_limits<float>::quiet_NaN());
cv::Mat1b is_nan = test != test;
return cv::countNonZero(is_nan);
}
int main() {
assert(16 == count_NaNs_in_image(4, 4)); // succeeds
assert(44 == count_NaNs_in_image(4, 11)); // fails with 12 NaNs found instead of 44?
return 0;
}
```
##### Workaround
Invert the comparison, ie
```.cpp
int count_not_not_nans_in_image(int rows, int cols) {
cv::Mat1f test(rows, cols);
test.setTo(std::numeric_limits<float>::quiet_NaN());
cv::Mat1b is_not_nan = test == test;
return cv::countNonZero(255 - is_not_nan);
}
``` | bug,priority: low,category: core,RFC | low | Major |
557,047,289 | vscode | Call Hierarchy view doesn't show number of callers | - VSCode Version: 1.42 Insiders
- OS Version: Windows 10
Found while testing #89386
Would be nice to know how many callers, perhaps in the title bar. Something like "X Callers of 'Area'".

| feature-request,callhierarchy | low | Minor |
557,057,113 | angular | Styles from external components are included outside the scope of a shadow DOM component | # π bug report
### Affected Package
The issue is caused by package `@angular/platform-browser`
### Is this a regression?
No
### Description
When deploying an Angular component with `encapsulation: ViewEncapsulation.ShadowDom`, everything regarding the component should be kept inside the `#shadow-root`. This includes all stylings.
#### Actual behavior
1. Styles added by `styleUrls` or `styles` in the `@Component` decorator are added to the `#shadow-root` βοΈ
2. Styles coming from external components like @angular/material are added to the `#shadow-root` βοΈ
3. Styles coming from external components are *also* added to document's `<head>` outside of the `#shadow-root` β
#### Expected behavior
Only 1. and 2. should happen.
#### Root cause analysis
As far as I can see, this is caused by the `ShadowDomRenderer` using the `DomSharedStylesHost` which adds `_doc.head` to its list of known `_hostNodes` in the constructor:
https://github.com/angular/angular/blob/d43187f7ef7e841d84ecf594e8e515a9d534daaa/packages/platform-browser/src/dom/shared_styles_host.ts#L39
## π¬ Minimal Reproduction
https://stackblitz.com/edit/angular-material-shadowdom-styles
In the `<head>` section there are `<style>` tags of Material components, which should only be inside the shadow DOM.
## π Your Environment
**Angular Version:**
<pre><code>
Angular CLI: 8.3.23
Node: 12.7.0
OS: win32 x64
Angular: 8.2.14
... animations, common, compiler, compiler-cli, core, forms
... language-service, platform-browser, platform-browser-dynamic
... router
Package Version
-----------------------------------------------------------
@angular-devkit/architect 0.803.23
@angular-devkit/build-angular 0.803.23
@angular-devkit/build-optimizer 0.803.23
@angular-devkit/build-webpack 0.803.23
@angular-devkit/core 8.3.23
@angular-devkit/schematics 8.3.23
@angular/cdk 8.2.3
@angular/cli 8.3.23
@angular/material 8.2.3
@ngtools/webpack 8.3.23
@schematics/angular 8.3.23
@schematics/update 0.803.23
rxjs 6.4.0
typescript 3.5.3
webpack 4.39.2
</code></pre>
| type: bug/fix,freq1: low,area: core,state: has PR,core: CSS encapsulation,P3 | medium | Critical |
557,067,561 | flutter | ButtonThemeData does not allow changing elevation, even through subclassing. | I'm building my second production app with Flutter, and one of our major goals with this project is to use more of the pre-built `material` widgets and existing theming capabilities. With our first app, most of the buttons are a `FlatButton` with a background color set. To adhere to our current goals, I want to replace those with `RaisedButton`, since `FlatButton` is defined as _not_ having a background color.
The issue we have at hand is this: if we use `FlatButton`, we cannot override the button color (`getButtonColor`), and if we use `RaisedButton`, we cannot override the elevation (`getElevation`), because neither are data-driven (they rely entirely on the incoming button's `Type` for internal logic) and in `theme_data.dart:337` Flutter rebuilds the button theme using the built-in `ButtonThemeData` type, so using an overriden `getButtonColor` or `getElevation` does not work as expected.
I'd rather this issue not turn into a discussion of _why_ I want a "raised" button to have an elevation of zero. That said, I'm open to any solution that results in:
- Using the built-in `material` Widgets as much as possible.
- Having access to a button with 0 elevation and a filled background. | framework,f: material design,c: proposal,P3,team-design,triaged-design | low | Minor |
557,075,846 | vscode | aggregate Bug/Feature Requests *in editor* for more accurate community upvoting | Currently, feature requests need to receive 20 upvotes within 60 days to be officially added to the Backlog.
This means that users who frequently browse open issues on GitHub will have a disproportionately large impact on which features are adopted.
Users of VSCode who _don't_ come here won't even know that voting system exists, so the **upvote stats will be skewed.**
VSCode is committed to being the most accessible editor on the planet, and one component of that should be making _influencing its development_ accessible to everyone who uses the editor, not just those adept-at or inclined-to browse GitHub issues.
-- Edited to separate the problem, above, from solution ideas, below, since the main thing I want to highlight is the problem of excluding some users from contributing to develop the editor. --
Thanks for considering!
| question-discussion | low | Critical |
557,080,949 | pytorch | Comments Separating Class Methods from Different Classes in C++ Files | ## π Feature
Throughout the PyTorch Distributed codebase (and PyTorch as a whole), a single C++ file may have member functions from several different classes. `process_group_agent.cpp` separates these with a simple one-line comment. This could be a good practice to standardize across the codebase for easier readability.
For Example:
```
////////////////////////////// Work /////////////////////////////////////
void ProcessGroup::Work::abort() {...}
bool ProcessGroup::Work::wait() {...}
////////////////////////// ProcessGroup /////////////////////////////////
ProcessGroup::ProcessGroup(int rank, int size) {...}
ProcessGroup::~ProcessGroup() {...}
``` | triaged,better-engineering | low | Minor |
557,085,385 | pytorch | Implement Backend-Agnostic RPC functionality in RpcAgent | ## π Feature
`ProcessGroupAgent` contains functionality that makes more sense in the `RpcAgent`. Moving general-purpose RPC functionality to the parent class will make for better abstractions in the RPC C++ codebase and ensures core functionality would be available for future communication backends.
## Motivation
In particular, the `futureTimeoutThread_` in ProcessGroupAgent, which cleans up timed out RPC's, the atomic variable `rpcRunning_`, and data structures to track stale RPC's such as the `futureTimeouts_` map are better implemented in a backend-agnostic fashion in `RpcAgent`. This will allow better integration with new features such as RPC Retries and will not need to be re-implemented for each new backend.
cc @pietern @mrshenli @pritamdamania87 @zhaojuanmao @satgera @gqchen @aazzolini @rohan-varma @xush6528 @jjlilley @osalpekar | triaged,better-engineering,module: rpc | low | Minor |
557,102,158 | godot | Android debugging via USB switches screen 3.2 | **Godot version:**
3.2
**OS/device including version:**
Mac OS 14.6/Android 10
**Issue description:**
When I debug my 2d game on my android phone via USB, the tab sometimes randomly switches between 2d and script
**Steps to reproduce:**
Debug game on android via USB, go to different tabs and scenes in the editor. It will randomly switch back between 2d and the script tabs
**Minimal reproduction project:**
| bug,topic:editor | low | Critical |
557,117,341 | TypeScript | Missing IntelliSense for generic parameters of decorators | **TypeScript Version:** 3.7.3 and nightly
**VSCode Version:** 1.42.0-insider (and 1.41.1)
**Search Terms:**
- Generic decorators
- Decorator parameter suggestion
**Example Code:**
In order to simplify the code, I used `keyof T` as parameter type. It should accept any type expression.
```ts
function GenericDecorator<T>(propertyOfClass: keyof T) {
return (target: T, prop: string, dec: PropertyDescriptor) => { ... }
}
export class TestClass {
propertyOne: object = new Object();
propertyTwo: string = "abcdef";
propertyThree: number = 123456;
@GenericDecorator("propertyOne")
testFunction() { ... }
}
```
**Expected behavior:**
Getting valid parameters suggested by IntelliSense.
```ts
@GenericDecorator(..suggestion..)
```
**Actual behavior:**
Compiler is working as expected. Wrong values are underlined / marked as an error, for example:
```ts
@GenericDecorator("propertyOne") // works as expected
...
@GenericDecorator("Not_a_property_of_TestClass") // fails as expected
```
**Playground Link:**
[https://www.typescriptlang.org/play/?experimentalDecorators=true#code/LAKAZgr...](https://www.typescriptlang.org/play/?experimentalDecorators=true#code/LAKAZgrgdgxgLgSwPZQAQHECmVMCcEwAimMSuAhnGQDwAqAfABQAOuSzecAngPJgDCAG3IBnEQC5UAa0xckYVLQCUqAN6hUm1LkxwIuNIzjlcAc12TaAGlSt2kkXHxRTNgCYlJABTYdc3YhEYfGYqXBUAXno1VABfUHiQUEwAD2YyOFQYYTFFTEchURE1UA0tOz9uHhxJJAAjACsSTIjUHAB3VB5G5sYlAG4yzQrOLlp2pAcnBBdUVoAicjqYDzB5wZAtW19R2gALHUxJKAgAWzq8OdQARgAmAGYAFgBWADYNodQAASwcfCISGRKGRGPMRv4xhN5kpPhNcFIZqZiKQKGE+iUkpstIktp8ftg8ARkUC0fMAHJIOAAfXIVPB3Cp8iptHycEKYmhnyglIA6mQES5iaiQSp1JitolEkA)
**Related Issues:** <!-- Did you find other bugs that looked similar? -->
N/A
If needed, I can add some example use cases for this request. | Bug,Domain: Completion Lists | low | Critical |
557,118,528 | TypeScript | VSCode JSDoc doesn't recognize members that are classes from default imported object. | <!-- π¨ STOP π¨ π¦π§π’π£ π¨ πΊπ»πΆπ· π¨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker. Even if you think you've found a *bug*, please read the FAQ first, especially the Common "Bugs" That Aren't Bugs section!
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ: https://github.com/Microsoft/TypeScript/wiki/FAQ
Please fill in the *entire* template below.
-->
<!--
Please try to reproduce the issue with the latest published version. It may have already been fixed.
For npm: `typescript@next`
This is also the 'Nightly' version in the playground: http://www.typescriptlang.org/play/?ts=Nightly
-->
**VSCode Version:** 1.41.1
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
jsdoc export import class object type
**Code**
a.js
```js
export default class A {
constructor() {
this.m = 2;
}
}
```
b.js
```js
export default class B {
constructor() {
this.n = 4;
}
}
```
lib.js
```js
import A from './a';
import B from './b';
const Lib = {
A,
B
};
export default Lib;
```
app.js
```js
import Lib from './lib';
/**
* @param {Lib.A} a
* @param {Lib.B} b
*/
function foo(a, b) {
}
function bar() {
const a = new Lib.A();
foo(new Lib.A(), new Lib.B());
}
```
**Expected behavior:**
When I hover over the `Lib` in the `@param` jsdoc of app.js `foo` function, the tooltip should read something similar to what happens in the bar function:
```
(alias) const Lib: {
A: typeof A;
B: typeof B;
}
import Lib
```
And when I hover over the `Lib.A` in the same area, the tooltip should read something similar to what happens in the bar function:
```
(property) A: new () => A
```
**Actual behavior:**
When I hover over the `Lib` or `Lib.A` in the `@param` jsdoc of app.js `foo` function, the types are `any`, not being recognized as a particular type.
The hovering over types in actual javascript within the bar function works great.
**Related Issues:** <!-- Did you find other bugs that looked similar? -->
I searched, but I'm not seeing any similar issue. Maybe I'm missing something obvious?
**Notes**
I don't believe I saw this behavior in earlier versions of VSCode (maybe even October or September 2019 releases). My code base hasn't changed, but the editor has been upgraded.
| Needs Investigation | low | Critical |
557,150,368 | kubernetes | Suggest `listType=map` if we can guess good key options | We currently fail with an error when generating the openapi if `listType` is missing on lists, and that's a good thing, but we can be even more useful and suggest `listType=map` along with `listMapKeys=...` if we can figure out a good key name.
For example, if we detect that the list is a list of structs that have a field name, that list is probably a map with name as the key, we should help the user and let them know.
This can (and probably should) happen in multiple places:
- [ ] kube-openapi generation. When we generate the kube-openapi, we could replace the existing error by a more useful error suggesting the map value.
- [ ] in kubebuilder, we could also suggest the key in kube-builder as we parse the tags.
- [ ] For CRDs, we could update the status to mention this is a possible useful configuration. | sig/api-machinery,lifecycle/frozen,wg/api-expression | low | Critical |
557,168,604 | flutter | Test that embedders can post native thread tasks recursively. | Not caught earlier because of LUCI [not exercising AOT modes](https://github.com/flutter/flutter/issues/49733). | a: tests,engine,P2,team-engine,triaged-engine | low | Minor |
557,168,799 | go | x/tools/gopls: replace "verboseOutput" with a log label | Rather than having the `"verboseOutput"` setting, we should tag some log messages with a verbose label. | gopls,Tools | low | Minor |
557,191,287 | flutter | Google Map indoor navigation floor selector style looks wrong with Dark mode | <!-- Thank you for using Flutter!
If you are looking for support, please check out our documentation
or consider asking a question on Stack Overflow:
* https://flutter.dev/
* https://api.flutter.dev/
* https://stackoverflow.com/questions/tagged/flutter?sort=frequent
If you have found a bug or if our documentation doesn't have an answer
to what you're looking for, then fill our the template below. Please read
our guide to filing a bug first: https://flutter.dev/docs/resources/bug-reports
-->
## Steps to Reproduce
<!-- You must include full steps to reproduce so that we can reproduce the problem. -->
To reproduce:
1- Add GoogleMaps plugin
2- Set mapType to normal
3- Zoom in to any building that supports indoor view
**Expected results:** <!-- what did you want to see? -->
The level selector at the bottom right side above the floating action button should be white.
**Actual results:** <!-- what did you see? -->
The selector is Black.
Note:
Sometimes when I run the app, it's white.
I have not tried it on android.
<img src="https://user-images.githubusercontent.com/25441876/73408485-5b407780-42ca-11ea-811e-2a23cbdae073.PNG" height="450" />
<details>
<summary>Logs</summary>
<!--
Run your application with `flutter run --verbose` and attach all the
log output below between the lines with the backticks. If there is an
exception, please see if the error message includes enough information
to explain how to solve the issue.
-->
```
PLATFORM_DEVELOPER_APPLICATIONS_DIR =
/Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/Applications
PLATFORM_DEVELOPER_BIN_DIR = /Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/usr/bin
PLATFORM_DEVELOPER_LIBRARY_DIR =
/Applications/Xcode.app/Contents/PlugIns/Xcode3Core.ideplugin/Contents/SharedSupport/Developer/Library
PLATFORM_DEVELOPER_SDK_DIR = /Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/SDKs
PLATFORM_DEVELOPER_TOOLS_DIR = /Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/Tools
PLATFORM_DEVELOPER_USR_DIR = /Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/usr
PLATFORM_DIR = /Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform
PLATFORM_DISPLAY_NAME = iOS
PLATFORM_NAME = iphoneos
PLATFORM_PREFERRED_ARCH = arm64
PLATFORM_PRODUCT_BUILD_VERSION = 17B102
PLIST_FILE_OUTPUT_FORMAT = binary
PLUGINS_FOLDER_PATH = Runner.app/PlugIns
PODS_BUILD_DIR = /Users/michelrahme/Development/Projects/concordia_navigation/build/ios
PODS_CONFIGURATION_BUILD_DIR = /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/Debug-iphoneos
PODS_PODFILE_DIR_PATH = /Users/michelrahme/Development/Projects/concordia_navigation/ios/.
PODS_ROOT = /Users/michelrahme/Development/Projects/concordia_navigation/ios/Pods
PRECOMPS_INCLUDE_HEADERS_FROM_BUILT_PRODUCTS_DIR = YES
PRECOMP_DESTINATION_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build/Debug-iphoneos/Runner.build/PrefixHeaders
PRESERVE_DEAD_CODE_INITS_AND_TERMS = NO
PRIVATE_HEADERS_FOLDER_PATH = Runner.app/PrivateHeaders
PRODUCT_BUNDLE_IDENTIFIER = com.michelrahme.concordiaNavigation
PRODUCT_BUNDLE_PACKAGE_TYPE = APPL
PRODUCT_MODULE_NAME = Runner
PRODUCT_NAME = Runner
PRODUCT_SETTINGS_PATH = /Users/michelrahme/Development/Projects/concordia_navigation/ios/Runner/Info.plist
PRODUCT_TYPE = com.apple.product-type.application
PROFILING_CODE = NO
PROJECT = Runner
PROJECT_DERIVED_FILE_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build/DerivedSources
PROJECT_DIR = /Users/michelrahme/Development/Projects/concordia_navigation/ios
PROJECT_FILE_PATH = /Users/michelrahme/Development/Projects/concordia_navigation/ios/Runner.xcodeproj
PROJECT_NAME = Runner
PROJECT_TEMP_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build
PROJECT_TEMP_ROOT =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
PROVISIONING_PROFILE_REQUIRED = YES
PUBLIC_HEADERS_FOLDER_PATH = Runner.app/Headers
RECURSIVE_SEARCH_PATHS_FOLLOW_SYMLINKS = YES
REMOVE_CVS_FROM_RESOURCES = YES
REMOVE_GIT_FROM_RESOURCES = YES
REMOVE_HEADERS_FROM_EMBEDDED_BUNDLES = YES
REMOVE_HG_FROM_RESOURCES = YES
REMOVE_SVN_FROM_RESOURCES = YES
RESOURCE_RULES_REQUIRED = YES
REZ_COLLECTOR_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build/Debug-iphoneos/Runner.build/ResourceManagerResources
REZ_OBJECTS_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build/Debug-iphoneos/Runner.build/ResourceManagerResources/Objects
SCAN_ALL_SOURCE_FILES_FOR_INCLUDES = NO
SCRIPTS_FOLDER_PATH = Runner.app/Scripts
SCRIPT_OUTPUT_STREAM_FILE =
/var/folders/bv/89l3gwms2kl3f4bx438x1m080000gn/T/flutter_build_log_pipe.fbGnxL/pipe_to_stdout
SDKROOT = /Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/SDKs/iPhoneOS13.2.sdk
SDK_DIR = /Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/SDKs/iPhoneOS13.2.sdk
SDK_DIR_iphoneos13_2 =
/Applications/Xcode.app/Contents/Developer/Platforms/iPhoneOS.platform/Developer/SDKs/iPhoneOS13.2.sdk
SDK_NAME = iphoneos13.2
SDK_NAMES = iphoneos13.2
SDK_PRODUCT_BUILD_VERSION = 17B102
SDK_VERSION = 13.2
SDK_VERSION_ACTUAL = 130200
SDK_VERSION_MAJOR = 130000
SDK_VERSION_MINOR = 200
SED = /usr/bin/sed
SEPARATE_STRIP = NO
SEPARATE_SYMBOL_EDIT = NO
SET_DIR_MODE_OWNER_GROUP = YES
SET_FILE_MODE_OWNER_GROUP = NO
SHALLOW_BUNDLE = YES
SHARED_DERIVED_FILE_DIR =
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/Debug-iphoneos/DerivedSources
SHARED_FRAMEWORKS_FOLDER_PATH = Runner.app/SharedFrameworks
SHARED_PRECOMPS_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/PrecompiledHeaders
SHARED_SUPPORT_FOLDER_PATH = Runner.app/SharedSupport
SKIP_INSTALL = NO
SOURCE_ROOT = /Users/michelrahme/Development/Projects/concordia_navigation/ios
SRCROOT = /Users/michelrahme/Development/Projects/concordia_navigation/ios
STRINGS_FILE_OUTPUT_ENCODING = binary
STRIP_BITCODE_FROM_COPIED_FILES = YES
STRIP_INSTALLED_PRODUCT = YES
STRIP_STYLE = all
STRIP_SWIFT_SYMBOLS = YES
SUPPORTED_DEVICE_FAMILIES = 1,2
SUPPORTED_PLATFORMS = iphonesimulator iphoneos
SUPPORTS_MACCATALYST = NO
SUPPORTS_TEXT_BASED_API = NO
SWIFT_OBJC_BRIDGING_HEADER = Runner/Runner-Bridging-Header.h
SWIFT_OPTIMIZATION_LEVEL = -Onone
SWIFT_PLATFORM_TARGET_PREFIX = ios
SWIFT_VERSION = 5.0
SYMROOT = /Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Products
SYSTEM_ADMIN_APPS_DIR = /Applications/Utilities
SYSTEM_APPS_DIR = /Applications
SYSTEM_CORE_SERVICES_DIR = /System/Library/CoreServices
SYSTEM_DEMOS_DIR = /Applications/Extras
SYSTEM_DEVELOPER_APPS_DIR = /Applications/Xcode.app/Contents/Developer/Applications
SYSTEM_DEVELOPER_BIN_DIR = /Applications/Xcode.app/Contents/Developer/usr/bin
SYSTEM_DEVELOPER_DEMOS_DIR = /Applications/Xcode.app/Contents/Developer/Applications/Utilities/Built Examples
SYSTEM_DEVELOPER_DIR = /Applications/Xcode.app/Contents/Developer
SYSTEM_DEVELOPER_DOC_DIR = /Applications/Xcode.app/Contents/Developer/ADC Reference Library
SYSTEM_DEVELOPER_GRAPHICS_TOOLS_DIR = /Applications/Xcode.app/Contents/Developer/Applications/Graphics Tools
SYSTEM_DEVELOPER_JAVA_TOOLS_DIR = /Applications/Xcode.app/Contents/Developer/Applications/Java Tools
SYSTEM_DEVELOPER_PERFORMANCE_TOOLS_DIR = /Applications/Xcode.app/Contents/Developer/Applications/Performance Tools
SYSTEM_DEVELOPER_RELEASENOTES_DIR = /Applications/Xcode.app/Contents/Developer/ADC Reference Library/releasenotes
SYSTEM_DEVELOPER_TOOLS = /Applications/Xcode.app/Contents/Developer/Tools
SYSTEM_DEVELOPER_TOOLS_DOC_DIR = /Applications/Xcode.app/Contents/Developer/ADC Reference
Library/documentation/DeveloperTools
SYSTEM_DEVELOPER_TOOLS_RELEASENOTES_DIR = /Applications/Xcode.app/Contents/Developer/ADC Reference
Library/releasenotes/DeveloperTools
SYSTEM_DEVELOPER_USR_DIR = /Applications/Xcode.app/Contents/Developer/usr
SYSTEM_DEVELOPER_UTILITIES_DIR = /Applications/Xcode.app/Contents/Developer/Applications/Utilities
SYSTEM_DEXT_INSTALL_PATH = /System/Library/DriverExtensions
SYSTEM_DOCUMENTATION_DIR = /Library/Documentation
SYSTEM_KEXT_INSTALL_PATH = /System/Library/Extensions
SYSTEM_LIBRARY_DIR = /System/Library
TAPI_VERIFY_MODE = ErrorsOnly
TARGETED_DEVICE_FAMILY = 1,2
TARGETNAME = Runner
TARGET_BUILD_DIR = /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/Debug-iphoneos
TARGET_NAME = Runner
TARGET_TEMP_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build/Debug-iphoneos/Runner.build
TEMP_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build/Debug-iphoneos/Runner.build
TEMP_FILES_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build/Debug-iphoneos/Runner.build
TEMP_FILE_DIR =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
/Runner.build/Debug-iphoneos/Runner.build
TEMP_ROOT =
/Users/michelrahme/Library/Developer/Xcode/DerivedData/Runner-edqwnvzxjabtjifwgjoqwcjzzttm/Build/Intermediates.noindex
TOOLCHAIN_DIR = /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain
TRACK_WIDGET_CREATION = true
TREAT_MISSING_BASELINES_AS_TEST_FAILURES = NO
UID = 501
UNLOCALIZED_RESOURCES_FOLDER_PATH = Runner.app
UNSTRIPPED_PRODUCT = NO
USER = michelrahme
USER_APPS_DIR = /Users/michelrahme/Applications
USER_LIBRARY_DIR = /Users/michelrahme/Library
USE_DYNAMIC_NO_PIC = YES
USE_HEADERMAP = YES
USE_HEADER_SYMLINKS = NO
USE_LLVM_TARGET_TRIPLES = YES
USE_LLVM_TARGET_TRIPLES_FOR_CLANG = YES
USE_LLVM_TARGET_TRIPLES_FOR_LD = YES
USE_LLVM_TARGET_TRIPLES_FOR_TAPI = YES
USE_RECURSIVE_SCRIPT_INPUTS_IN_SCRIPT_PHASES = YES
VALIDATE_PRODUCT = NO
VALIDATE_WORKSPACE = NO
VALID_ARCHS = arm64 arm64e armv7 armv7s
VERBOSE_PBXCP = NO
VERBOSE_SCRIPT_LOGGING = YES
VERSIONING_SYSTEM = apple-generic
VERSIONPLIST_PATH = Runner.app/version.plist
VERSION_INFO_BUILDER = michelrahme
VERSION_INFO_FILE = Runner_vers.c
VERSION_INFO_STRING = "@(#)PROGRAM:Runner PROJECT:Runner-1"
WRAPPER_EXTENSION = app
WRAPPER_NAME = Runner.app
WRAPPER_SUFFIX = .app
WRAP_ASSET_PACKS_IN_SEPARATE_DIRECTORIES = NO
XCODE_APP_SUPPORT_DIR = /Applications/Xcode.app/Contents/Developer/Library/Xcode
XCODE_PRODUCT_BUILD_VERSION = 11C504
XCODE_VERSION_ACTUAL = 1131
XCODE_VERSION_MAJOR = 1100
XCODE_VERSION_MINOR = 1130
XPCSERVICES_FOLDER_PATH = Runner.app/XPCServices
YACC = yacc
arch = armv7
variant = normal
2020-01-29 19:01:13.440 xcodebuild[98874:2979610] +dataWithFirstBytes:1024
ofFile:"/Users/michelrahme/Development/flutter/.pub-cache/hosted/pub.dartlang.org/flutter_statusbar_text_color-0.1.2/LICEN
SE" failed, errno = 2
2020-01-29 19:01:14.161 xcodebuild[98874:2979689] DTDeviceKit: deviceType from 00008006-000C192C3C11002E was NULL
2020-01-29 19:01:14.290 xcodebuild[98874:2979610] [MT] DTDeviceKit: deviceType from 00008006-000C192C3C11002E was NULL
2020-01-29 19:01:14.292 xcodebuild[98874:2979610] [MT] DTDeviceKit: deviceType from 00008006-000C192C3C11002E was NULL
2020-01-29 19:01:14.295 xcodebuild[98874:2979610] [MT] DTDeviceKit: deviceType from 00008006-000C192C3C11002E was NULL
[ +489 ms] Installing and launching...
[ ] Debugging is enabled, connecting to observatory
[ +3 ms] executing: /Users/michelrahme/Development/flutter/bin/cache/artifacts/ios-deploy/ios-deploy --id 00008030-001269C01446802E --bundle
build/ios/iphoneos/Runner.app --no-wifi --justlaunch --args --enable-dart-profiling --enable-checked-mode --verify-entry-points
[ +12 ms] [....] Waiting for iOS device to be connected
[ +9 ms] [....] Using 00008030-001269C01446802E (D421AP, D421AP, uknownos, unkarch) a.k.a. 'iPhone'.
[ ] ------ Install phase ------
[ ] [ 0%] Found 00008030-001269C01446802E (D421AP, D421AP, uknownos, unkarch) a.k.a. 'iPhone' connected through USB, beginning install
[ +262 ms] [ 5%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/META-INF/ to device
[ ] [ 5%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/META-INF/com.apple.ZipMetadata.plist to device
[ ] [ 5%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/_CodeSignature/ to device
[ ] [ 5%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/_CodeSignature/CodeResources to device
[ ] [ 5%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/[email protected] to
device
[ ] [ 5%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/[email protected] to
device
[ ] [ 5%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/ to
device
[ ] [ 5%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/[email protected] to device
[ ] [ 5%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/bubble_left.png to device
[ ] [ 5%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/[email protected] to device
[ ] [ 5%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/[email protected] to device
[ ] [ 5%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/ic_error.png to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/[email protected] to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ to
device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/de.lproj/
to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/de.lproj/
GMSCore.strings to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/he.lproj/
to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/he.lproj/
GMSCore.strings to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/en_AU.lpr
oj/ to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/en_AU.lpr
oj/GMSCore.strings to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ar.lproj/
to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ar.lproj/
GMSCore.strings to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/el.lproj/
to device
[ ] [ 6%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/el.lproj/
GMSCore.strings to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ja.lproj/
to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ja.lproj/
GMSCore.strings to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/iw.lproj/
to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/iw.lproj/
GMSCore.strings to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_qu_dir
[email protected] to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/en.lproj/
to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/en.lproj/
GMSCore.strings to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/uk.lproj/
to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/uk.lproj/
GMSCore.strings to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_qu_dir
ection_mylocation.png to device
[ ] [ 7%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/es_419.lp
roj/ to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/es_419.lp
roj/GMSCore.strings to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSSprite
s-0-1x.png to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_compas
[email protected] to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_compas
[email protected] to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/es_MX.lpr
oj/ to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/es_MX.lpr
oj/GMSCore.strings to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_compas
s_needle_large.png to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/zh_CN.lpr
oj/ to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/zh_CN.lpr
oj/GMSCore.strings to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_qu_dir
[email protected] to device
[ ] [ 8%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/DroidSans
Merged-Regular.ttf to device
[ +3 ms] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/nb.lproj/
to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/nb.lproj/
GMSCore.strings to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/es.lproj/
to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/es.lproj/
GMSCore.strings to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_e
ndcap.png to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/pt_BR.lpr
oj/ to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/pt_BR.lpr
oj/GMSCore.strings to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/da.lproj/
to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/da.lproj/
GMSCore.strings to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_locati
[email protected] to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/it.lproj/
to device
[ ] [ 9%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/it.lproj/
GMSCore.strings to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/sk.lproj/
to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/sk.lproj/
GMSCore.strings to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSNavSpr
ites-0-3x.png to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/pt_PT.lpr
oj/ to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/pt_PT.lpr
oj/GMSCore.strings to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_co
mpass_night.png to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_my
_location.png to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_16-4
.png to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ms.lproj/
to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ms.lproj/
GMSCore.strings to device
[ ] [ 10%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_256-
64.png to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_co
mpass.png to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/sv.lproj/
to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/sv.lproj/
GMSCore.strings to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_8-2.
png to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_s
top_and_go.png to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/cs.lproj/
to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/cs.lproj/
GMSCore.strings to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_s
topped.png to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_ba
ckground.png to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSNavNig
htModeSprites-0-1x.png to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_my
[email protected] to device
[ ] [ 11%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/en_IN.lpr
oj/ to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/en_IN.lpr
oj/GMSCore.strings to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ko.lproj/
to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ko.lproj/
GMSCore.strings to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSNavSpr
ites-0-2x.png to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_s
low.png to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_locati
[email protected] to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSNavSpr
ites-0-1x.png to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/no.lproj/
to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/no.lproj/
GMSCore.strings to device
[ ] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/Tharlon-R
egular.ttf to device
[ +100 ms] [ 12%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/hu.lproj/
to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/hu.lproj/
GMSCore.strings to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_s
top_and_go_night.png to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/zh_HK.lpr
oj/ to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/zh_HK.lpr
oj/GMSCore.strings to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_4-1.
png to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_ba
[email protected] to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_compas
[email protected] to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/Assets.ca
r to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/tr.lproj/
to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/tr.lproj/
GMSCore.strings to device
[ ] [ 13%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_f
ast_night.png to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/pl.lproj/
to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/pl.lproj/
GMSCore.strings to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/zh_TW.lpr
oj/ to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/zh_TW.lpr
oj/GMSCore.strings to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSNavNig
htModeSprites-0-2x.png to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSNavNig
htModeSprites-0-3x.png to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/en_GB.lpr
oj/ to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/en_GB.lpr
oj/GMSCore.strings to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/dav_one_w
ay_16_256.png to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_2-1.
png to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/vi.lproj/
to device
[ ] [ 14%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/vi.lproj/
GMSCore.strings to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_ba
[email protected] to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_32-8
.png to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/lv.lproj/
to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/lv.lproj/
GMSCore.strings to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/lt.lproj/
to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/lt.lproj/
GMSCore.strings to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ru.lproj/
to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ru.lproj/
GMSCore.strings to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/polyline_
colors_texture_dim.png to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_s
topped_night.png to device
[ ] [ 15%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/fr_CA.lpr
oj/ to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/fr_CA.lpr
oj/GMSCore.strings to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_128-
32.png to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/fr.lproj/
to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/fr.lproj/
GMSCore.strings to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/fi.lproj/
to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/fi.lproj/
GMSCore.strings to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/id.lproj/
to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/id.lproj/
GMSCore.strings to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/nl.lproj/
to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/nl.lproj/
GMSCore.strings to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/th.lproj/
to device
[ ] [ 16%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/th.lproj/
GMSCore.strings to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_64-1
6.png to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSSprite
s-0-3x.png to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/polyline_
colors_texture.png to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_compas
s_needle.png to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/pt.lproj/
to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/pt.lproj/
GMSCore.strings to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_f
ast.png to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_s
low_night.png to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/road_1-1.
png to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ic_locati
on_off.png to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ro.lproj/
to device
[ ] [ 17%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ro.lproj/
GMSCore.strings to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/traffic_s
tartcap.png to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/Info.plis
t to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/GMSSprite
s-0-2x.png to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_co
[email protected] to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/button_co
[email protected] to device
[ +63 ms] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/hr.lproj/
to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/hr.lproj/
GMSCore.strings to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/hi.lproj/
to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/hi.lproj/
GMSCore.strings to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ca.lproj/
to device
[ ] [ 18%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCoreResources.bundle/ca.lproj/
GMSCore.strings to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/default_marker.png to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/[email protected] to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/oss_licenses_maps.txt.gz to
device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/bubble_right.png to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/[email protected] to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/[email protected] to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/[email protected] to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCacheStorage.momd/ to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCacheStorage.momd/Storage.mom
to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCacheStorage.momd/VersionInfo.
plist to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCacheStorage.momd/StorageWithT
ileProto.mom to device
[ ] [ 19%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCacheStorage.momd/StorageWithT
ileVersionID.mom to device
[ ] [ 20%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/GMSCacheStorage.momd/StorageWithT
ileProto.omo to device
[ ] [ 20%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/GoogleMaps.bundle/Info.plist to device
[ ] [ 20%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon29x29.png to
device
[ ] [ 20%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/[email protected] to
device
[ ] [ 20%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Runner to device
[+1574 ms] [ 23%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/[email protected] to
device
[ ] [ 23%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/[email protected] to
device
[ ] [ 23%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon29x29~ipad.png to
device
[ ] [ 23%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/[email protected] to
device
[ ] [ 23%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/[email protected] to
device
[ ] [ 23%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/[email protected] to
device
[ ] [ 24%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/ to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/Main.storyboardc/ to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/Main.storyboardc/UIViewController-BYZ-38
-t0r.nib/ to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/Main.storyboardc/UIViewController-BYZ-38
-t0r.nib/objects-13.0+.nib to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/Main.storyboardc/UIViewController-BYZ-38
-t0r.nib/runtime.nib to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/Main.storyboardc/BYZ-38-t0r-view-8bC-Xf-
vdC.nib/ to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/Main.storyboardc/BYZ-38-t0r-view-8bC-Xf-
vdC.nib/objects-13.0+.nib to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/Main.storyboardc/BYZ-38-t0r-view-8bC-Xf-
vdC.nib/runtime.nib to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/Main.storyboardc/Info.plist to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/LaunchScreen.storyboardc/ to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/LaunchScreen.storyboardc/01J-lp-oVM-view
-Ze5-6b-2t3.nib/ to device
[ ] [ 24%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/LaunchScreen.storyboardc/01J-lp-oVM-view
-Ze5-6b-2t3.nib/objects-13.0+.nib to device
[ ] [ 25%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/LaunchScreen.storyboardc/01J-lp-oVM-view
-Ze5-6b-2t3.nib/runtime.nib to device
[ ] [ 25%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/LaunchScreen.storyboardc/UIViewControlle
r-01J-lp-oVM.nib/ to device
[ ] [ 25%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/LaunchScreen.storyboardc/UIViewControlle
r-01J-lp-oVM.nib/objects-13.0+.nib to device
[ ] [ 25%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/LaunchScreen.storyboardc/UIViewControlle
r-01J-lp-oVM.nib/runtime.nib to device
[ ] [ 25%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Base.lproj/LaunchScreen.storyboardc/Info.plist to
device
[ ] [ 25%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Assets.car to device
[ +1 ms] [ 25%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppFrameworkInfo.plist to
device
[ ] [ 25%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon40x40@2x~ipad.png
to device
[ ] [ 25%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon76x76@2x~ipad.png
to device
[ ] [ 25%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon83.5x83.5@2x~ipad.png to device
[ ] [ 25%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon29x29@2x~ipad.png
to device
[ ] [ 26%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon40x40~ipad.png to
device
[ ] [ 26%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/ to device
[ ] [ 26%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/libswiftObjectiveC.dylib to device
[ +15 ms] [ 26%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/libswiftCore.dylib to device
[+1308 ms] [ 29%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/libswiftCoreGraphics.dylib to device
[ +52 ms] [ 29%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/flutter_statusbar_text_color.framework/
to device
[ ] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/flutter_statusbar_text_color.framework/_
CodeSignature/ to device
[ ] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/flutter_statusbar_text_color.framework/_
CodeSignature/CodeResources to device
[ ] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/flutter_statusbar_text_color.framework/f
lutter_statusbar_text_color to device
[ +13 ms] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/flutter_statusbar_text_color.framework/I
nfo.plist to device
[ ] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/libswiftDispatch.dylib to device
[ +82 ms] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/libswiftCoreFoundation.dylib to device
[ +16 ms] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/Flutter.framework/ to device
[ ] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/Flutter.framework/_CodeSignature/ to
device
[ ] [ 30%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/Flutter.framework/_CodeSignature/CodeRes
ources to device
[ ] [ 31%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/Flutter.framework/icudtl.dat to device
[ +70 ms] [ 31%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/Flutter.framework/Flutter to device
[+2920 ms] [ 39%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/Flutter.framework/Info.plist to device
[ ] [ 39%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/
to device
[ ] [ 39%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/_CodeSignature/ to device
[ ] [ 40%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/_CodeSignature/CodeResourc
es to device
[ ] [ 40%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/App to device
[ ] [ 40%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/ to device
[ ] [ 40%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/LICENSE to
device
[ +56 ms] [ 40%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/vm_snapshot
_data to device
[ ] [ 40%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/isolate_sna
pshot_data to device
[ +229 ms] [ 41%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/AssetManife
st.json to device
[ ] [ 41%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/kernel_blob
.bin to device
[+1706 ms] [ 45%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/FontManifes
t.json to device
[ ] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/packages/
to device
[ ] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/packages/cu
pertino_icons/ to device
[ ] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/packages/cu
pertino_icons/assets/ to device
[ ] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/packages/cu
pertino_icons/assets/CupertinoIcons.ttf to device
[ ] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/fonts/ to
device
[ ] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/flutter_assets/fonts/Mater
ialIcons-Regular.ttf to device
[ +2 ms] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/App.framework/Info.plist to device
[ ] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/libswiftDarwin.dylib to device
[ +23 ms] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/location.framework/ to device
[ ] [ 46%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/location.framework/_CodeSignature/ to
device
[ ] [ 47%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/location.framework/_CodeSignature/CodeRe
sources to device
[ ] [ 47%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/location.framework/location to device
[ +39 ms] [ 47%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/location.framework/Info.plist to device
[ ] [ 47%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/libswiftFoundation.dylib to device
[ +661 ms] [ 49%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/path_provider.framework/ to device
[ ] [ 49%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/path_provider.framework/_CodeSignature/
to device
[ ] [ 49%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/path_provider.framework/_CodeSignature/C
odeResources to device
[ ] [ 49%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/path_provider.framework/path_provider to
device
[ +14 ms] [ 49%] Copying
/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Frameworks/path_provider.framework/Info.plist to
device
[ ] [ 49%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon20x20~ipad.png to
device
[ ] [ 49%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/embedded.mobileprovision
to device
[ ] [ 49%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon20x20@2x~ipad.png
to device
[ ] [ 49%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/Info.plist to device
[ ] [ 49%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/PkgInfo to device
[ ] [ 49%] Copying /Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app/AppIcon76x76~ipad.png to
device
[ +285 ms] [ 52%] CreatingStagingDirectory
[ ] [ 57%] ExtractingPackage
[ ] [ 60%] InspectingPackage
[ +31 ms] [ 60%] TakingInstallLock
[ +41 ms] [ 65%] PreflightingApplication
[ +74 ms] [ 65%] InstallingEmbeddedProfile
[ +6 ms] [ 70%] VerifyingApplication
[ +659 ms] [ 75%] CreatingContainer
[ +4 ms] [ 80%] InstallingApplication
[ +4 ms] [ 85%] PostflightingApplication
[ +2 ms] [ 90%] SandboxingApplication
[ +8 ms] [ 95%] GeneratingApplicationMap
[ +103 ms] [100%] Installed package build/ios/iphoneos/Runner.app
[ +118 ms] ------ Debug phase ------
[ ] Starting debug of 00008030-001269C01446802E (D421AP, D421AP, uknownos, unkarch) a.k.a. 'iPhone' connected through USB...
[ +521 ms] [ 0%] Looking up developer disk image
[ +14 ms] [ 95%] Developer disk image mounted successfully
[ +345 ms] [100%] Connecting to remote debug server
[ ] -------------------------
[ +30 ms] (lldb) command source -s 0 '/tmp/27DA1917-BB9B-42AA-9589-23A78DB405FE/fruitstrap-lldb-prep-cmds-00008030_001269C01446802E'
[ ] Executing commands in '/tmp/27DA1917-BB9B-42AA-9589-23A78DB405FE/fruitstrap-lldb-prep-cmds-00008030_001269C01446802E'.
[ ] (lldb) platform select remote-ios --sysroot '/Users/michelrahme/Library/Developer/Xcode/iOS DeviceSupport/13.3 (17C54)/Symbols'
[ ] Platform: remote-ios
[ ] Connected: no
[ ] SDK Path: "/Users/michelrahme/Library/Developer/Xcode/iOS DeviceSupport/13.3 (17C54)/Symbols"
[ ] (lldb) target create "/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app"
[+4079 ms] Current executable set to '/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos/Runner.app' (arm64).
[ ] (lldb) script
fruitstrap_device_app="/private/var/containers/Bundle/Application/52E74235-C290-43F3-A596-47ED5058738A/Runner.app"
[ +209 ms] (lldb) script fruitstrap_connect_url="connect://127.0.0.1:57223"
[ ] (lldb) script fruitstrap_output_path=""
[ ] (lldb) script fruitstrap_error_path=""
[ ] (lldb) target modules search-paths add /usr "/Users/michelrahme/Library/Developer/Xcode/iOS DeviceSupport/13.3
(17C54)/Symbols/usr" /System "/Users/michelrahme/Library/Developer/Xcode/iOS DeviceSupport/13.3 (17C54)/Symbols/System"
"/private/var/containers/Bundle/Application/52E74235-C290-43F3-A596-47ED5058738A"
"/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos"
"/var/containers/Bundle/Application/52E74235-C290-43F3-A596-47ED5058738A"
"/Users/michelrahme/Development/Projects/concordia_navigation/build/ios/iphoneos" /Developer "/Users/michelrahme/Library/Developer/Xcode/iOS
DeviceSupport/13.3 (17C54)/Symbols/Developer"
[ +20 ms] (lldb) command script import "/tmp/27DA1917-BB9B-42AA-9589-23A78DB405FE/fruitstrap_00008030_001269C01446802E.py"
[ +3 ms] (lldb) command script add -f fruitstrap_00008030_001269C01446802E.connect_command connect
[ ] (lldb) command script add -s asynchronous -f fruitstrap_00008030_001269C01446802E.run_command run
[ ] (lldb) command script add -s asynchronous -f fruitstrap_00008030_001269C01446802E.autoexit_command autoexit
[ ] (lldb) command script add -s asynchronous -f fruitstrap_00008030_001269C01446802E.safequit_command safequit
[ ] (lldb) connect
[ +27 ms] (lldb) run
[ +231 ms] success
[ ] (lldb) safequit
[ +110 ms] Process 21374 detached
[ +45 ms] Application launched on the device. Waiting for observatory port.
[ +3 ms] Checking for advertised Dart observatories...
[+5022 ms] Checking for available port on com.michelrahme.concordiaNavigation._dartobservatory._tcp.local
[ +2 ms] Checking for authentication code for com.michelrahme.concordiaNavigation._dartobservatory._tcp.local
[ +3 ms] Attempting to forward device port 61069 to host port 1024
[ ] executing: /Users/michelrahme/Development/flutter/bin/cache/artifacts/usbmuxd/iproxy 1024 61069 00008030-001269C01446802E
[+1007 ms] Forwarded port ForwardedPort HOST:1024 to DEVICE:61069
[ +4 ms] Installing and launching... (completed in 22.3s)
[ +7 ms] Connecting to service protocol: http://127.0.0.1:1024/wZckrDVUIIc=/
[ +144 ms] Successfully connected to service protocol: http://127.0.0.1:1024/wZckrDVUIIc=/
[ +6 ms] Sending to VM service: getVM({})
[ +4 ms] Result: {type: VM, name: vm, architectureBits: 64, hostCPU: Unknown, operatingSystem: ios, targetCPU: arm64, version: 2.7.0 (Mon
Dec 2 20:10:59 2019 +0100) on "ios_arm64", _profilerMode: VM, _nativeZoneMemoryUsage: 0, pid: 21374, startTime: 1580342493960,...
[ +4 ms] Sending to VM service: getIsolate({isolateId: isolates/82715888898403})
[ +3 ms] Sending to VM service: _flutter.listViews({})
[ +21 ms] Result: {type: Isolate, id: isolates/82715888898403, name: main, number: 82715888898403, _originNumber: 82715888898403, startTime:
1580342493984, _heaps: {new: {type: HeapSpace, name: new, vmName: Scavenger, collections: 3, avgCollectionPeriodMillis: 2000...
[ +12 ms] Result: {type: FlutterViewList, views: [{type: FlutterView, id: _flutterView/0x108020c20, isolate: {type: @Isolate, fixedId: true,
id: isolates/82715888898403, name: main.dart$main-82715888898403, number: 82715888898403}}]}
[ +5 ms] DevFS: Creating new filesystem on the device (null)
[ +1 ms] Sending to VM service: _createDevFS({fsName: concordia_navigation})
[ +10 ms] Result: {type: FileSystem, name: concordia_navigation, uri:
file:///private/var/mobile/Containers/Data/Application/2132C1F2-15CC-4CD5-B447-6A56DA75A986/tmp/concordia_navigationNQdfGQ/concordia_navigatio
n/}
[ ] DevFS: Created new filesystem on the device
(file:///private/var/mobile/Containers/Data/Application/2132C1F2-15CC-4CD5-B447-6A56DA75A986/tmp/concordia_navigationNQdfGQ/concordia_navigati
on/)
[ +1 ms] Updating assets
[ +71 ms] Syncing files to device iPhone...
[ +1 ms] Scanning asset files
[ +1 ms] <- reset
[ ] Compiling dart to kernel with 0 updated files
[ +7 ms] /Users/michelrahme/Development/flutter/bin/cache/dart-sdk/bin/dart
/Users/michelrahme/Development/flutter/bin/cache/artifacts/engine/darwin-x64/frontend_server.dart.snapshot --sdk-root
/Users/michelrahme/Development/flutter/bin/cache/artifacts/engine/common/flutter_patched_sdk/ --incremental --target=flutter
-Ddart.developer.causal_async_stacks=true --output-dill /var/folders/bv/89l3gwms2kl3f4bx438x1m080000gn/T/flutter_tool.onZFDQ/app.dill
--packages /Users/michelrahme/Development/Projects/concordia_navigation/.packages -Ddart.vm.profile=false -Ddart.vm.product=false
--bytecode-options=source-positions,local-var-info,debugger-stops,instance-field-initializers,keep-unreachable-code,avoid-closure-call-instruc
tions --enable-asserts --track-widget-creation --filesystem-scheme org-dartlang-root
[ +4 ms] <- compile package:concordia_navigation/main.dart
[+5699 ms] Updating files
[ +79 ms] DevFS: Sync finished
[ ] Syncing files to device iPhone... (completed in 5,795ms, longer than expected)
[ ] Synced 0.9MB.
[ +1 ms] Sending to VM service: _flutter.listViews({})
[ +2 ms] Result: {type: FlutterViewList, views: [{type: FlutterView, id: _flutterView/0x108020c20, isolate: {type: @Isolate, fixedId: true,
id: isolates/82715888898403, name: main.dart$main-82715888898403, number: 82715888898403}}]}
[ ] <- accept
[ ] Connected to _flutterView/0x108020c20.
[ +1 ms] π₯ To hot reload changes while running, press "r". To hot restart (and rebuild state), press "R".
[ ] An Observatory debugger and profiler on iPhone is available at: http://127.0.0.1:1024/wZckrDVUIIc=/
[ ] For a more detailed help message, press "h". To detach, press "d"; to quit, press "q".
```
<!--
Run `flutter analyze` and attach any output of that command below.
If there are any analysis errors, try resolving them before filing this issue.
-->
```
Analyzing concordia_navigation...
info β’ The value of the field '_controller' isn't used β’ lib/screens/map.dart:22:23 β’ unused_field
1 issue found. (ran in 2.0s)
```
<!-- Finally, paste the output of running `flutter doctor -v` here. -->
```
[β] Flutter (Channel stable, v1.12.13+hotfix.7, on Mac OS X 10.15.2 19C57, locale en-CA)
β’ Flutter version 1.12.13+hotfix.7 at /Users/michelrahme/Development/flutter
β’ Framework revision 9f5ff2306b (3 days ago), 2020-01-26 22:38:26 -0800
β’ Engine revision a67792536c
β’ Dart version 2.7.0
[β] Android toolchain - develop for Android devices (Android SDK version 29.0.2)
β’ Android SDK at /Users/michelrahme/Library/Android/sdk
β’ Android NDK location not configured (optional; useful for native profiling support)
β’ Platform android-29, build-tools 29.0.2
β’ Java binary at: /Applications/Android Studio.app/Contents/jre/jdk/Contents/Home/bin/java
β’ Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b49-5587405)
β’ All Android licenses accepted.
[β] Xcode - develop for iOS and macOS (Xcode 11.3.1)
β’ Xcode at /Applications/Xcode.app/Contents/Developer
β’ Xcode 11.3.1, Build version 11C504
β’ CocoaPods version 1.8.4
[β] Android Studio (version 3.5)
β’ Android Studio at /Applications/Android Studio.app/Contents
β’ Flutter plugin version 42.1.1
β’ Dart plugin version 191.8593
β’ Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b49-5587405)
[β] Connected device (1 available)
β’ iPhone β’ 00008030-001269C01446802E β’ ios β’ iOS 13.3
β’ No issues found!
```
</details>
| platform-ios,p: maps,package,has reproducible steps,P2,found in release: 1.21,found in release: 2.0,found in release: 2.2,team-ios,triaged-ios | low | Critical |
557,205,782 | terminal | InputEngineTest::RoundTripTest Fails | Roses are red,
Violets are blue,
InputEngineTest::RoundTripTest fails,
So I left a TODO. | Product-Conpty,Area-VT,Issue-Task | low | Minor |
557,206,606 | godot | intersect_ray() ignores any TileMap cells with collision polygons after a call to TileMap.set_cell() | **Godot version:** 3.2
**OS/device including version:** Manjaro Linux 18.1.5
**Issue description:** If you change a cell (on a TileMap) from a tile with collision polygon to a tile without collision polygon, the next call to Physics2DDirectSpaceState.intersect_ray() will not collide with any TileMap cells. This only happens if a call to intersect_ray() is made during next _physics_process tick right after set_cell() call had happened on a _process() tick.

I believe it is a regression, since this does not happen on version 3.1.2.
**Steps to reproduce:**
1) During _process() tick, call TileMap.set_cell() and change a cell from a tile with collision polygon to a tile without.
2) On the next _physics_process() tick, call Physics2DDirectSpaceState.intersect_ray() directed at a TileMap cell with a collision polygon.
**Minimal reproduction project:**
[ReproductionProject.zip](https://github.com/godotengine/godot/files/4131627/ReproductionProject.zip)
| bug,confirmed,topic:physics,regression,topic:2d | low | Minor |
557,213,307 | flutter | Desktop app frame rate isn't synced to display | I'm running on a 120hz monitor, I've scheduled a persistent frame callback and can see that I get ~60fps. It would be great if Flutter could use something like CADisplayLink on MacOS to sync frame rendering with the display.
Gist of a simple app showing this: https://gist.github.com/luigi-rosso/0e3ed273e52764bf1d7bdacca49cccd2

I can see in various [sites](https://jsfiddle.net/greggman/ULxVp/), apps, and games that the display is actually refreshing at 120hz.

Previously filed here: https://github.com/google/flutter-desktop-embedding/issues/656
Fix status:
* [X] Windows (note: has the right rate, but may be out of phase)
* [x] macOS
* [ ] Linux | engine,c: performance,platform-mac,platform-linux,a: desktop,perf: speed,has reproducible steps,P2,found in release: 3.19,found in release: 3.22,team-macos,triaged-macos | low | Critical |
557,250,858 | tensorflow | Ability to calculate projected memory usage for a given model | <em>Please make sure that this is a feature request. As per our [GitHub Policy](https://github.com/tensorflow/tensorflow/blob/master/ISSUES.md), we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:feature_template</em>
**System information**
- TensorFlow version (you are using): 2.1
- Are you willing to contribute it (Yes/No): Yes, but I would likely need assistance/guidance
**Describe the feature and the current behavior/state.**
The closest feature I can find to this is `tf.keras.Model.summary()`. This feature calculate the trainable and non-trainable parameters per-layer but doesn't attempt to calculate any memory usage statistics.
The proposed feature would extend `tf.keras.Model.summary()` by also calculating memory requirements per-layer and for the complete model. If multiple compute devices are used in the model then a per-device memory breakdown would also be useful.
**Will this change the current api? How?**
This could either just add functionality to `tf.keras.Model.summary()` without changing the interface, or it could add extra parameters to `tf.keras.Model.summary()`. Alternatively, a new function could be added, whatever is deemed most appropriate.
**Who will benefit with this feature?**
Anyone with a desire to get an idea about how much memory their model will need for either training or inference. It should be possible for someone to determine whether the model they wish to train/evaluate will fit into RAM (either system of GPU) that they have available.
**Any Other info.**
My focus is on the `tf.keras.Model.summary()` API, but this feature should probably be extended to other means of model creation.
| stat:awaiting tensorflower,type:feature,comp:keras,TF 2.1 | low | Critical |
557,264,367 | puppeteer | request.headers() does not give all headers that are used when making the request | ```js
const browser = await puppeteer.launch();
const page = await browser.newPage();
page.on('request', async (request) => {
await request.headers();
});
await page.goto('http://gajus.com');
```
just gives:
```
'upgrade-insecure-requests': '1',
'user-agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) HeadlessChrome/79.0.3945.0 Safari/537.36',
'sec-fetch-user': '?1'
```
whereas the actual headers include additional headers such as: accept, accept-language and accept encoding.
Related:
* https://github.com/puppeteer/puppeteer/issues/3436 (closed without a solution) | bug,upstream,confirmed,P3 | medium | Critical |
557,276,889 | rust | ExitCodeExt: Please provide an extension method for the "raw" exit code | On Unix platforms, `ExitCode` wraps the exit code received from the system, and provides helper functions to distinguish "exited normally with exit code" from "exited with signal number". However, some command-line tools (those that execute another program) want to exit with the same exit status as the program they run. Please consider providing a function to obtain the "raw" exit status, suitable to pass directly to `sys::process::exit`. | T-libs-api,C-feature-request | low | Minor |
557,287,288 | TypeScript | `lib.d.ts` isn't ordered appropriately for signature help | 
This extremely broad signature help is occurring because `lib.d.ts` is ordered from oldest to newest. Unfortunately that means that users end up getting complex signatures that they never cared about. Even if we came up with a *Matcher* interface, it's not clear that that's even useful.
| Bug | low | Minor |
557,289,318 | TypeScript | Define protocol type names for well-known symbols | 
This is pretty hard to grok. Maybe we should have a named type for types like
```
{
[Symbol.match](string: string): RegExpMatchArray;
}
```
It's not totally clear what we'd call these - they're awkward. Something like `RegExpMatcher`? | Suggestion,Experience Enhancement | low | Minor |
557,300,448 | go | cmd/compile: prove misses obvious facts about constants | I noticed that left shifts by a constant are not being marked as bounded by prove. Prove consults ft.limits, but ft.limits doesn't contain limits for constants. Something similar happens for some other ops with easy limits.
This is pretty straightforward to fix: Write a wrapper around ft.limits accesses that generates limits in cases like these.
I have a prototype CL of this, and it mostly helps. But before it can be mailed, I need to investigate and fix a few more minor regressions from it. (The compiler is quite sensitive to the order in which optimizations occur.)
And before doing that, I wanted to check in about whether this was even the right kind of fix here.
cc @zdjones @rasky
| Performance,NeedsInvestigation,compiler/runtime | low | Minor |
557,323,639 | pytorch | Error tracing custom autograd.Function | ## π Bug
I got an error when I tried to use jit to trace my autograd Function. In previous Pytorch versions(1.1.0), this works, but recently I updated to 1.4.0, and this stopped working. I'm not exactly sure which version exactly did this stopped working.
## To Reproduce
Steps to reproduce the behavior:
1. Make sure you have Pytorch1.4.0
2. Make a custom autograd Function. In my case, I'm using this one
```
class raw(autograd.Function):
@staticmethod
def forward(ctx, inp):
ctx.a = (inp * inp + 1).reciprocal()
ctx.b = ctx.a.sqrt()
return inp * ctx.b
@staticmethod
def backward(ctx, grad_output):
return grad_output * ctx.a * ctx.b
```
3. Trace this Function.
```
from torch import jit
jit.trace(raw.apply, example_inputs=tc.randn(1))
```
4. The error should orginate from the line of tracing and look like this:
```
......................
jit.trace(raw.apply, example_inputs=tc.randn(1))
File "...\Python37\lib\site-packages\torch\jit\__init__.py", line 903, in trace
name = _qualified_name(func)
File "...\Python37\lib\site-packages\torch\_jit_internal.py", line 696, in _qualified_name
"__module__ can't be None.".format(name))
RuntimeError: Could not get qualified name for class 'apply': __module__ can't be None.
```
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
## Expected behavior
I expect the tracing to work like how the previous versions did.
<!-- A clear and concise description of what you expected to happen. -->
## Environment
Please copy and paste the output from our
[environment collection script](https://raw.githubusercontent.com/pytorch/pytorch/master/torch/utils/collect_env.py)
(or fill out the checklist below manually).
You can get the script and run it with:
```
wget https://raw.githubusercontent.com/pytorch/pytorch/master/torch/utils/collect_env.py
# For security purposes, please check the contents of collect_env.py before running it.
python collect_env.py
```
PyTorch version: 1.4.0+cpu
Is debug build: No
CUDA used to build PyTorch: None
OS: Microsoft Windows 10 Home
GCC version: Could not collect
CMake version: Could not collect
Python version: 3.7
Is CUDA available: No
CUDA runtime version: No CUDA
GPU models and configuration: No CUDA
Nvidia driver version: No CUDA
cuDNN version: No CUDA
Versions of relevant libraries:
[pip3] numpy==1.16.3
[pip3] torch==1.4.0+cpu
[pip3] torchvision==0.5.0+cpu
[conda] Could not collect
## Additional context
I'm bad at using Github so if I missed something please tell me :)
<!-- Add any other context about the problem here. -->
cc @suo | oncall: jit,triaged | low | Critical |
557,356,610 | go | strings/bytes: LastIndexByte is significantly slower than IndexByte | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13.1 darwin/amd64
</pre>
### Does this issue reproduce with the latest release?
yes.
### What operating system and processor architecture are you using (`go env`)?
Both Linux/MacOS
### What did you do?
I was using multipart.NewReader() to process multi-part responses from Cloud REST API.
It turned out that ~1/3 of profile is spent in mime/multipart/multipart.go :: scanUntilBoundary() -> bytes.LastIndexByte().
After looking into it, it is no wonder as bytes.LastIndexByte() is not using any optimisations and compiled into simple loop iterating over bytes, no REP SCASB instruction is used on Intel (nor SSE).
### What did you expect to see?
bytes.LastIndexByte() to use SSE or at least REP SCASB optimised code.
### What did you see instead?
simple byte to byte loop in asm code.
| Performance,help wanted,NeedsFix | low | Major |
557,395,671 | go | x/website/cmd/golangorg: Add Content Security Policy | CSP is an important protection against some of the higher risk web vulnerabilities and the official Go website doesn't currently adopt it.
Moreover CSP is a internal requirement for any website hosted on *.google.eTLD and the Go website is currently also hosted on golang.google.cn.
I can take care of fixing this or finding someone that can work on it if the proposal is accepted.
/cc @dmitshur @andybons | NeedsFix | low | Minor |
557,400,291 | rust | Exponential trait selection when compiling a crate using combine 4 | Originally reported in https://github.com/Marwes/combine/issues/284 . It appears that the changes made between version 3 and 4 in https://github.com/Marwes/combine . Made trait selection exponential in some cases. The main change that I would suspect causing this is that combine-3 had `Input` as an associated type whereas combine-4 uses a type parameter.
The following minimized repo reproduces the slowdown, removing a few arguments from the `choice!` macro makes it compile quickly https://github.com/Marwes/combine-slow-compile . The commit before `master` contains a version with combine-3 which compiles instantaneously (after dependencies are compiled). (diff https://github.com/Marwes/combine-slow-compile/commit/21cf38a5429e32a703c2bc2ea02cbcbaf2985b01)
 | A-trait-system,I-compiletime,T-compiler,C-bug | low | Major |
557,426,404 | flutter | [image_picker][iOS] freeze when clicking the camera take photo button multiple times quickly | Using plugin [image_picker], when presenting the camera view, clicking the take photo button multiple times quickly will lead to the application freeze at the preview page with "Retake"/"Use Photo" bottom sheet. No response when clicking the buttons on the bottom sheet.
The plugin works fine if only pressing the take photo button once.
Platform: iOS 13.3
Flutter Doctor:
```
Doctor summary (to see all details, run flutter doctor -v):
[β] Flutter (Channel stable, v1.12.13+hotfix.5, on Mac OS X 10.14.6 18G95,
locale en-HK)
[!] Android toolchain - develop for Android devices (Android SDK version 28.0.3)
! Some Android licenses not accepted. To resolve this, run: flutter doctor
--android-licenses
[β] Xcode - develop for iOS and macOS (Xcode 10.3)
[β] Android Studio (version 3.4)
[β] Connected device (1 available)
```
Logs:
```
2020-01-30 19:03:59.924768+0800 projectg_demo_ios[719:82926] [VERBOSE-2:ui_dart_state.cc(157)] Unhandled Exception: PlatformException(multiple_request, Cancelled by a second request, null)
#0 StandardMethodCodec.decodeEnvelope (package:flutter/src/services/message_codecs.dart:569:7)
#1 MethodChannel.invokeMethod (package:flutter/src/services/platform_channel.dart:321:33)
<asynchronous suspension>
#2 ImagePicker.pickImage (package:image_picker/image_picker.dart:64:40)
#3 InputBarWidgetState._openCameraBottomSheet.<anonymous closure>.<anonymous closure> (package:project_g/ui/screens/conversation/input/widget_input_bar.dart:470:51)
#4 BSCameraSelectorWidget.build.<anonymous closure> (package:project_g/ui/widgets/widget_bs_camera_selector.dart:28:15)
#5 _InkResponseState._handleTap (package:flutter/src/material/ink_well.dart:706:14)
#6 _InkResponseState.build.<anonymous closure> (package:flutter/src/material/ink_well.dart:789:36)
#7 GestureRecognizer.invokeCallback (package:flutter/src/gestures/recognizer.dart:182:24)
#8 TapGestureRecognizer.handleTapUp (package:flutter/src/gestures/tap.dart:486:11)
#9 BaseTapGestureRecognizer._checkUp (package:flutter/src/gestures/tap.dart:264:5)
#10 BaseTapGestureRecognizer.acceptGesture (package:flutter/src/gestures/tap.dart:236:7)
#11 GestureArenaManager.sweep (package:flutter/src/gestures/arena.dart:156:27)
#12 GestureBinding.handleEvent (package:flutter/src/gestures/binding.dart:222:20)
#13 GestureBinding.dispatchEvent (package:flutter/src/gestures/binding.dart:198:22)
#14 GestureBinding._handlePointerEvent (package:flutter/src/gestures/binding.dart:156:7)
#15 GestureBinding._flushPointerEventQueue (package:flutter/src/gestures/binding.dart:102:7)
#16 GestureBinding._handlePointerDataPacket (package:flutter/src/gestures/binding.dart:86:7)
#17 _rootRunUnary (dart:async/zone.dart:1138:13)
#18 _CustomZone.runUnary (dart:async/zone.dart:1031:19)
#19 _CustomZone.runUnaryGuarded (dart:async/zone.dart:933:7)
#20 _invoke1 (dart:ui/hooks.dart:273:10)
#21 _dispatchPointerDataPacket (dart:ui/hooks.dart:182:5)
2020-01-30 19:03:59.936992+0800 projectg_demo_ios[719:82775] Warning: Attempt to present <UIImagePickerController: 0x106007000> on <FlutterViewController: 0x105882c00> whose view is not in the window hierarchy!
```
Thank you for help. | platform-ios,c: performance,p: image_picker,package,perf: speed,has reproducible steps,P2,found in release: 3.7,found in release: 3.9,team-ios,triaged-ios | low | Major |
557,434,080 | flutter | Focus gets lost on Android TV | Run the code below. It displays two raised buttons and a text field and another raised button.
You can use the DPad (or the cursor keys if trying this on the simulator) to change the focus between both buttons. I changed their focus color to make this very obvious. You can active buttons with RETURN. You cannot activate them using the SELECT button on the DPad but that's something I can fix myself and but part of the bug report.
However, once you focus the text field, you cannot get your focus back and the app ist stuck until you stop and start it again. A hot restart is not enough to free the focus.
Once the text field has the focus, the keyboard opens. On the simulator, it has a DONE button. Selecting that button (don't press ENTER, that's cheating) closes the keyboard βΒ but only if the text field has no `onEditingComplete` callback! As a workaround, overriding `onSubmitted` is triggered on DONE and then the keyboard still closes.
I would expect that `onEditingComplete` also works.
Adding `f.requestFocus();` to the `onSubmitted` callback (`f` is the focus node of the third raised button) is a workaround to get the focus back:
```dart
TextField(
onSubmitted: (s) => f.requestFocus();
),
```
If I now again enter the text field, this warnings are printed:
```
W/IInputConnectionWrapper( 4205): getTextBeforeCursor on inactive InputConnection
W/IInputConnectionWrapper( 4205): getTextAfterCursor on inactive InputConnection
W/IInputConnectionWrapper( 4205): getTextBeforeCursor on inactive InputConnection
W/IInputConnectionWrapper( 4205): getTextAfterCursor on inactive InputConnection
W/IInputConnectionWrapper( 4205): deleteSurroundingText on inactive InputConnection
W/IInputConnectionWrapper( 4205): commitText on inactive InputConnection
```
Something is wrong here.
My expectation is that the text field works as with a native Android app:
* when focused, it is displayed as "highlighted" but no keyboard opens
* when selected while focused, the keyboard opens, you can edit until DONE is selected and the keyboard closes while the text field stays focused
* while focused, you can use UP to move to the previous button (after moving the cursor to the beginning)
* while focused, you can use DOWN to move to the next button (after moving the cursor to the end)
* while focused, you can use LEFT and RIGHT to move the cursor.
```dart
import 'package:flutter/material.dart';
void main() => runApp(MyApp());
class MyApp extends StatelessWidget {
@override
Widget build(BuildContext context) {
return MaterialApp(
home: MyHomePage(),
);
}
}
class MyHomePage extends StatefulWidget {
MyHomePage({Key key}) : super(key: key);
@override
_MyHomePageState createState() => _MyHomePageState();
}
class _MyHomePageState extends State<MyHomePage> {
final f = FocusNode();
@override
Widget build(BuildContext context) {
return Scaffold(
body: Center(
child: Column(
mainAxisAlignment: MainAxisAlignment.center,
children: <Widget>[
RaisedButton(
focusColor: Colors.yellow,
onPressed: () => print("1"),
child: Text("Hallo!"),
),
RaisedButton(
focusColor: Colors.amber,
onPressed: () => print("2"),
child: Text("Hallo!"),
),
SizedBox(
width: 200,
child: TextField(
onSubmitted: (s) {
print(s);
f.requestFocus();
},
),
),
RaisedButton(
focusNode: f,
focusColor: Colors.lime,
onPressed: () => print("3"),
child: Text("Hallo!"),
),
],
),
),
);
}
}
```
<details>
<summary>Logs</summary>
<!--
Run your application with `flutter run --verbose` and attach all the
log output below between the lines with the backticks. If there is an
exception, please see if the error message includes enough information
to explain how to solve the issue.
-->
```
```
<!--
Run `flutter analyze` and attach any output of that command below.
If there are any analysis errors, try resolving them before filing this issue.
-->
```
```
<!-- Finally, paste the output of running `flutter doctor -v` here. -->
```
[β] Flutter (Channel master, v1.14.7-pre.38, on Mac OS X 10.14.6 18G103, locale de-DE)
β’ Flutter version 1.14.7-pre.38 at /Users/sma/Work/flutter
β’ Framework revision 92f7e16312 (vor 9 Stunden), 2020-01-29 17:51:31 -0800
β’ Engine revision 6007c17fd2
β’ Dart version 2.8.0 (build 2.8.0-dev.5.0 fc3af737c7)
[β] Android toolchain - develop for Android devices (Android SDK version 28.0.3)
β’ Android SDK at /Users/sma/Library/Android/sdk
β’ Android NDK location not configured (optional; useful for native profiling support)
β’ Platform android-28, build-tools 28.0.3
β’ Java binary at: /Applications/Android Studio.app/Contents/jre/jdk/Contents/Home/bin/java
β’ Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b49-5587405)
β’ All Android licenses accepted.
[β] Android Studio (version 3.5)
β’ Android Studio at /Applications/Android Studio.app/Contents
β’ Flutter plugin version 42.1.1
β’ Dart plugin version 191.8593
β’ Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b49-5587405)
[β] VS Code (version 1.41.1)
β’ VS Code at /Applications/Visual Studio Code.app/Contents
β’ Flutter extension version 3.7.1
[β] Connected device (6 available)
β’ AFTB β’ 192.168.178.57:5555 β’ android-arm β’ Android 5.1.1 (API 22)
β’ sdk google atv x86 β’ emulator-5554 β’ android-x86 β’ Android 7.0 (API 24) (emulator)
```
</details>
| a: text input,framework,f: material design,f: focus,P2,team-framework,triaged-framework | low | Critical |
557,470,612 | node | Macos Derived exceptions seem to be affected by RTTI settings in the gypfile | I don't know whether it will affect any nodejs user, but since it may and I found it out the hard way I'm putting this here:
```
'xcode_settings': {
'GCC_ENABLE_CPP_RTTI': 'NO'
}
```
will lead to that derived exeptions in a program linked against that library may not work anymore; so this tiny test programm linked against that library:
```
#include <stdexcept>
#include <iostream>
int main(int argc, char* argv[]) {
std::cout << "start test" << std::endl;
try {
std::cout << "throw The Error" << std::endl;
throw std::runtime_error(std::string("test"));
}
catch (std::exception const& x) {
std::cout << "catch error" << std::endl;
}
return 1;
}
```
may result in `catch error` not being printed, but a runtime exception happening.
If you don't find this usefull - simply close it. | build,macos | low | Critical |
557,487,232 | pytorch | MKLDNN doesnt work and is slower than normal cpu mode | ## π Bug
MKLDNN seems not to have any effect on runtime performance. Its even some times, slower than normal cpu mode!
The issue exists both on 1.3.1 and 1.4.0.
## To Reproduce
Steps to reproduce the behavior:
Run this on a supporting platform such as Linux (e.g. ubuntu 18.04)
```python
import torch
print(f'Pytorch version : {torch.__version__}')
print(*torch.__config__.show().split("\n"), sep="\n")
from torchvision import models
from torch.utils import mkldnn as mkldnn_utils
import time
def forward(net, use_mkldnn=False, iteration=1, batch_size=10):
net.eval()
batch = torch.rand(batch_size, 3,224,224)
if use_mkldnn:
net = mkldnn_utils.to_mkldnn(net)
batch = batch.to_mkldnn()
start_time = time.time()
for i in range(iteration):
net(batch)
return time.time() - start_time
net = models.resnet18(False)
iter_cnt = 100
batch_size = 1
no_mkldnn = forward(net, False, iter_cnt, batch_size)
with_mkldnn = forward(net, True, iter_cnt, batch_size)
print(f"time-normal: {no_mkldnn:.4f}s")
print(f"time-mkldnn: {with_mkldnn:.4f}s")
print(f"mkldnn is {with_mkldnn/no_mkldnn:.2f}x slower!")
```
Or simply run this on a [Google Colab : ](https://colab.research.google.com/drive/1OKBSZY5rc5w9QIKfyo5trS3JTgBEvpPp)
## Expected behavior
Should be 10x or more faster than the normal CPU mode.
## Environment
- PyTorch Version (e.g., 1.0): 1.4.0+cpu
- OS (e.g., Linux): 18.04.3 LTS
- How you installed PyTorch (`conda`, `pip`, source): pip
- Build command you used (if compiling from source): -
- Python version: tested on both 3.6 / 3.7.3
- CUDA/cuDNN version: tested both on gpu and cpu builds: 10.1.243 / 7.6.5
- GPU models and configuration: None
- Any other relevant information: -
cc @gujinghui @PenghuiCheng @XiaobingSuper @jianyuh | triaged,module: mkldnn | low | Critical |
557,503,340 | flutter | video_player: Autoplay support | Im aware browser might block autoplay videos but this is only partially true.
This is also stated in the article youve referenced:
> Chrome's autoplay policies will change in April of 2018 and I'm here to tell you why and how this is going to affect **video playback with sound**
[https://developers.google.com/web/updates/2017/09/autoplay-policy-changes](https://developers.google.com/web/updates/2017/09/autoplay-policy-changes)
Coming from webdev, adding `playsinline` and `muted` attributes will enable autoplaying in most browsers.
> Chrome's autoplay policies are simple:
> - Muted autoplay is always allowed.
> [...]
## Use case
Many
## Proposal
Autoplay videos.
| c: new feature,platform-web,p: video_player,package,P3,team-web,triaged-web | low | Minor |
557,537,599 | godot | Group related functions doesn't work well with tool enabled | **Godot version:**
3.2
**OS/device including version:**
ubuntu 19.10
**Issue description:**
I'm creating a plugin related to group managing, but when use a function like node.add_to_group(groupname) the groups built'in tab doesnt recognize this (sometimes it works, but if you first add a number string:

**Steps to reproduce:**
Create a plugin
Create a way to add_groups in-editor
ImprovedGroups.gd
```
tool
extends EditorPlugin
class_name ImprovedGroups
var dock
func _enter_tree():
# Initialization of the plugin goes here
# Load the dock scene and instance it
dock = load("res://addons/improved_groups/Groups.tscn").instance()
# Add the loaded scene to the docks
add_control_to_dock(DOCK_SLOT_RIGHT_UR, dock)
dock.configure(self)
func _exit_tree():
# Clean-up of the plugin goes here
# Remove the dock
remove_control_from_docks(dock)
# Erase the control from the memory
dock.free()
func get_selected_node()-> Node:
return get_editor_interface().get_selection().get_selected_nodes()[0]
```
ManagerGroups.gd (the dock scene root script)
```
tool
extends Control
class_name ManagerGroups
export var path_group_name:NodePath
export var path_apply_children:NodePath
export var path_add:NodePath
var le_group_name:LineEdit
var cb_apply_children:CheckBox
var bt_add:Button
var improved_groups:ImprovedGroups = null
func configure(caller:ImprovedGroups):
le_group_name = get_node(path_group_name)
cb_apply_children = get_node(path_apply_children)
bt_add = get_node(path_add)
improved_groups = caller
print("configured")
pass # Replace with function body.
func add_group():
var selected_node = improved_groups.get_selected_node()
var add_to_children = cb_apply_children.pressed
var group_name = le_group_name.text
if group_name == "":
printerr("Group name cant be empty")
return
if selected_node == null:
printerr("Select a node first")
return
if (selected_node.is_in_group(group_name)):
printerr("Node "+selected_node.name+" already in this group")
return
else:
selected_node.add_to_group(group_name,true)
prints("Added Group: ",group_name," to node: ",selected_node.name)
le_group_name.text = ""
pass
func _on_bt_add_pressed():
add_group()
pass # Replace with function body.
```
** Update
checked godot's source, possibily could be cuz of this line:
groups_editor.cpp line 195
```
void GroupDialog::_add_group(String p_name) {
if (!is_visible()) {
return; // No need to edit the dialog if it's not being used.
}
``` | bug,topic:editor | low | Major |
557,570,206 | storybook | Setting the default width/height of the addons panel? | In Storybook 5.3, I would like to change the default width/height (depending on the orientation) of the addons panel to be able to have the components and the story source side by side, like in the screenshot.
<img width="1552" alt="Screen Shot 2020-01-30 at 16 27 08" src="https://user-images.githubusercontent.com/425406/73463474-964fb300-437d-11ea-8c52-e699276d5dc7.png">
Is there a way to accomplish this currently? I could not find any option in the documentation. | feature request,ui,has workaround | medium | Major |
557,581,726 | youtube-dl | Add www.paramountnetwork.it/ | Hi, is to add https://www.paramountnetwork.it/ for example https://www.paramountnetwork.it/film/218i3r/agatha-christie-le-due-verita-parte-1
thank you | site-support-request,question | low | Minor |
557,613,860 | react | Bug: Render-phase update to another root causes an over-rendering loop | ```js
let container1 = document.createElement('div');
let container2 = document.createElement('div');
function Root1() {
_setX(x => x + 1);
return 'Hello';
}
let _setX = () => {};
function Root2() {
console.log('Root 2')
let [x, setX] = React.useState(0);
_setX = setX;
return <div>{x}</div>
}
const root1 = ReactDOM.createRoot(container1);
root1.render(<Root1 />);
const root2 = ReactDOM.createRoot(container2);
root2.render(<Root2 />);
```
Expected: `Root2` renders twice.
Actual: `Root2` renders a non-deterministic number of times.
This bisects to https://github.com/facebook/react/pull/15755, which removed the `5000` clamping. Without it, we get updates that have a slightly larger timeout than `5000`, and so the loop doesn't stop until we get to `5000`. | Type: Regression | low | Critical |
557,655,704 | go | cmd/compile: possible latent codegen issue on amd64 removing zero extensions | I'm not sure whether it is possible to trigger this bug right now, but I think there may be an issue lurking.
Consider code like:
```go
func f(x uint32) uint64 {
return uint64(x & 0xFFFFFFFF)
}
```
The outmost op is `ZeroExt32to64`, which gets lowered to `MOVLQZX`. The innermost op ends up being `(ANDLconst [0xFFFFFFFF] x)`.
Then this rule triggers: `(MOVLQZX x) && zeroUpper32Bits(x,3) -> x`. `ANDLconst` is among the ops listed as zeroing the upper 32 bits of a register, so we eliminate the zero extension.
Then this rule triggers: `(ANDLconst [c] x) && int32(c)==-1 -> x`, eliminating the `ANDLconst`.
This leaves us with just `x`, but without any zero extension, which could leave junk in the top half of the register.
As it stands, this doesn't happen in this function, because the And32 is eliminated during generic optimization. But there are ways to sneak an `& -1` past generic optimization, e.g. with a constant shift, which is how I discovered this issue. We are also saved in this case by using a MOVL to load the value from an arg. But we could be using a computed value instead. So I haven't convinced myself that this couldn't actually cause bad code generation right now for just the right input.
There are two possible diagnoses/fixes:
* zeroUpper32Bits assumes that the inner op actually gets executed, which is wrong, in which case we need to pare back the list of accepted ops, probably to only loads, which I think cannot be eliminated at that point (?), since dse occurs earlier
* the ANDLconst rule is unsound because eliminating it eliminates the side-effects of zeroing the top half of the register
I'm strongly inclined to the first diagnosis and fix. Thoughts?
cc @randall77 @cherrymui
| NeedsInvestigation,compiler/runtime | low | Critical |
557,657,891 | godot | Joystick Navigation broken for OptionButton | Using the new Godot v3.2 release, with Ubuntu 18.04.3 LTS
I've been messing around with 3.2, and I noticed that I can now assign Joy Axis inputs to the ui_left/right/up/down actions, and for the most part the menu navigation works as expected. Pushing a direction on the joystick only moves the selection once until the joystick position is reset, which seems like the expected behavior to me (whereas in the last version of Godot the selection would jump every single time the joystick moved at all).
However, I still notice a problem when I use an OptionButton, where after I select it and move the joystick, it retains the problematic behavior from before where any slight movements cause the selection within the OptionButton menu to jump. To me, the expected behavior would be for the selection to only move up or down the first time the joystick is moved up or down, until its position is reset, just like with other UI navigation. I suspect that this problematic behavior may occur in other UI elements too though I haven't checked yet.
Steps to reproduce:
1. Create a new project
2. Modify the Input Map by adding `Device 0, Axis 0 - (Left Stick Left).` to ui_left, etc. for all 4 ui directions
3. Create a new Control scene, add a container, then add an OptionButton to that container
4. Add a number of items to the OptionButton
5. (Optional) Add a script to the control to grab the focus of the OptionButton on _ready
6. Run the project, then use a controller (I used an XBox One S controller) to click on the OptionButton
7. Move the joystick up and down to see how the selection jumps too quickly
Sample project (includes one of each button type):
[JoystickUITest.zip](https://github.com/godotengine/godot/files/4135469/JoystickUITest.zip)
I did notice some other focus problems in that sample project, but I can go ahead and create new issues for those if needed (to summarize, MenuButtons and LinkButtons aren't focused when navigating past them, even if they are explicitly defined as neighbors). | bug,confirmed,usability,topic:input,topic:gui | low | Critical |
557,677,354 | flutter | Cupertino pickers are not focusable with a keyboard | @goderbauer
@gspencergoog
@lorenzOliveto | framework,a: accessibility,f: date/time picker,f: cupertino,a: desktop,customer: soldier,f: focus,has reproducible steps,P2,found in release: 3.3,found in release: 3.7,team-design,triaged-design | low | Major |
557,720,958 | go | all: update standard-library dependencies at the start and end of each development cycle | The `x` dependencies vendored into the standard library (via `src/go.mod` and `src/cmd/go.mod`) should be updated for each code freeze, so that we can apply any needed fixes to those dependencies without _also_ pulling in unnecessary changes or requiring significant backporting work (see, for example, #36851).
This task, like #11811, #12042, and the API audit (#36167, #32813), should occur regularly in each development cycle. Probably we should update the dependencies at the beginning of each code freeze, and again when we reopen the tree at the end of each code freeze.
CC @golang/osp-team | NeedsFix,early-in-cycle,release-blocker,recurring | high | Critical |
557,727,316 | godot | Viewport as texture GLES2 - weird glitch | **Godot version**: 3.2
**GLES2**
with GLES3 it works fine
OS/device including version: Windows 10/ASUS VivoBook (doesn't work on other devices too)
Issue description:
Assigning viewport to quad mesh doesn't work in a particular scenario.
Steps to reproduce:
Create scene with camera, quad mesh and viewport with label assigned to the quad mesh.

Set the environment "Mode" to custom color and choose black.

Run the game.

Everything works well with different color

| bug,topic:rendering | low | Major |
557,731,328 | terminal | The Resize Window escape sequence doesn't support omitted and zero parameters | # Environment
```none
Windows build number: Version 10.0.18362.535
```
# Steps to reproduce
1. Open a bash shell in conshot
2. Set the window size to 80x25 with `printf "\e[8;25;80t"`
3. Set just the height parameter to 30 with `printf "\e[8;30t"`
4. Set just the width parameter to 100 with `printf "\e[8;;100t"`
5. Set both width and height to zero with `printf "\e[8;0;0t"`
# Expected behavior
According to the [XTerm documentation](https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h4-Functions-using-CSI-_-ordered-by-the-final-character-lparen-s-rparen:CSI-Ps;Ps;Ps-t:Ps-=-8;height;width.2519), omitted parameters should reuse the current height or width, while zero parameters should use the display's height or width.
* Step 2 should set the window size to 80x25
* Step 3 should set the window size to 80x30 (i.e. just changing the height)
* Step 4 should set the window size to 100x30 (i.e. just changing the width)
* Step 5 should set the window size to full screen.
# Actual behavior
Step 2 works as expected, but the remaining steps have no effect.
That said, this functionality doesn't seem to be widely supported (the only terminals I've seen that got it right were XTerm and Mintty), so it's probably not that important. | Product-Conhost,Area-VT,Issue-Bug | low | Major |
557,746,647 | TypeScript | Parameter types not being inferred in a class method whose type is indexed from an interface | <!-- π¨ STOP π¨ π¦π§π’π£ π¨ πΊπ»πΆπ· π¨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker. Even if you think you've found a *bug*, please read the FAQ first, especially the Common "Bugs" That Aren't Bugs section!
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ: https://github.com/Microsoft/TypeScript/wiki/FAQ
Please fill in the *entire* template below.
-->
<!--
Please try to reproduce the issue with the latest published version. It may have already been fixed.
For npm: `typescript@next`
This is also the 'Nightly' version in the playground: http://www.typescriptlang.org/play/?ts=Nightly
-->
**TypeScript Version:** 3.7.4
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
Index type
Parameter type of method
Inferring parameter types
**Code**
```ts
interface Parameters1 {
paramA: number;
paramB: string;
}
interface Parameters2 {
paramC: boolean;
}
interface SomeFunctions {
foo(options: Parameters1): void;
bar(options: Parameters2): void;
}
class AClass implements SomeFunctions {
public foo(options: Parameters1): void { }
public bar(options: Parameters2): void { }
}
const aClassInstance = new AClass();
function callMethod<K extends keyof SomeFunctions>(
methodName: K,
options: Parameters<SomeFunctions[K]>[0]
) {
// 1. Does not work:
// Argument of type 'Parameters<SomeFunctions[K]>[0]' is not assignable to parameter of type 'Parameters1 & Parameters2'.
aClassInstance[methodName](options);
// 2. Also does not work
// Argument of type 'Parameters<SomeFunctions[K]>[0]' is not assignable to parameter of type 'Parameters1 & Parameters2'.
// Type 'Parameters<SomeFunctions[K]>[0]' is not assignable to type 'Parameters1'.
(aClassInstance[methodName] as SomeFunctions[K])(options);
// 3. Does work, even though it is essentially the same thing as 2
type MethodType = (o: typeof options) => ReturnType<SomeFunctions[K]>
(aClassInstance[methodName] as MethodType)(options);
}
```
**Expected behavior:**
Case 2 should work. Possibly also case 1.
Typescript should know that options is the correct type
**Actual behavior:**
Only case 3 works, even though it is more or less the same as case 2
**Playground Link:** <!-- A link to a TypeScript Playground "Share" link which demonstrates this behavior -->
[Playground](https://www.typescriptlang.org/play/?ssl=2&ssc=22&pln=1&pc=1#)
**Related Issues:** <!-- Did you find other bugs that looked similar? -->
| Needs Investigation | low | Critical |
557,764,099 | pytorch | Can we add support for Enum in scripted models? | Enums are not supported right now
import torch
from torch import nn
from enum import Enum
class Testing(Enum):
lala = "lala"
class Mo(nn.Module):
def __init__(self, la):
super().__init__()
self.la = la
def forward(self, x):
if self.la == Testing.lala:
return x
return x+1
saved_mo = torch.jit.script(mo)
RuntimeError:
Module 'Mo' has no attribute 'la' (This attribute exists on the Python module, but we failed to convert Python type: 'Testing' to a TorchScript type.):
File "<ipython-input-18-76956f0fdf61>", line 7
def forward(self, x):
if self.la == Testing.lala:
~~~~~~~ <--- HERE
return x
return x+1
cc @suo | oncall: jit,triaged | low | Critical |
557,779,750 | terminal | Safer SafeArrays | It'd be nice if we could be safer with SafeArrays. We use them a lot in our accessibility model.
Some ideas include:
- deleting automatically if anything goes wrong
- easy conversions to/from vectors/arrays
I submitted an issue on wil [here](https://github.com/microsoft/wil/issues/114). If it gets implemented there first, we should use that. If not, we should just create it here and then pass it up to wil. | Help Wanted,Product-Terminal,Issue-Task,Area-CodeHealth | low | Minor |
557,805,259 | vscode | Untitled file title gets in my way when trying to save | - Open an untitled file with some contents
- Try to save it in the custom file dialog
- The generated title is inserted into the input every time I complete a folder segment

| under-discussion,simple-file-dialog,workbench-untitled-editors | low | Major |
557,806,873 | electron | effective test for select-client-certificate | the current test for `select-client-certificate` doesn't actually exercise the code path unless there is a client certificate installed on the machine, which there isn't on CI.
fixing this might involve either a) figuring out how to install client certs on our CI machines in a programmatic way, or b) exposing new APIs to control the client cert store so that we can test this effectively. | enhancement :sparkles: | low | Minor |
557,808,762 | terminal | Additional Testing for UiaTextRange | # Description of the new feature/enhancement
After #4018, more tests were added to UiaTextRange. However, there are still gaps in our testing and they require some more thought and additional refactoring. Here's some that I have in mind:
- [x] `MoveByWord`
- [x] `MoveEndpointByUnitWord`
- [ ] `FindText`
- [ ] `GetBoundingRectangles`
- [ ] `GetText`
The word navigation tests are currently relying more on the tests within `TextBuffer`. So they might not need to be as cumbersome.
FindText relies on our search module so that also isn't a high priority.
This is more of an opportunity to compare our accessibility model to others' and make sure we're not doing anything wonky (and preventing us from getting to that point).
These new tests also serve as a good opportunity to use `TEST_METHOD_PROPERTY`. Ideally, we could just have a standard set of tests across all text units and just iterate through those units.
| Area-Accessibility,Product-Terminal,Issue-Task,Area-CodeHealth | low | Minor |
557,846,959 | go | runtime: allow map hashes with different tradeoffs | <!--
Please answer these questions before submitting your issue. Thanks!
For questions please use one of our forums: https://github.com/golang/go/wiki/Questions
-->
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13.4 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
Yes
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
go version go1.13.4 linux/amd64
esr@snark:~/WWW/reposurgeon$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/home/esr/.cache/go-build"
GOENV="/home/esr/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/home/esr/go"
GOPRIVATE=""
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/usr/lib/go-1.13"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/usr/lib/go-1.13/pkg/tool/linux_amd64"
GCCGO="/usr/bin/gccgo"
AR="ar"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD="/home/esr/WWW/reposurgeon/go.mod"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build890210982=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
Wrote a complicated data transformation using maps heavily. Observed that after I did enough optimization, map hashing costs actually competed with GC as a performance bottleneck.
This is where the standard bug template stops being useful. The problem I see is that the choice of aeshash512 is more expensive than is right for my application, and I can't fix that.
I'm not arguing with the choice to optimize for cryptographic hardness over performance by default; given Golang's intended role as a language for network servers this was sensible. But my application - reposurgeon - doesn't need that hardening, and is paying a significant performance cost for it.
Therefore, this RFE: Offer a runtime switch that allows plugging in a weaker but faster hash.
| Performance,NeedsInvestigation,compiler/runtime | low | Critical |
557,848,583 | TypeScript | intellisense to reference computed field directly | <!-- π¨ STOP π¨ π¦π§π’π£ π¨ πΊπ»πΆπ· π¨
Half of all issues filed here are duplicates, answered in the FAQ, or not appropriate for the bug tracker.
Please help us by doing the following steps before logging an issue:
* Search: https://github.com/Microsoft/TypeScript/search?type=Issues
* Read the FAQ, especially the "Common Feature Requests" section: https://github.com/Microsoft/TypeScript/wiki/FAQ
-->
## Search Terms
suggestions computed key
interface computed key
intellisense computed key
## Suggestion
<!-- A summary of what you'd like to see added or changed -->
Currently when using a computed field on a type, intellisense does not show it in the typescript playground, and in VSCode it suggests using `obj[varName]` style.
## Use Cases
I declare some const vars to alias keys in an sql query, and then I want to use those const vars to build a dynamic interface. It's a lot more readable and easier to write if I can get intellisense for dot notation rather than bracket notation
## Examples
```ts
function TestA() {
const compKey = "bar";
const obj = {} as unknown as {
foo: string;
[compKey]: number;
};
//got intellisense for this
const a = obj.foo;
//no intellisense for this in ts-playground,
//in VSCode intellisense suggests obj[compKey]
//desired intellisense would be obj.var
const b = obj.bar;
//this intellisense is shown in VSCode, but not the playground
const c = obj[compKey];
}
//Work around
function TestB() {
let obj;
{
const compKey = "bar";
obj = {} as unknown as {
foo: string;
[compKey]: number;
};
}
//got intellisense for this
const a = obj.foo;
//no intellisense for this in ts-playground, but works fine in VSCode
const b = obj.bar;
//error as expected as compKey out of scope
const c = obj[compKey];
}
```
http://www.typescriptlang.org/play/?ssl=39&ssc=2&pln=1&pc=1#code/GYVwdgxgLglg9mABAFQKYGcoEEAUBKRAbwChEzEIFMK4BbABwGlUBPRAXkQCIAjAQwBOXANylylMNTg8AVhyIBfRH3SJwAazBwA7khVEx5csDhwAXIkwCYYAOaijRgNqUGzFgF0LYELR6oBB3IFUUNEAHpw2zgoRBsoVAAbRJh0VElURBMBRCgAC1SwiWo+eWkZADoTOFCjSK04sATk1PS0rLgc-NTG3PQAWnpEvhZbAThwABMAGjDIm0QANQBlAGE4Scz4pJS0jMsQW1sMKFVylzomVg858M30GAFUScbm3bbM7QnEl-9EcoqADdBEUqLEeGVZBV+IFEMRbt1VNsWnt2j10HkdEgFit1ptpogeCBYlpYvlMkMRmMJmBJqDJLEIJCZBc3NdRAp4ZEAOqddTKcZTYigSCwBAoE4AIXwBiMiVQsXKQTIJEc4jBNDZbE4vEEIjCRnK8kISn0Gi0umUqlVasc1QsVhs9gNttZV083l8-kCLrIITCnNu0ViyPe+2yuQKqjhRmKsVKnAB1Vq5HqcFeO1a4c6kZ6C1Og2Go0FtIJRNiXwE6lUwBsWyQuI2qHp1AhiahMOEcNuAXGOX0qAAHvRUNBnlbNe7-sT-sBLJQRy3Gcy3e4PBygA
## Checklist
My suggestion meets these guidelines:
* [x] This wouldn't be a breaking change in existing TypeScript/JavaScript code
* [x] This wouldn't change the runtime behavior of existing JavaScript code
* [x] This could be implemented without emitting different JS based on the types of the expressions
* [x] This isn't a runtime feature (e.g. library functionality, non-ECMAScript syntax with JavaScript output, etc.)
* [x] This feature would agree with the rest of [TypeScript's Design Goals](https://github.com/Microsoft/TypeScript/wiki/TypeScript-Design-Goals).
| Bug | low | Critical |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.