id
int64 393k
2.82B
| repo
stringclasses 68
values | title
stringlengths 1
936
| body
stringlengths 0
256k
⌀ | labels
stringlengths 2
508
| priority
stringclasses 3
values | severity
stringclasses 3
values |
|---|---|---|---|---|---|---|
596,776,848 |
godot
|
[Bullet] 3d physics stepping is two frames late for bullet physics
|
**Godot version:**
3.2.1-stable
**OS/device including version:**
Mac
**Issue description:**
3d physics stepping is two frames late using bullet physics. Expected would be that it is one frame late.
**Steps to reproduce:**
Apply either velocity or force to a 3d rigid body. You will notice that the object moves two frames after you apply the velocity.
**Minimal reproduction project:**
Create a 3d scene and add rigid body with this script.
```gdscript
extends RigidBody
var frame = 0
func _ready():
pass # Replace with function body.
func _physics_process(delta):
match frame:
10:
linear_velocity = Vector3(1,0,0)
12:
linear_velocity = Vector3(0,0,0)
force_update_transform()
print("frame: " + str(frame) + " vel " + str(linear_velocity) +" " + str(transform.origin))
frame += 1
```
The output is
```
frame: 9 vel (0, 0, 0) (0, 0, 0)
frame: 10 vel (1, 0, 0) (0, 0, 0)
frame: 11 vel (1, 0, 0) (0, 0, 0)
frame: 12 vel (0, 0, 0) (0.1, 0, 0)
frame: 13 vel (0, 0, 0) (0.2, 0, 0)
frame: 14 vel (0, 0, 0) (0.2, 0, 0)
```
Expected would be that body has moved at frame 11.
** Note **
Using godot physics is the update is only one frame late. Hence bullet physics works differently.
Create equivalent script for 2d. You will notice that the body moves on frame 11. Hence, 2d physics works correctly.
|
bug,confirmed,topic:physics,topic:3d
|
low
|
Major
|
596,794,679 |
material-ui
|
[RFC] Chip markup
|
Regarding https://material-ui.com/components/chips/#chip-2
## Problem
Chips currently have both a bad a11y and UX story:
- delete action is inaccessible on iOS devices (and requires knowledge of keyboard shortcuts in any other screen reader)
- deleteable Chips are announced as buttons
- clicking the delete icon triggers the same ripple as clicking the Chip itself. The ripple shouldn't bubble to elements that are not activated.
- a deleteable button is by default selectable but not a basic Chip. Why?
- demos illustrates a trailing icon that is not used as the delete action. Having it trigger onDelete is misleading (adressed in #16097)
## Current implementation
This uses terminology from the accessibility tree. A `<button>` does not necessarily represent a `HTMLButtonElement`
- Basic: `<generic tabindex="-1">Basic</generic>`
- clickable: `<button>Clickable</button>`
- deleteable: `<button>Deleteable <SVGRoot aria-hidden="true" focusable="false">...</SVGRoot></button>`
- deletable + clickable: `<button tabindex="0">Clickable deleteable <SVGRoot aria-hidden="true" focusable="false">...</SVGRoot></button>`
## Proposal
- don't nest interactive elements
- whether the delete action should be in tab order is a hard question. I lean towards having it in tab order and making it easily customizale (for grid-like implementations or Autocomplete)
### Basic
```diff
-<generic tabindex="-1">Basic</generic>
+<generic>Basic</generic>
```
### Clickable
These are fine.
### deleteable
```diff
-<button>Deleteable <SVGRoot aria-hidden="true" focusable="false">...</SVGRoot></button>
+<generic>Deleteable<button><SVGRoot>...</SVGRoot></button></generic>
```
### deletable + clickable
```diff
-<button tabindex="0">Clickable deleteable <SVGRoot aria-hidden="true" focusable="false">...</SVGRoot></button>
+<generic><button tabindex="0">Clickable deleteable</button><button tabindex="0"><SVGRoot aria-hidden="true" focusable="false">...</SVGRoot></button></generic>
```
## Context
#17708
Closes #19468
|
accessibility,discussion,component: chip,RFC
|
medium
|
Major
|
596,815,605 |
flutter
|
Assert that timeline events are sorted in the Flutter driver
|
Observatory should sort the timeline events before we decide to assert
|
tool,t: flutter driver,c: proposal,P3,team-tool,triaged-tool
|
low
|
Minor
|
596,835,464 |
go
|
runtime/pprof: TestCPUProfileLabel is flaky on aix-ppc64
|
[2020-04-08T18:37:38-d5e1b7c/aix-ppc64](https://build.golang.org/log/1b609fea568da538d9a0f53ea44c16b37dde01b8)
[2020-03-17T01:24:51-7ec4adb/aix-ppc64](https://build.golang.org/log/67294431057b77bc013c73313025b5462c29398e)
[2019-12-20T20:12:18-b234fdb/aix-ppc64](https://build.golang.org/log/1ad954ec72490dbd3d8e36a82da5453cf4ae5cd4)
[2019-11-19T19:21:57-e0306c1/aix-ppc64](https://build.golang.org/log/533ca5f7f8ff334f011d8c8822d324b4ac2453d6)
[2019-09-26T17:37:02-a37f2b4/aix-ppc64](https://build.golang.org/log/c2b1566354f26d77038d3cddf3ba182b1dc1b577)
Not clear to me whether this is a builder issue, an AIX platform issue, or a timing-sensitive test.
CC @hyangah for `pprof`, @trex58 @Helflym for `aix`, @andybons @toothrot @cagedmantis @dmitshur for builders.
|
Builders,NeedsInvestigation,OS-AIX,compiler/runtime
|
low
|
Minor
|
596,877,103 |
flutter
|
[web] Write e2e comprehensive test for pointers
|
Given the relatively high number of pointer-related regressions we've had in the past, it's important to have e2e tests for it. Some regressions were desktop-only (e.g. [right-click](https://github.com/flutter/engine/pull/15103)), some were mobile-only (e.g. [semantics placeholder](https://github.com/flutter/engine/pull/17595)), and some affected both (e.g. [pointer data sanitization](https://github.com/flutter/engine/pull/14082)).
#### Requirements
* The test has to send real browser events, not framework events.
#### Scenarios
* Clicking/tapping
* Hovering
* Dragging
* Right-clicking (context menu)
* Scrolling (wheel events)
The unit tests in `pointer_binding_test.dart` have a lot of edge cases that have regressed before.
|
a: tests,team,framework,platform-web,P2,team-web,triaged-web
|
low
|
Minor
|
596,884,605 |
TypeScript
|
the type of super() should be instance, not void
|
**TypeScript Version:** 3.8.3
**Search Terms:**
**Code**
```ts
class Sub extends Super {
constructor () { return super(); }
}
```
**Expected behavior:** no error
**Actual behavior:** `TS2322`
**Playground Link:**
**Related Issues:**
|
Needs Investigation,Has Repro
|
low
|
Critical
|
596,922,234 |
angular
|
Ivy: :leave animations trigger when navigating away from component
|
# 🐞 Bug report
### Affected Package
@angular/animations (presumably)
### Is this a regression?
No, happens in all versions of 9.x with Ivy enabled.
### Description
When an animation query is specified with `:leave`, it should only fire when an item is added/removed per the documentation. In 9.x with Ivy enabled, however, all `:leave` animations fire whenever the component is navigated away from, even though nothing is being added/removed. This is especially annoying because the new component won't load until the animations have finished.
You can disable Ivy in the repro and see that it does what you would expect--`:leave` doesn't fire when navigating away from the component.
## 🔬 Minimal Reproduction
https://stackblitz.com/edit/angular-udwujp
## 🔥 Exception or Error
N/A
## 🌍 Your Environment
**Angular Version:**
<pre><code>Angular CLI: 9.1.1
Node: 13.12.0
OS: win32 x64
Angular: 9.1.1
... animations, cli, common, compiler, compiler-cli, core, forms
... language-service, platform-browser, platform-browser-dynamic
... router
Ivy Workspace: Yes
Package Version
-----------------------------------------------------------
@angular-devkit/architect 0.901.1
@angular-devkit/build-angular 0.901.1
@angular-devkit/build-optimizer 0.901.1
@angular-devkit/build-webpack 0.901.1
@angular-devkit/core 9.1.1
@angular-devkit/schematics 9.1.1
@ngtools/webpack 9.1.1
@schematics/angular 9.1.1
@schematics/update 0.901.1
rxjs 6.5.5
typescript 3.8.3
webpack 4.42.0
</code></pre>
**Anything else relevant?**
N/A
|
area: animations,state: confirmed,state: needs more investigation,P3
|
low
|
Critical
|
596,998,674 |
opencv
|
I tried to write json data larger than 10k to FileStorage, but the write failed.why? help?
|
##### System information (version)
- OpenCV => 4.2
- Operating System / Platform => Windows 64 Bit
- Compiler => Visual Studio 2017
##### Detailed description
I tried to write json data larger than 10k to FileStorage, but the write failed.why? help?
##### Steps to reproduce
I tried to write json data larger than 10k to FileStorage, but the write failed.
```.cpp
bool writeCacheJson(const std::string &filePath, const std::string &key, std::string &content) {
if(content.empty() || key.empty()){
return false;
}
std::string file = filePath + "/facepp_landmark.json";
try {
cv::FileStorage fs(file, cv::FileStorage::WRITE | cv::FileStorage::MEMORY);
if (!fs.isOpened()) {
std::cerr << "cannot open " << file << " for write" << std::endl;
return false;
}
std::string json = std::string(content);
fs.write(key, json);
fs.release(); // 关闭文件，释放内存
} catch (...) {
std::cerr << "cannot open facepp_landmark.json for write" << std::endl;
return false;
}
return true;
}
```
error:OpenCV(3.4.6) /Volumes/build-storage/build/3_4_iOS-mac/opencv/modules/core/src/persistence_json.cpp:837: error: (-5:Bad argument) The written string is too long in function 'icvJSONWriteString'
Is there any other solution？
|
feature,priority: low,category: core
|
low
|
Critical
|
597,012,139 |
vscode
|
Create a data science "working group"
|
Refs: #93604
There was a request from the March Extension Authors call to create a sort of working group of data science extension authors (Julia, R, Python, etc) to share and collaborate on cross-cutting requirements and for VS Code to learn in be informed from the baseline infrastructure required for VS Code to support innovation in the data science space.
/cc @davidanthoff
|
question-discussion
|
low
|
Major
|
597,070,575 |
rust
|
[mir-opt] New optimization: eliminate concrete type -> trait object conversions
|
Basically I am suggesting to turn
```rust
let x = SomeStruct;
let y: &dyn SomeTrait = &x;
y.some_method();
```
into
```rust
let x = SomeStruct;
let y: &SomeStruct = &x;
x.some_method()
```
While the above code is likely never written, you can easily get this after inlining some function that e.g. returns a `Box<dyn Trait>` where the type that was turned into a trait object is always the same. I case the function is a trait method, we can't change the function into one returning `impl Trait`. This optimization would do the transformation for us.
cc @rust-lang/wg-mir-opt
|
T-compiler,C-feature-request,A-coercions,A-mir-opt,A-trait-objects
|
low
|
Minor
|
597,121,351 |
opencv
|
[OpenCL] Support for INTER_CUBIC resize
|
I tested INTER_CUBIC resize on my PC.
- OpenCV 3.4.5
- Win10 , Intel UHD 620
- Visual Studio 2017
With 720P input and 4K output, I got the report below.
- 12.09ms in case IPP is enabled
- 26.55ms in case IPP is disabled
Then I tried to use UMAT to speed up by OpenCL, but nothing happened.
I found the conditional statement in function `ocl_resize` which means INTER_CUBIC is not supported.
```
if( !(cn <= 4 &&
(interpolation == INTER_NEAREST || interpolation == INTER_LINEAR ||
(interpolation == INTER_AREA && inv_fx >= 1 && inv_fy >= 1) )) )
return false;
```
Is there a plan to add INTER_CUBIC to OpenCL function `resize.cl`?
|
optimization,feature,category: imgproc,category: ocl
|
low
|
Minor
|
597,131,484 |
material-ui
|
[Checkbox] a11y - Indeterminate checkbox state announces only as ticked or unticked rather than mixed
|
<!-- Provide a general summary of the issue in the Title above -->
<!--
Thank you very much for contributing to Material-UI by creating an issue! ❤️
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 😯
<!-- Describe what happens instead of the expected behavior. -->
When I use indeterminate=true on MUI Checkbox component I would expect a screenreader to announce its state as `mixed` rather than ticked or unticked as per aria guidelines.
https://www.w3.org/TR/wai-aria-practices-1.1/examples/checkbox/checkbox-2/checkbox-2.html
## Expected Behavior 🤔
<!-- Describe what should happen. -->
I would expect an indeterminate checkbox to be announced as `mixed` instead of ticked or unticked, as per the example on https://www.w3.org/TR/wai-aria-practices-1.1/examples/checkbox/checkbox-2/checkbox-2.html
## Steps to Reproduce 🕹
<!--
Provide a link to a live example (you can use codesandbox.io) and an unambiguous set of steps to reproduce this bug.
Include code to reproduce, if relevant (which it most likely is).
You should use the official codesandbox template as a starting point: https://material-ui.com/r/issue-template
If you have an issue concerning TypeScript please start from this TypeScript playground: https://material-ui.com/r/ts-issue-template
Issues without some form of live example have a longer response time.
-->
Steps:
1. Turn on VoiceOver on MacOS.
2. Navigate to https://material-ui.com/components/checkboxes/#basic-checkboxes in Chrome.
3. Tab navigate to the indeterminate checkbox and use the space key to tick/untick the box.
4. Will always be announced as ticked or unticked.
## Context 🔦
<!--
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 🌎
<!--
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 |
| ----------- | ------- |
| Mac OS | Catalina 10.15.4 |
| Chrome | Version 80.0.3987.163 (Official Build) (64-bit) |
|
accessibility,component: checkbox
|
medium
|
Critical
|
597,192,379 |
TypeScript
|
ThisType intellisense does not agree with TypeScript code validator
|
Think I've found a bug with ThisType and/or VS Code. When I use ThisType in a valid context it works fine as expected, but when I change that context to include an override on one of the methods, it seems the intellisense doesn't pick up the override and uses ThisType as the type of this within this method. However, the validator validates the type of this as the overridden one and erros when you use a property from the intellisense. Really hard to explain but easy to see in the playground link below.
**TypeScript Version:** 3.8.3
**Search Terms:** ThisType not working
**Code**
```ts
interface CustomThisType {
bar: string;
}
interface CustomThisType2 {
baz: boolean;
}
function create<T extends {}>(definition: T & ThisType<CustomThisType> & { myCustomFunction(this: CustomThisType2, ...args: any[]): any }): T {
return null as any;
}
create({
myTestFunction() {
this.bar; // works fine
},
myCustomFunction() {
this.baz; // intellisense does not show "baz" but it does not error when used
this.bar; // intellisense shows "bar" but errors when used
}
});
```
**Expected behavior:**
I would expect the intellisense to show baz and not bar.
**Actual behavior:**
Intellisense shows bar but then creates an error when you use it.
**Playground Link:**
https://www.typescriptlang.org/play/#code/JYOwLgpgTgZghgYwgAgMIFcDOYD2BbAFQAthMCBPABxQG8AoZZAIzigC5lspQBzAbjoBfOnVCRYiFBmz5ipCtQBMyeoxYAvDkxw4ANhDggBwujHQgEYYDhDIEUA5AA8BZBAAekEABNMKwQB8ABTeEDCgwFY2HK4AZMhyZFQQTtK4hCRJ1AHI8TTIeORp+ABi5pbWIEFgmRzFGfLJigA0yAB0Haw8mByG5ADaALoAlL0g5MiCowkqDMgOYOhQtiDourrIcH59xiL2jhBBqgXkBBDYZRZRVcOzjIw1pG0sUHzIAPTvyADuOFAA1n5wiAIHNBM0RIxCvVLhUbEFbscHplnnB1G9PsgxBB1qQICBMChvDhzsgQDgwJwiDhvsgAEQaOnMdCUyLIYmk8mU6BQP4-Ij45BYCDeObIp4vDFfbG4wkElCYanfPwM1hMpgstxQXlQPzfAW2YWixjCKZ8IA
**Related Issues:**
None that I can see
|
Bug,Domain: Completion Lists
|
low
|
Critical
|
597,205,477 |
go
|
runtime: cgo calls are way faster when enabling CPU profile
|
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.14.2 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
It was present with 1.13.4 and uprading to latest release confirms it
### 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/j-lallement/.cache/go-build"
GOENV="/home/j-lallement/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOINSECURE=""
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/home/j-lallement/.go/"
GOPRIVATE=""
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="/home/j-lallement/Ubuntu/Dev/zfs/zsys/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-build141739689=/tmp/go-build -gno-record-gcc-switches"
</pre></details>
### What did you do?
1. We started having a contention on some of our code (https://github.com/ubuntu/zsys) which is using go-libzfs when opening the ZFS pools.
This issues a lot of cgo calls to a library which then proceed to syscall. We expected some performance issues.
2. The following code:
```
dsZFS, err := newZ.libzfs.DatasetOpenAll()
if err != nil {
return fmt.Errorf(i18n.G("can't list datasets: %v"), err)
}
```
Was running for 13.5s in average in our system (with a lot of datasets)
3. We thus wanted to start doing CPU profile and did:
```
f, _ := os.Create("/tmp/cpuprofile")
defer f.Close()
pprof.StartCPUProfile(f)
dsZFS, err := newZ.libzfs.DatasetOpenAll()
if err != nil {
return fmt.Errorf(i18n.G("can't list datasets: %v"), err)
}
pprof.StopCPUProfile()
```
The surprise was that the whole call was then taking less than a second! This is fully reproducible on our machine: No change apart from enabling or disabling CPU profile goes from 13.5s without CPU profile to one second with it enabled.
[cpu.pprof.txt](https://github.com/golang/go/files/4455704/cpu.pprof.txt)
The result of DatasetOpenAll isn’t identical with and without CPU profile enabled, there are less objects returns (see the difference in call stack of number of `openchildren`)
4. We started to trace without and with CPU profile enabled. This confirms the time difference:
a. Without CPU Profile enabled:
```
t, _ := os.Create("/tmp/trace")
defer t.Close()
//f, _ := os.Create("/tmp/cpuprofile")
//defer f.Close()
//pprof.StartCPUProfile(f)
trace.Start(t)
dsZFS, err := newZ.libzfs.DatasetOpenAll()
if err != nil {
return fmt.Errorf(i18n.G("can't list datasets: %v"), err)
}
trace.Stop()
//pprof.StopCPUProfile()
```
Here is the trace (taking 15s): [trace_without_cpuprofile.txt](https://github.com/golang/go/files/4455742/trace_without_cpuprofile.txt)
b. With CPU Profile enabled:
```
t, _ := os.Create("/tmp/trace")
defer t.Close()
f, _ := os.Create("/tmp/cpuprofile")
defer f.Close()
pprof.StartCPUProfile(f)
trace.Start(t)
dsZFS, err := newZ.libzfs.DatasetOpenAll()
if err != nil {
return fmt.Errorf(i18n.G("can't list datasets: %v"), err)
}
trace.Stop()
pprof.StopCPUProfile()
```
Here is the trace (taking 3s): [trace_with_cpuprofile.txt](https://github.com/golang/go/files/4455739/trace_with_cpuprofile.txt)
There is a little bit more on the 15s trace of GC calls, but nothing drastic that can explain the difference (38ms vs 4.4ms). We can see there is a lot more gaps though in Execution time.
The other surprise is that we are blocking way more in syscalls (2.3s vs 0.4s).
https://github.com/bicomsystems/go-libzfs doesn’t seem to do anything particular in case it detects it’s being under profile, is there anything CPU profiler is doing to influence the Cgo calls or am I missing anything obvious?
|
Performance,NeedsInvestigation,compiler/runtime
|
low
|
Critical
|
597,206,777 |
pytorch
|
[docs] Matrix transpose missing in formula in nn.functional.bilinear docs
|
It should say `x_2^T` if I'm not mistaken: https://github.com/pytorch/pytorch/commit/883628cb5c31c89323cef52e96310c92abecfcbe#r38390009
|
module: docs,triaged
|
low
|
Minor
|
597,207,988 |
node
|
Node 13 on MacOS cannot dyld to /usr/local/lib
|
* **Version**: v13.12.0
* **Platform**: MacOS
* **Subsystem**: Native modules
I'm running into an issue where node 13 cannot dyld to a binary in `/usr/local/lib`.
Context: I'm maintaining the nodejs foundationdb bindings. The way the native module works is nodejs loads a local napi module, which in turn looks for a dynamic library (`libfdb_c.dylib` on macos). This is normally added by the [fdb installer](https://www.foundationdb.org/download/) into `/usr/local/lib`.
This works fine under node 8 through to node 12, but under node 13 dyld can't find the dynamic library when it loads the native module.
Weirdly it works in my CI environment in [travisci](https://travis-ci.org/github/josephg/node-foundationdb/jobs/672768998), which confuses me greatly.
I can't find the code in nodejs which manages DYLD search paths (or where / how that might have changed in node 13). I can probably manually bisect node 13 versions using nvm if need be, but I'm hoping some other people have some ideas of things to try.
And it starts working again if I manually add `DYLD_LIBRARY_PATH=/usr/local/lib` to my environment. But I'd really rather avoid requiring that my users do that too.
```
$ node
Welcome to Node.js v13.12.0.
Type ".help" for more information.
> require('foundationdb')
Could not load native module. Make sure the foundationdb client is installed and
(on windows) in your PATH. https://www.foundationdb.org/download/
Uncaught:
Error: dlopen(/Users/josephg/temp/node_modules/foundationdb/prebuilds/darwin-x64/node.napi.node, 1): Library not loaded: libfdb_c.dylib
Referenced from: /Users/josephg/temp/node_modules/foundationdb/prebuilds/darwin-x64/node.napi.node
Reason: image not found
at Object.Module._extensions..node (internal/modules/cjs/loader.js:1197:18)
at Module.load (internal/modules/cjs/loader.js:996:32)
at Function.Module._load (internal/modules/cjs/loader.js:896:14)
at Module.require (internal/modules/cjs/loader.js:1036:19)
at require (internal/modules/cjs/helpers.js:72:18)
at load (/Users/josephg/temp/node_modules/node-gyp-build/index.js:21:10)
```
|
macos,addons
|
low
|
Critical
|
597,221,340 |
pytorch
|
torch.ceil, torch.floor should accept a dtype argument
|
## 🚀 Feature
When a user is doing `torch.ceil` and `torch.floor`, it is likely that the user is asking for a long tensor instead of a floating tensor
`torch.ceil(x, dtype=torch.long)` would make this possible.
cc @mruberry @rgommers
|
triaged,module: numpy,module: ux
|
low
|
Minor
|
597,243,544 |
rust
|
ICEs + RUST_BACKTRACE=1 should interleave the backtrace and query stack.
|
The interleaving can be done by looking for frames with a certain function path (that we need to make sure show up w/o debuginfo, so that this is useful on nightlies).
Downside: harder to see the aggregate query stack in one place (unless we duplicate it).
Upside: easier to find where a query was invoked from.
cc @Zoxc
|
C-enhancement,A-diagnostics,T-compiler,A-query-system,D-diagnostic-infra,A-backtrace
|
low
|
Critical
|
597,272,954 |
create-react-app
|
PUBLIC_URL is not the same as homepage, but only the pathname
|
<!--
Please note that your issue will be fixed much faster if you spend about
half an hour preparing it, including the exact reproduction steps and a demo.
If you're in a hurry or don't feel confident, it's fine to report bugs with
less details, but this makes it less likely they'll get fixed soon.
In either case, please use this template and fill in as many fields below as you can.
Note that we don't provide help for webpack questions after ejecting.
You can find webpack docs at https://webpack.js.org/.
-->
### Describe the bug
I have uploaded all the static resources to CDN, include .js and .css files. So I have to set the 'homepage' as the host address of CDN server, which make the urls of scripts in 'index.html' to be right.
However, the urls injected to 'index.html' is the pathname of 'homepage' that I set in package.json, after upload react-scripts to 3.4.1.
And there is no such problem when I using version 3.3.0. I found that the actual homepage has been change after this PR [feat(react-scripts): allow PUBLIC_URL in develoment mode](https://github.com/facebook/create-react-app/pull/7259)
So I want to know, is it a bug that will be fixed? Or is there any method that can make the PUBLIC_URL to be a complete url.
### Did you try recovering your dependencies?
<!--
Your module tree might be corrupted, and that might be causing the issues.
Let's try to recover it. First, delete these files and folders in your project:
* node_modules
* package-lock.json
* yarn.lock
Then you need to decide which package manager you prefer to use.
We support both npm (https://npmjs.com) and yarn (http://yarnpkg.com/).
However, **they can't be used together in one project** so you need to pick one.
If you decided to use npm, run this in your project directory:
npm install -g npm@latest
npm install
This should fix your project.
If you decided to use yarn, update it first (https://yarnpkg.com/en/docs/install).
Then run in your project directory:
yarn
This should fix your project.
Importantly, **if you decided to use yarn, you should never run `npm install` in the project**.
For example, yarn users should run `yarn add <library>` instead of `npm install <library>`.
Otherwise your project will break again.
Have you done all these steps and still see the issue?
Please paste the output of `npm --version` and/or `yarn --version` to confirm.
-->
no, but downgrade to 3.3.0
### Which terms did you search for in User Guide?
<!--
There are a few common documented problems, such as watcher not detecting changes, or build failing.
They are described in the Troubleshooting section of the User Guide:
https://facebook.github.io/create-react-app/docs/troubleshooting
Please scan these few sections for common problems.
Additionally, you can search the User Guide itself for something you're having issues with:
https://facebook.github.io/create-react-app/
If you didn't find the solution, please share which words you searched for.
This helps us improve documentation for future readers who might encounter the same problem.
-->
The docs still use the complete url as example [Step 1: Add homepage to package.json](https://create-react-app.dev/docs/deployment#step-1-add-homepage-to-packagejson)
### Environment
<!--
To help identify if a problem is specific to a platform, browser, or module version, information about your environment is required.
This enables the maintainers quickly reproduce the issue and give feedback.
Run the following command in your React app's folder in terminal.
Note: The result is copied to your clipboard directly.
`npx create-react-app --info`
Paste the output of the command in the section below.
-->
irrelevant, i guess
### Steps to reproduce
<!--
How would you describe your issue to someone who doesn’t know you or your project?
Try to write a sequence of steps that anybody can repeat to see the issue.
-->
(Write your steps here:)
1. set homepage in package.json like 'https://example.com/a/b/c'
2. yarn build
3. the url of script in index.html will be '/a/b/c/xxx.js', which will not find the correct resource
### Expected behavior
<!--
How did you expect the tool to behave?
It’s fine if you’re not sure your understanding is correct.
Just write down what you thought would happen.
-->
The PUBLIC_URL should be the same as homepage in package.json
### Actual behavior
<!--
Did something go wrong?
Is something broken, or not behaving as you expected?
Please attach screenshots if possible! They are extremely helpful for diagnosing issues.
-->
The PUBLIC_URL is the pathname of homepage
### Reproducible demo
<!--
If you can, please share a project that reproduces the issue.
This is the single most effective way to get an issue fixed soon.
There are two ways to do it:
* Create a new app and try to reproduce the issue in it.
This is useful if you roughly know where the problem is, or can’t share the real code.
* Or, copy your app and remove things until you’re left with the minimal reproducible demo.
This is useful for finding the root cause. You may then optionally create a new project.
This is a good guide to creating bug demos: https://stackoverflow.com/help/mcve
Once you’re done, push the project to GitHub and paste the link to it below:
-->
no demo
<!--
What happens if you skip this step?
We will try to help you, but in many cases it is impossible because crucial
information is missing. In that case we'll tag an issue as having a low priority,
and eventually close it if there is no clear direction.
We still appreciate the report though, as eventually somebody else might
create a reproducible example for it.
Thanks for helping us help you!
-->
|
issue: bug report
|
medium
|
Critical
|
597,276,327 |
go
|
x/tools/cmd/fiximports: don't use .biz extension in testdata
|
<!--
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 go1.14.2 windows/amd64
</pre>
### Does this issue reproduce with the latest release?
### What operating system and processor architecture are you using (`go env`)?
<details><summary><code>go env</code> Output</summary><br><pre>
set GO111MODULE=on
set GOARCH=amd64
set GOBIN=
set GOCACHE=C:\Users\bert.hooyman\AppData\Local\go-build
set GOENV=C:\Users\bert.hooyman\AppData\Roaming\go\env
set GOEXE=.exe
set GOFLAGS=
set GOHOSTARCH=amd64
set GOHOSTOS=windows
set GOINSECURE=
set GONOPROXY=
set GONOSUMDB=
set GOOS=windows
set GOPATH=C:\Users\bert.hooyman\go
set GOPRIVATE=
set GOPROXY=https://proxy.golang.org,direct
set GOROOT=c:\go
set GOSUMDB=sum.golang.org
set GOTMPDIR=
set GOTOOLDIR=c:\go\pkg\tool\windows_amd64
set GCCGO=gccgo
set AR=ar
set CC=gcc
set CXX=g++
set CGO_ENABLED=1
set GOMOD=NUL
set CGO_CFLAGS=-g -O2
set CGO_CPPFLAGS=
set CGO_CXXFLAGS=-g -O2
set CGO_FFLAGS=-g -O2
set CGO_LDFLAGS=-g -O2
set PKG_CONFIG=pkg-config
set GOGCCFLAGS=-m64 -mthreads -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=C:\Users\BERT~1.HOO\AppData\Local\Temp\go-build874685950=/tmp/go-build -gno-record-gcc-switches
</pre></details>
### What did you do?
<!--
If possible, provide a recipe for reproducing the error.
A complete runnable program is good.
A link on play.golang.org is best.
-->
go get -v golang.org/x/tools/gopls
### What did you expect to see?
successful tool install
### What did you see instead?
go: downloading golang.org/x/tools v0.0.0-20200408132156-9ee5ef7a2c0d
-> unzip C:\Users\bert.hooyman\go\pkg\mod\cache\download\golang.org\x\tools\@v\v0.0.0-20200408132156-9ee5ef7a2c0d.zip: mkdir C:\Users\bert.hooyman\go\pkg\mod\golang.org\x\[email protected]\cmd\fiximports\testdata\src\titanic.biz: Access is denied.
go: golang.org/x/tools/gopls upgrade => v0.3.4
-> unzip C:\Users\bert.hooyman\go\pkg\mod\cache\download\golang.org\x\tools\@v\v0.0.0-20200316194252-fafb6e2e8a4a.zip: mkdir C:\Users\bert.hooyman\go\pkg\mod\golang.org\x\[email protected]\cmd\fiximports\testdata\src\titanic.biz: Access is denied.
go get golang.org/x/tools/gopls: unzip C:\Users\bert.hooyman\go\pkg\mod\cache\download\golang.org\x\tools\@v\v0.0.0-20200316194252-fafb6e2e8a4a.zip: mkdir C:\Users\bert.hooyman\go\pkg\mod\golang.org\x\[email protected]\cmd\fiximports\testdata\src\titanic.biz: Access is denied.
The problem I am seeing is essentially: mkdir titanic.biz
.biz is a ransomware extension Symantec will block it.
My system has Symantec Enpoint Protection, and my employer won't let me switch it off.
.biz is a bad choice for a testdata folder name suffix.
|
NeedsInvestigation,Tools
|
low
|
Critical
|
597,298,681 |
storybook
|
Source-loader: Breaks when stories reference other stories before they are defined
|
**Describe the bug**
waring: @storybook/source-loader was applied to a file which does not contain a story. Please check your webpack configuration and make sure to apply @storybook/source-loader only to files containg stories. Related file:"
**To Reproduce**
Steps to reproduce the behavior:
1. run is [repo](https://github.com/0x0006e/StoryBook/tree/bug/10363)
**Expected behavior**
no waring
**Screenshots**
**Code snippets**
```js
export function AllButton() {
return (
<>
<DefaultButton />
<PrimaryButton />
<SuccessButton />
</>
);
}
export function DefaultButton() {
return (
<Button
onClick={action("clicked")}
text={text("Default Button", "Default Button")}
/>
);
}
export function PrimaryButton() {
return (
<Button
onClick={action("clicked")}
text={text("Primary Button", "Primary Button")}
type="primary"
/>
);
}
export function SuccessButton() {
return (
<Button
onClick={action("clicked")}
text={text("Success Button", "Success Button")}
type="success"
/>
);
}
```
**System:**
```bash
Environment Info:
System:
OS: macOS 10.15.4
CPU: (12) x64 Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz
Binaries:
Node: 13.12.0 - ~/.nvm/versions/node/v13.12.0/bin/node
Yarn: 1.22.0 - /usr/local/bin/yarn
npm: 6.14.4 - ~/.nvm/versions/node/v13.12.0/bin/npm
Browsers:
Chrome: 80.0.3987.163
Safari: 13.1
npmPackages:
@storybook/addon-actions: ^5.3.18 => 5.3.18
@storybook/addon-console: ^1.2.1 => 1.2.1
@storybook/addon-docs: ^5.3.18 => 5.3.18
@storybook/addon-knobs: ^5.3.18 => 5.3.18
@storybook/addon-links: ^5.3.18 => 5.3.18
@storybook/addon-notes: ^5.3.18 => 5.3.18
@storybook/addons: ^5.3.18 => 5.3.18
@storybook/preset-create-react-app: ^2.1.1 => 2.1.1
@storybook/react: ^5.3.18 => 5.3.18
```
**Additional context**
Add any other context about the problem here.
|
bug,has workaround,source-loader,csf
|
medium
|
Critical
|
597,300,195 |
pytorch
|
[discussion] Refactor spectral_norm to use the newly merged lowrank solvers and proposal for Linear Algebra Cookbook page
|
LOBPCG variants were merged in https://github.com/pytorch/pytorch/pull/29488
However, spectral_norm [implementation](https://github.com/pytorch/pytorch/blob/master/torch/nn/utils/spectral_norm.py#L2) contains an older power iteration implementation to get an estimate for leading eigenvalue.
I propose (also in https://github.com/pytorch/pytorch/issues/8049#issuecomment-607253487) to either:
1) surface the power iteration / implement randomized power iteration to the user - it's useful reference code (for debugging linear layer properties), but it's one another solver to maintain, although not a big amount of [code](https://github.com/facebook/fbpca/blob/master/fbpca.py#L1503)
2) replace existing power iteration by the new LOBPCG call (discussed in the end of https://github.com/pytorch/pytorch/issues/8049 and deemed quite practical by @lobpcg)
In general I think it'd be nice to have some dedicated page "Linear Algebra PyTorch Cookbook" and detail for typical linear algebra tasks (linear system solvers, eigenproblem solvers, ...) what methods are currently supported (wrt sparsity, symmetric, p.s.d, problem sizes, differentiability, space/time complexity, cpu/gpu support, batchability, what external libraries are used, is parallelization across batches used, numerical stability considerations, etc)
cc @vincentqb @vishwakftw @SsnL @jianyuh @nikitaved
|
triaged,module: linear algebra
|
low
|
Critical
|
597,306,823 |
vue
|
SSR fails to render component inside v-else of a v-if with v-html
|
### Version
2.6.11
### Reproduction link
[https://github.com/tuomassalo/vue-ssr-v-html-bug](https://github.com/tuomassalo/vue-ssr-v-html-bug)
### Steps to reproduce
- clone the repo
- run `npm run dev`
- open `localhost:8080`
- observe Console log.
### What is expected?
I expect SSR to render "bar: Bar!", as the client-side does.
Or, I'd like to get an eslint warning that this is a bad idea (if that is the problem).
### What is actually happening?
`App.vue` fails to render `bar-component` on the server. Instead, it outputs `<bar-component></bar-component>`, and the dev server gives the warning "The client-side rendered virtual DOM tree is not matching server-rendered content."
---
The key part of `App.vue` is this:
```
<div v-if="foo" v-html="'Foo.'"/>
<div v-else>
bar: <bar-component/>
</div>
```
My original component was naturally longer. I ran into this problem after changing the `v-if` line from something like:
<div v-if="foo">{{ foo }}</div>
To:
<div v-if="foo" v-html="foo"/>
... which *seemed* innocuous to me.
Finally, apologies for posting a very possible duplicate.
<!-- generated by vue-issues. DO NOT REMOVE -->
|
bug,has workaround,feat:ssr
|
low
|
Critical
|
597,309,944 |
youtube-dl
|
MotoGp site support request
|
<!--
######################################################################
WARNING!
IGNORING THE FOLLOWING TEMPLATE WILL RESULT IN ISSUE CLOSED AS INCOMPLETE
######################################################################
-->
## Checklist
<!--
Carefully read and work through this check list in order to prevent the most common mistakes and misuse of youtube-dl:
- First of, make sure you are using the latest version of youtube-dl. Run `youtube-dl --version` and ensure your version is 2020.03.24. If it's not, see https://yt-dl.org/update on how to update. Issues with outdated version will be REJECTED.
- Make sure that all provided video/audio/playlist URLs (if any) are alive and playable in a browser.
- Make sure that site you are requesting is not dedicated to copyright infringement, see https://yt-dl.org/copyright-infringement. youtube-dl does not support such sites. In order for site support request to be accepted all provided example URLs should not violate any copyrights.
- Search the bugtracker for similar site support requests: http://yt-dl.org/search-issues. DO NOT post duplicates.
- Finally, put x into all relevant boxes (like this [x])
-->
- [x] I'm reporting a new site support request
- [x] I've verified that I'm running youtube-dl version **2020.03.24**
- [x] I've checked that all provided URLs are alive and playable in a browser
- [x] I've checked that none of provided URLs violate any copyrights
- [x] I've searched the bugtracker for similar site support requests including closed ones
## Example URLs
<!--
Provide all kinds of example URLs support for which should be included. Replace following example URLs by yours.
-->
- Single video: https://www.youtube.com/watch?v=BaW_jenozKc
- Single video: https://youtu.be/BaW_jenozKc
- Playlist: https://www.youtube.com/playlist?list=PL4lCao7KL_QFVb7Iudeipvc2BCavECqzc
## Description
Could you please add support to MotoGP website.
Using this URL, returns "unsupported website" error:
https://www.motogp.com/en/video_gallery/2020/04/09/free-10-of-the-best-races-from-home-heroes/329106?n=103440
However, if I inspect element the video URL is pretty obvious there and I can easily get the link:
https://apphdod2idevo-vh.akamaihd.net/i/,2012/08/24/2007_16_AUS_MGP_RACE_IF_Mixed_forSD43_ST_clean.mp4,2012/08/24/2007_16_AUS_MGP_RACE_IF_Mixed_forSD43_HR_clean.mp4,.csmil/master.m3u8?hdnea=st=1586441215~exp=1586452015~acl=/i/,2012/08/24/2007_16_AUS_MGP_RACE_IF_Mixed_forSD43_ST_clean.mp4,2012/08/24/2007_16_AUS_MGP_RACE_IF_Mixed_forSD43_HR_clean.mp4,*~hmac=8765a8f8d917a038e3ffaab0802f5da2ca76ec3c77b75bc3456ad488158bc112
This URL download just fine by the way with youtube-dl, so it should be easy to add support for it.
Would be great if Youtube-dl could support it out of the box.
Thank you for your hard work.
|
site-support-request
|
low
|
Critical
|
597,329,068 |
go
|
cmd/link: sigbus/segfault on ARM using AzCopy
|
### What version of Go are you using (`go version`)?
`go version go1.14.2 linux/arm`
### Does this issue reproduce with the latest release?
Yes.
### What operating system and processor architecture are you using (`go env`)?
````
GO111MODULE=""
GOARCH="arm"
GOBIN=""
GOCACHE="/home/pi/.cache/go-build"
GOENV="/home/pi/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="arm"
GOHOSTOS="linux"
GOINSECURE=""
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/usr/local/go"
GOPRIVATE=""
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/usr/local/go"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/usr/local/go/pkg/tool/linux_arm"
GCCGO="gccgo"
GOARM="7"
AR="ar"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
GOMOD="/home/pi/azure-storage-azcopy-10.3.4/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 -marm -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build366671593=/tmp/go-build -gno-record-gcc-switches"
````
### What did you do?
Compiled and used azcopy. azcopy will SOMETIMES segfault. Opened an issue ( https://github.com/Azure/azure-storage-azcopy/issues/882 ) and developers suggest it is an issue with golang on ARM.
````
wget https://github.com/Azure/azure-storage-azcopy/archive/10.3.4.zip
tar zxvf 10.3.4.zip
cd azure-storage-azcopy-10.3.4/
go build
go install
````
### What did you expect to see?
Completed upload.
### What did you see instead?
````
pi@pbsorca:~ $ azure-storage-azcopy copy --recursive "$RAWFS" "$RAWFS_SAS" --log-level warning
INFO: Scanning...
Job 768e77f4-9f04-ee48-7fa4-b287ee176e20 has started
Log file is located at: /home/pi/.azcopy/768e77f4-9f04-ee48-7fa4-b287ee176e20.log
0.0 %, 0 Done, 0 Failed, 10000 Pending, 0 Skipped, 10000 Total (scanning...), unexpected fault address 0x0
fatal error: fault
[signal SIGBUS: bus error code=0x1 addr=0x0 pc=0x1296c]
goroutine 164 [running]:
runtime.throw(0x60c151, 0x5)
/usr/local/go/src/runtime/panic.go:1116 +0x5c fp=0x539b8fc sp=0x539b8e8 pc=0x460f0
runtime.sigpanic()
...
````
See attached for full trace.
[azcopy-golang-seg.txt](https://github.com/golang/go/files/4456753/azcopy-golang-seg.txt)
|
NeedsInvestigation,compiler/runtime
|
low
|
Critical
|
597,343,743 |
pytorch
|
Updating a ModelDict instance with another ModelDict instance generates error.
|
## 🐛 Bug
<!-- A clear and concise description of what the bug is. -->
When merging two torch.nn.ModuleDict instances A and B via A.update(B), the following error is generated:
```
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
<ipython-input-157-967ebf8b86d0> in <module>
10 print(type(model2))
11
---> 12 model1.update(model2)
c:\python36\envs\modern_pointcn\lib\site-packages\torch\nn\modules\container.py in update(self, modules)
348 raise ValueError("ModuleDict update sequence element "
349 "#" + str(j) + " has length " + str(len(m)) +
--> 350 "; 2 is required")
351 self[m[0]] = m[1]
352
ValueError: ModuleDict update sequence element #0 has length 5; 2 is required
```
It looks like a ModuleDict is not recognized as a mapping.
## To Reproduce
Steps to reproduce the behavior:
1. import torch
2. Create two instances of torch.nn.ModuleDict A and B.
3. Call A.update(B)
<!-- If you have a code sample, error messages, stack traces, please provide it here as well -->
Code to reproduce:
```
model1 = torch.nn.ModuleDict(OrderedDict([
('conv1', torch.nn.Conv2d(1,20,5)),
('relu1', torch.nn.ReLU())]))
model2 = torch.nn.ModuleDict(OrderedDict([
('conv1', torch.nn.Conv2d(1,20,5)),
('relu1', torch.nn.ReLU())]))
model1.update(model2)
```
## Expected behavior
<!-- A clear and concise description of what you expected to happen. -->
A ModuleDict can be appended to another ModuleDict using the update() function.
## Environment
PyTorch version: 1.4.0
Is debug build: No
CUDA used to build PyTorch: 10.1
OS: Microsoft Windows 10 Enterprise
GCC version: Could not collect
CMake version: version 3.15.3
Python version: 3.6
Is CUDA available: Yes
CUDA runtime version: 9.0.176
GPU models and configuration: GPU 0: GeForce GTX 980
Nvidia driver version: 432.00
cuDNN version: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\bin\cudnn64_7.dll
Versions of relevant libraries:
[pip3] numpy==1.18.2
[pip3] torch==1.4.0
[pip3] torchvision==0.5.0
## Additional context
<!-- Add any other context about the problem here. -->
|
module: nn,triaged
|
low
|
Critical
|
597,351,056 |
go
|
encoding/xml: memoize names during decode
|
Counterpart of #32779 for `encoding/xml`, but tracking it as a performance improvement instead of a proposal.
When an xml is decoded, many of the same element and attribute names appear over and over in the XML and get copied and reconstructed in the result over and over as well. This was previously partially discussed in [#21823](https://github.com/golang/go/issues/21823#issuecomment-376000825), where @nussjustin and @a-h proposed some good initial implementation:
```go
// Get name: /first(first|second)*/
// Do not set d.err if the name is missing (unless unexpected EOF is received):
// let the caller provide better context.
func (d *Decoder) name() (s string, ok bool) {
d.buf.Reset()
if !d.readName() {
return "", false
}
// Now we check the characters.
b := d.buf.Bytes()
if s, ok = d.names[string(b)]; ok {
return s, ok
}
if !isName(b) {
d.err = d.syntaxError("invalid XML name: " + string(b))
return "", false
}
s = string(b)
d.names[s] = s
return s, true
}
```
|
Performance,NeedsInvestigation
|
low
|
Critical
|
597,367,517 |
go
|
runtime: timeout in TestCgoCCodeSIGPROF and TestCgoExecSignalMask on linux-ppc64le-buildlet
|
[2020-04-09T01:16:43-376472d/linux-ppc64le-buildlet](https://build.golang.org/log/9a3e965f29fdf5f0cb91232628959d4edc68574c)
CC @laboger @ceseo for `linux-ppc64le`, @aclements @ianlancetaylor for `runtime_test.TestCgo*`, @andybons for possible builder slowness
|
Builders,NeedsInvestigation,arch-ppc64x,compiler/runtime
|
low
|
Major
|
597,368,965 |
excalidraw
|
Enhance arbitrary shapes with background
|
The new feature to fill arbitrary shapes with a background as discussed in #1262 and implemented by #1315 is amazing. That being said it would be even more useful to me if it flexibility was added.
While I can draw something like this: (a showcased in the tweet)
It is impossible for me to draw a real star which I would expect to look like this:
This would require hiding parts of the line inside of the shape and filling the full area. I actually needed this when trying to draw a heart. All shapes that have an inward crinkle ar otherwise impossible to fill as the lines always curve.
You thus have to cross into your shape to achieve the shape you are aiming for.
|
enhancement
|
low
|
Major
|
597,391,773 |
PowerToys
|
KBM Editor migrate to XAML / rejuv from code behind UX
|
# Summary of the enhancement
<!--
A clear and concise description of what the problem is that the new feature would solve.
Describe why and how a user would use this new functionality (if applicable).
-->
Keyboard Manager's UI is currently completely implemented in the code behind without using any XAML files. This should be migrated to XAML for WinUI 3.
|
Help Wanted,Product-Keyboard Shortcut Manager,Area-Quality,Area-User Interface,Cost-Large,UI refresh
|
low
|
Major
|
597,394,971 |
flutter
|
Allow specifying the target platform for flutter run on Android
|
This would allow an application to be installed and run as a 32bit arm even if the phone hardware supports 64bit. This can be important for app size, or if the app depends on other native libraries that are only distributed as 32bit.
An example PR that started this work is: https://github.com/flutter/flutter/pull/51631
|
c: new feature,tool,P3,team-tool,triaged-tool
|
low
|
Minor
|
597,411,296 |
TypeScript
|
`typeof` expressions should return an equivalent of a guard
|
<!-- 🚨 STOP 🚨 𝗦𝗧𝗢𝗣 🚨 𝑺𝑻𝑶𝑷 🚨
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
typeof guard array filter
<!-- List of keywords you searched for before creating this issue. Write them down here so that others can find this suggestion more easily -->
## Suggestion
Currently, `typeof` expressions return a boolean as expected. However, they are also special in that they narrow the types of their operands. In a similar vein, we can use type guards as the return value of functions to narrow(or assert) the types of their parameters:
```ts
declare const broad: string | number | object;
if(typeof broad === 'number') {
// broad is now a number, not a string nor an object
}
declare function isNumber(param: any): param is number;
if(isNumber(broad) {
// broad is now a number, not a string nor an object
}
```
This is all fine, but suppose you write a very specific function:
```ts
// Current return type: boolean
// Wish it was `param is number`
const isNumber = (param: any) => typeof param === 'number'
if (isNumber(broad)) {
// param is possibly a string or object?!
}
```
This is IMO a special case that should be handled; particularly, when a `typeof` expression(or any expression that narrows a parameter's type) is the return value of a function, TS should infer that the function is a guard.
Notice that, unlike the discussion in #6015, this issue is about the **inferred** type of a function which has a return value of a narrowing-expression. We could go further:
```ts
const isObject = (o: unknown) => typeof o === 'object';
// type for isObject == (o:unknown) => o is object
const isElement = (o: object) => o instanceof Element;
// type for isElement == (o:object) => o is Element
const unknownIsElement = (o: unknown) => isObject(o) && isElement(o);
// type for unknownIsElement == (o:unknown) => o is Element
const numberIsElement = (o:unknown) => isNumber(o) && isElement(o); // error: o is number so isElement can't be called
```
(by the way, I searched but couldn't find a similar one. If this is a duplicate I apologize!)
<!-- A summary of what you'd like to see added or changed -->
## Use Cases
The most immediate use case is for array `filter`:
```ts
// error: (string|object)[] not assignable to string[]
const stringElements: string[] = elements.filter(e => typeof e === 'string');
// Works, but repetitive
const stringElements: string[] = elements.filter((e): e is string => typeof e === 'string');
```
I'm sure this type of change would make guards far more powerful as well, since we could create functions combine JS-checks(`typeof`, `instanceof`) or even TS-checks(like discriminant variable checking) without having to explicitly write the guard expression(which, IMO is as good as a type assertion at the moment).
## Examples
(see above)
## Checklist
My suggestion meets these guidelines:
* [ ] 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).
This is possibly a breaking change, but IMO a good one as it would highlight actual programming errors for code that breaks
|
Suggestion,In Discussion
|
low
|
Critical
|
597,425,802 |
go
|
encoding/xml: unable to handle utf-16-encoded file without manual manipulation of source bytes
|
<!--
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 go1.14 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/mccolljr/Library/Caches/go-build"
GOENV="/Users/mccolljr/Library/Application Support/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="darwin"
GOINSECURE=""
GONOPROXY=""
GONOSUMDB=""
GOOS="darwin"
GOPATH="/Users/mccolljr/go"
GOPRIVATE=""
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/Users/mccolljr/go/src/github.com/golang/go"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/Users/mccolljr/go/src/github.com/golang/go/pkg/tool/darwin_amd64"
GCCGO="gccgo"
AR="ar"
CC="clang"
CXX="clang++"
CGO_ENABLED="1"
GOMOD="/Users/mccolljr/go/src/github.com/orthly/3oxz/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 -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=/var/folders/4g/0y_btbcj46v3x478swzt64140000gn/T/go-build229162746=/tmp/go-build -gno-record-gcc-switches -fno-common"
</pre></details>
### What did you do?
I have a file like the following, encoded in `utf-16` (BOM is little endian) on disk:
```xml
<?xml version="1.0" encoding="utf-16"?>
<SomeValidXML></SomeValidXML>
```
When I read the file from disk, the bytes (including the line containing `<?xml version="1.0" encoding="utf-16"?>`), are encoded in utf-16.
I wanted to parse the full file, with no modification, using the `encoding/xml` package.
### What did you expect to see?
I expected to be able to either A: transform the file's bytes to `utf8` and pass that reader to `xml.NewDecoder` to successfully parse the utf8 data as xml, or B: pass the `utf16`-encoded bytes to `xml.NewDecider` and provide a `CharsetReader` to successfully parse the `utf16` data as XML.
### What did you see instead?
There were a couple of error cases.
1. When I pass the `resultOfOsOpen` directly to `xml.NewDecoder`, with or without setting `CharsetReader` to `charset.NewReaderLabel`: `XML syntax error on line 1: invalid UTF-8`
2. When I pass the `utf-8` reader returned by `charset.NewReader(resultOfOsOpen, "text/xml")` to `xml.NewDecoder`: `xml: encoding "utf-16" declared but Decoder.CharsetReader is nil`
3. When I pass the `utf-8` reader returned by `charset.NewReader(resultOfOsOpen, "text/xml")` to `xml.NewDecoder`, AND set `CharsetReader` to `charset.NewReaderLabel`: The (now `utf-8`-encoded) data is interpreted as `utf-16` and I the decoder reads the file as gibberish.
It seems to me that the `encoding/xml` package expects the line containing `<?xml version="1.0" encoding="utf-16"?>` to be in some encoding that resembles valid `utf-8`-encoded text in order to read the encoding line and properly parse the rest of the file, OR for the line to be removed if manual transformation of the input is done beforehand (like with `utf-16`, which cannot be read as valid `utf-8` text).
Am I missing something? Is there a way to do this without modifying the input bytes?
|
NeedsInvestigation
|
low
|
Critical
|
597,439,031 |
godot
|
Scrolling in RichTextLabel ignores visible_characters property (fixed in `master`)
|
**Godot version:**
3.2.1.stable.official
**OS/device including version:**
Linux
**Issue description:**
If text does not fit in height and `visible_characters` is used (for the typewriter effect), then the text is not displayed as expected.
1. `scroll_active == true and scroll_following == false`:
There is an opportunity to scroll through unnecessary empty space.
2. `scroll_active == false and scroll_following == true`:
First you see empty space, then (when visible_characters is incremented by a script) you see the end of the phrase.
Proposed solution: `RichTextLabel` should only consider visible lines when scrolling.
**Steps to reproduce:**
**Minimal reproduction project:**
[rtl_scroll_bug.zip](https://github.com/godotengine/godot/files/4457595/rtl_scroll_bug.zip)
|
bug,topic:gui
|
low
|
Critical
|
597,488,107 |
terminal
|
closeTab should present a confirmation dialog
|
# Description of the new feature/enhancement
`closeTab` should present a confirmation dialog if multiple panes are open. This would prevent users from accidentally closing all panes in a tab by accident.
|
Help Wanted,Area-UserInterface,Product-Terminal,Issue-Task,In-PR
|
low
|
Major
|
597,505,799 |
three.js
|
OutlineEffect lines not meeting at corner
|
##### Description of the problem
I've been working to implement an OutlineEffect for a Collada file and have been having problems with the lines not meeting at the corners. I modified the Toon example that you created with cubes instead of spheres and am having the same problem so it doesn't seem to be my implementation: [https://jsfiddle.net/4170Lv2z/1/](https://jsfiddle.net/4170Lv2z/1/)
And here's a screenshot of my model:
<img width="939" alt="Screen Shot 2020-04-09 at 3 25 51 PM" src="https://user-images.githubusercontent.com/17128916/78932967-75d75f80-7a76-11ea-98eb-ea90dc80b8b7.png">
Is there a quick fix for this?
##### Three.js version
- [ ] Dev
- [x] r115
- [ ] ...
##### Browser
- [x] All of them
- [ ] Chrome
- [ ] Firefox
- [ ] Internet Explorer
##### OS
- [x] All of them
- [ ] Windows
- [ ] macOS
- [ ] Linux
- [ ] Android
- [ ] iOS
##### Hardware Requirements (graphics card, VR Device, ...)
|
Needs Investigation,Post-processing
|
medium
|
Major
|
597,512,219 |
pytorch
|
Illegal memory access / CUDNN_STATUS_MAPPING_ERROR on Quadro 8000
|
## 🐛 Bug
When training the model from https://github.com/rosinality/stylegan2-pytorch on a Quadro RTX 8000, in Pytorch 1.4, cuda 10.1, cudnn 7.6.5 I get the error:
`RuntimeError: cuDNN error: CUDNN_STATUS_MAPPING_ERROR`
This only happens for very large image sizes and large batch sizes, and appears to be exactly the error from [#22496 ](https://github.com/pytorch/pytorch/issues/22496).
To fix it, I tried installing pytorch from master, currently 1.6.0a0+e311e53, along with cuda 10.2, cudnn 7.6.5. Now I get an error under the same circumstances (i.e. only for large image and batch sizes), but the error is different:
```
File "train.py", line 232, in train
path[0].backward()
File "/home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/tensor.py", line 198, in backward
torch.autograd.backward(self, gradient, retain_graph, create_graph)
File "/home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/autograd/__init__.py", line 98, in backward
Variable._execution_engine.run_backward(
RuntimeError: CUDA error: an illegal memory access was encountered (launch_unrolled_kernel at /home/zack/setup/pytorch/aten/src/ATen/native/cuda/CUDALoops.cuh:242)
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0x6c (0x7f8ffcb652cc in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libc10.so)
frame #1: void at::native::gpu_kernel_impl<__nv_hdl_wrapper_t<false, true, __nv_dl_tag<void (*)(at::TensorIterator&, bool), &at::native::copy_device_to_device, 4u>, float (float)> >(at::TensorIterator&, __nv_hdl_wrapper_t<false, true, __nv_dl_tag<void (*)(at::TensorIterator&, bool), &at::native::copy_device_to_device, 4u>, float (float)> const&) + 0x1a94 (0x7f8f182b5764 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #2: void at::native::gpu_kernel<__nv_hdl_wrapper_t<false, true, __nv_dl_tag<void (*)(at::TensorIterator&, bool), &at::native::copy_device_to_device, 4u>, float (float)> >(at::TensorIterator&, __nv_hdl_wrapper_t<false, true, __nv_dl_tag<void (*)(at::TensorIterator&, bool), &at::native::copy_device_to_device, 4u>, float (float)> const&) + 0x15b (0x7f8f182b6e4b in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #3: at::native::copy_device_to_device(at::TensorIterator&, bool) + 0x675 (0x7f8f18294eb5 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #4: <unknown function> + 0x1e4800c (0x7f8f1829800c in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #5: <unknown function> + 0x98d99b (0x7f8f29f8199b in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #6: <unknown function> + 0x98a913 (0x7f8f29f7e913 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #7: at::native::copy_(at::Tensor&, at::Tensor const&, bool) + 0x54 (0x7f8f29f80694 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #8: at::native::contiguous(at::Tensor const&, c10::MemoryFormat) + 0x4a0 (0x7f8f2a252600 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #9: <unknown function> + 0xf86e88 (0x7f8f2a57ae88 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #10: <unknown function> + 0xff1d13 (0x7f8f2a5e5d13 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #11: <unknown function> + 0x137faa3 (0x7f8f177cfaa3 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #12: <unknown function> + 0x3163957 (0x7f8f195b3957 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #13: at::native::cudnn_convolution_transpose_backward_input(at::Tensor const&, at::Tensor const&, c10::ArrayRef<long>, c10::ArrayRef<long>, c10::ArrayRef<long>, long, bool, bool) + 0xbb (0x7f8f195b3e7b in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #14: <unknown function> + 0x31d3b7e (0x7f8f19623b7e in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #15: <unknown function> + 0x322be1e (0x7f8f1967be1e in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #16: at::native::cudnn_convolution_transpose_backward(at::Tensor const&, at::Tensor const&, at::Tensor const&, c10::ArrayRef<long>, c10::ArrayRef<long>, c10::ArrayRef<long>, c10::ArrayRef<long>, long, bool, bool, std::array<bool, 2ul>) + 0x50f (0x7f8f195b6ecf in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #17: <unknown function> + 0x31d37a5 (0x7f8f196237a5 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #18: <unknown function> + 0x322c0a0 (0x7f8f1967c0a0 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cuda.so)
frame #19: <unknown function> + 0x2b44c5a (0x7f8f2c138c5a in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #20: <unknown function> + 0x2ba21b0 (0x7f8f2c1961b0 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #21: torch::autograd::generated::CudnnConvolutionTransposeBackward::apply(std::vector<at::Tensor, std::allocator<at::Tensor> >&&) + 0x41e (0x7f8f2bdcfdce in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #22: <unknown function> + 0x2bc4f5e (0x7f8f2c1b8f5e in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #23: torch::autograd::Engine::evaluate_function(std::shared_ptr<torch::autograd::GraphTask>&, torch::autograd::Node*, torch::autograd::InputBuffer&) + 0x1748 (0x7f8f2c1b3d88 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #24: torch::autograd::Engine::thread_main(std::shared_ptr<torch::autograd::GraphTask> const&, bool) + 0x4df (0x7f8f2c1b4c1f in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #25: torch::autograd::Engine::thread_init(int, std::shared_ptr<torch::autograd::ReadyQueue> const&) + 0xc0 (0x7f8f2c1aa200 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_cpu.so)
frame #26: torch::autograd::python::PythonEngine::thread_init(int, std::shared_ptr<torch::autograd::ReadyQueue> const&) + 0x50 (0x7f8ff7b4b440 in /home/zack/anaconda3/envs/pt16/lib/python3.8/site-packages/torch/lib/libtorch_python.so)
frame #27: <unknown function> + 0xc819d (0x7f8fff2d319d in /home/zack/anaconda3/envs/pt16/bin/../lib/libstdc++.so.6)
frame #28: <unknown function> + 0x76db (0x7f900410f6db in /lib/x86_64-linux-gnu/libpthread.so.0)
frame #29: clone + 0x3f (0x7f9003e3888f in /lib/x86_64-linux-gnu/libc.so.6)
```
While the code compile custom ops, I've confirmed that if I replace the custom ops with the native versions the error is still the same (though the frames are a little different). The error in both cases happens when calling backwards on a loss which includes a gradient term (i.e. pytorch is taking a second derivative). The exact location is here: https://github.com/rosinality/stylegan2-pytorch/blob/ab37d5da06bc80257b87a0725628cb9f9f8dee18/train.py#L152. If I remove all of these loss terms, the error goes away, so I suspect the error lies in a combination of the architecture of the generator, large tensors, and second derivatives.
So, while this is similar to [#22496 ](https://github.com/pytorch/pytorch/issues/22496) and a few other errors which appear to have been resolved in updated pytorch versions, it seems under these circumstances there are still issues.
## To Reproduce
Steps to reproduce the behavior:
I can make a minimal example if it would be helpful, but is there something obvious I should try first?
## Expected behavior
No errors.
## Environment
```
PyTorch version: 1.6.0a0+e311e53
Is debug build: No
CUDA used to build PyTorch: 10.2
OS: Ubuntu 18.04.4 LTS
GCC version: (Ubuntu 7.4.0-1ubuntu1~18.04.1) 7.4.0
CMake version: version 3.17.0
Python version: 3.8
Is CUDA available: Yes
CUDA runtime version: 10.2.89
GPU models and configuration:
GPU 0: Quadro RTX 8000
GPU 1: Quadro RTX 8000
GPU 2: Quadro RTX 8000
GPU 3: Quadro RTX 8000
GPU 4: Quadro RTX 8000
GPU 5: Quadro RTX 8000
GPU 6: Quadro RTX 8000
GPU 7: Quadro RTX 8000
Nvidia driver version: 440.44
cuDNN version: /usr/lib/x86_64-linux-gnu/libcudnn.so.7.6.5
Versions of relevant libraries:
[pip3] numpy==1.13.3
[pip3] torch==1.4.0
[pip3] torchvision==0.5.0
[conda] mkl 2019.0 pypi_0 pypi
[conda] mkl-include 2019.0 pypi_0 pypi
[conda] numpy 1.18.2 pypi_0 pypi
[conda] torch 1.6.0a0+e311e53 pypi_0 pypi
[conda] torchvision 0.6.0a0+dec8628 pypi_0 pypi
```
cc @csarofeen @ptrblck
|
module: cudnn,triaged
|
low
|
Critical
|
597,533,122 |
godot
|
The FileSystem dock highlights the main scene, but doesn't update if the main scene changes
|
Godot 3.2.1
Following this QA post:
I just realized the FileSystem dock highlights the main scene with a blue color (by default). However, it doesn't update when I change that scene.
Forcing a rebuild of the tree (by changing its layout or restarting the editor) makes it update.
On a side note, I'm wondering what benefit there is to do something like that. It seems quite peculiar, and not immediately obvious.
|
enhancement,topic:editor
|
low
|
Minor
|
597,539,365 |
flutter
|
[gen_l10n] Support deferred loading of localizations in non-Flutter web platforms
|
With https://github.com/flutter/flutter/pull/53824, Flutter web now supports deferred loading of localization files, allowing for lazy loading of localization files as they are needed rather than loading them all on start up. This is generally useful for applications with support for many locales, since we may not want to download or load all localization files until they are needed.
We should look into seeing if some similar optimizations can be done for non-web platforms.
|
c: new feature,framework,a: internationalization,a: size,c: proposal,perf: app size,P3,team-framework,triaged-framework
|
low
|
Minor
|
597,551,689 |
TypeScript
|
Type inference in index type assignment is skipped
|
<!-- 🚨 STOP 🚨 𝗦𝗧𝗢𝗣 🚨 𝑺𝑻𝑶𝑷 🚨
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:** 3.8.3
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:** index indexed type assignment initialize initializer initialization inference object narrow narrowing assertion
**Code**
```ts
type One = { name: string; };
type Many = Array<One>;
type OneOrMany = One | Many;
var one: OneOrMany = { name: "sam" };
one.name = "pat"
var many: OneOrMany = [{ name: "sam" },{ name: "ash" }];
many[0].name = "pat"
// all is well here
interface OOMDict { [key: string]: OneOrMany; }
var oomdict2: OOMDict = {};
oomdict2.one = { name: "sam" };
oomdict2.one.name = "pat";
oomdict2.many = [{ name: "sam" }, { name: "ash" }];
oomdict2.many[0].name = "pat";
// all is well here
// testing resolution within an index type initializer
var oomdict: OOMDict = { one: { name: "sam" }, many: [{ name: "sam" }, { name: "ash" }] };
oomdict.one.name = "pat";
// ERROR: Property 'name' does not exist on type 'OneOrMany'.
// ERROR: Property 'name' does not exist on type 'Many'.(2339)
oomdict.many[0].name = "pat";
// ERROR: Element implicitly has an 'any' type because expression of type '0' can't be used to index type 'OneOrMany'.
// ERROR: Property '0' does not exist on type 'OneOrMany'.(7053)
```
**Expected behavior:** Initialization / assignment of an OOMDict object effects type inference on its OneOrMany values identically to initialization or assignment of a OneOrMany variable.
**Actual behavior:** Values in an OOMDict either do not narrow based on their contents and type inference (oomdict.many does not narrow from OneOrMany to Many) or they narrow to the wrong specific type (oomdict.one narrows from OneOfMany to Many instead of One).
**Playground Link:** https://www.typescriptlang.org/play/?ssl=1&ssc=1&pln=56&pc=1#code/PTAECMEMGcEsGNQBcCeAHAptAUKzoB5AOw1AF5QBvUIyAWwwC5RokAnWIgcwG5QBfHrnSkAspCIpyoAIJs2kFAB5iGAHxC8pVQTbjJ01aAA+ofSiHYQyLEk5dQbLAHsANgFc7zoqGcAzZAALbRJdc2QRCBgMABNfH3hvJAwiJBwAN0g2eKZCUL0JKQpqWgZmACJoenKBIWtIJCDYaFA0Z05G71JAmFBIVydIGKlaeWcAd1jQPzZnOjyMMMLkZwWo6CnvUFg00ETUlLSAGj6-ZOznbM5WDCHfAKRgvugNtmS4nQKDLT7wZ3SMNgugA6UqkCjlNANcrYTLZOiFZifcIUADaJXouUq1QEJwxZVA5RggRq-AAunUwAiDD0WqNZpM4jM5gslt9VuEoBs4lsdi19slUjhqShUQAGMmgzHSSHQ7Ai4EItAACgAHuQ1KBVQBKKxgfqubYtSauVzYPU2Vj2RwuDxeHz+IIhRZfKQ-cY7QKcPo+TgxDDqrTYDoYNh+SDwbQEUQAEQQjWoqIA1hgUMxWBxuGSkflzHx+LCsr45jF4wAmJHRuPwRrFQRAkvl4FdUDWFvNUBJogTHxIVbgUiQVmuvqNR4dtodbCgGfSfFYqp0UmUp3POBcIgMVI0LIMzYkaazebI5Z9hYNuilmtl5skKUMGVQpDlITORvXxXLawio2d7vjXt+0HYdwgaJoWknVJp1nNF5wqRdSTxHcCSJaASQECkLUeQcXlgDct0aekJimH9mWPXNTw5QoLyvJAbxFcVJTBR9oRXA1fxNM0sNsa0nGgNxPFgLYPXHHwJG2Ih-UDSJOB2WB+lgAAvUNCwud8kErWN4znHJmDgwkENxUARWYdFkIXHF+CQsEKmJUkyVqLCvRaGB103Q5QFpUAADkCAAFR3MZGVXH5HTfS941vUhLmLCKa0-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-Bi5UMcaFMpG9fj3CbjBtSoaCZ3Civk1TByKEX5ft5QClWzAVCkckGp89IE7oG1xpglcNAaaQR6ooZljn1fdTX7ppjpSa1iFpFykFDLMNgOVfj-EBBaWW1ohx+DKjWISPhJ7ZEyPzGIDRkESRuHcVW6BrTYUIFWbSfIMCuD8Ngf08BXBZFIAgogSCtjoO+g0DAypN7xhbjWbUzB0jtBiNgfYbt1IABYdJdD0p3QeS4jK937piLusirLD0UYSOyGFHIsMwckDhYjtSWCAA
**Related Issues:** https://github.com/Microsoft/TypeScript/issues/1706 https://github.com/microsoft/TypeScript/issues/5089
|
Needs Investigation
|
low
|
Critical
|
597,575,584 |
flutter
|
Expose Android setSystemGestureExclusionRects and getSystemGestureExclusionRects
|
We should expose the [setSystemGestureExclusionRects](https://developer.android.com/reference/android/view/View#setSystemGestureExclusionRects(java.util.List%3Candroid.graphics.Rect%3E)) and [getSystemGestureExclusionRects](https://developer.android.com/reference/android/view/View#getSystemGestureExclusionRects()) Android APIs, which were introduced with Android 10 and its full gesture navigation functionality.
|
c: new feature,platform-android,framework,f: gestures,c: proposal,P3,team-android,triaged-android
|
low
|
Major
|
597,585,341 |
TypeScript
|
Include documentation from a variable's type when you hover the variable
|
Issue Type: <b>Feature request</b>
I had hoped to see the description for the type when hovering over the value via intellisense, etc.
When I hover over `context.slot` I can see "The slot the disk is in." yet when I try hovering over `const slot` I see just the type. Is there anything to enable this? Even setting `slot` to `Slot` manually it still only shows the type.
```ts
/** The slot the disk is in. */
type Slot = string;
interface Context {
/** The slot the disk is in. */
slot?: Slot
}
const func = (context: Context): Context => {
const x = context.slot;
const slot: Slot = context.slot;
return {
slot: x || slot
};
};
func({
slot: '1'
}).slot;
```
VS Code version: Code 1.43.2 (0ba0ca52957102ca3527cf479571617f0de6ed50, 2020-03-24T07:34:57.037Z)
OS version: Darwin x64 19.2.0
<details>
<summary>System Info</summary>
|Item|Value|
|---|---|
|CPUs|Intel(R) Core(TM) i5 CPU M 520 @ 2.40GHz (4 x 2400)|
|GPU Status|2d_canvas: unavailable_software<br>flash_3d: disabled_software<br>flash_stage3d: disabled_software<br>flash_stage3d_baseline: disabled_software<br>gpu_compositing: disabled_software<br>metal: disabled_off<br>multiple_raster_threads: enabled_on<br>oop_rasterization: disabled_off<br>protected_video_decode: disabled_off<br>rasterization: disabled_software<br>skia_renderer: disabled_off_ok<br>video_decode: disabled_software<br>viz_display_compositor: enabled_on<br>viz_hit_test_surface_layer: disabled_off_ok<br>webgl: unavailable_software<br>webgl2: unavailable_software|
|Load (avg)|1, 1, 2|
|Memory (System)|8.00GB (0.79GB free)|
|Process Argv|.|
|Screen Reader|no|
|VM|0%|
</details><details><summary>Extensions (27)</summary>
Extension|Author (truncated)|Version
---|---|---
go-outliner|766|0.1.20
npm-intellisense|chr|1.3.0
path-intellisense|chr|1.4.2
jsrefactor|cms|2.20.5
vue-peek|dar|1.0.2
vscode-eslint|dba|2.1.2
EditorConfig|Edi|0.14.5
vscode-npm-script|eg2|0.3.11
flow-for-vscode|flo|1.5.0
todo-tree|Gru|0.0.171
docthis|joe|0.7.1
vscode-home-assistant|kee|1.6.1
dotenv|mik|1.0.1
vscode-docker|ms-|1.0.0
csharp|ms-|1.21.16
atom-keybindings|ms-|3.0.6
cpptools|ms-|0.27.0
Go|ms-|0.13.1
vscode-typescript-next|ms-|3.9.20200408
go-doc|msy|0.1.1
hide-gitignored|npx|1.1.0
vetur|oct|0.24.0
vscode-graphql|Pri|0.2.14
json-template|sto|0.3.1
vscode-icons|vsc|10.0.0
vscode-wakatime|Wak|4.0.0
vscode-go-autotest|win|1.6.0
(2 theme extensions excluded)
</details>
<!-- generated by issue reporter -->
|
Suggestion,Awaiting More Feedback
|
medium
|
Major
|
597,585,825 |
flutter
|
WidgetSpans in RTL text laying out LTR if they occur on the same line
|
## Steps to Reproduce
<!-- You must include full steps to reproduce so that we can reproduce the problem. -->
Run the following app:
```
import 'package:flutter/material.dart';
void main() => runApp(App());
class App extends StatelessWidget {
@override
Widget build(BuildContext context) {
return MaterialApp(
title: 'Widget Test',
home: TestWidgetSpans(),
);
}
}
class TestWidgetSpans extends StatelessWidget {
@override
Widget build(BuildContext context) {
return Directionality(
textDirection: TextDirection.rtl,
child: Scaffold(
body: Center(
child: Text.rich(
TextSpan(
text: 'هذا اختبار',
style: TextStyle(
backgroundColor: Colors.grey.withOpacity(0.5),
fontSize: 30,
),
children: [
WidgetSpan(
child: Container(
width: 30.0,
color: Colors.blue.withOpacity(0.5),
child: Center(
child: Text('1'),
),
),
),
TextSpan(text: ' و '),
WidgetSpan(
child: Container(
width: 50.0,
color: Colors.red.withOpacity(0.5),
child: Center(
child: Text('2'),
),
),
),
TextSpan(text: ' ثم '),
WidgetSpan(
child: Container(
width: 70.0,
color: Colors.green.withOpacity(0.5),
child: Center(
child: Text('3'),
),
),
),
TextSpan(text: ' ، لكنه معطل'),
],
),
),
),
),
);
}
}
```
**Expected results:** <!-- what did you want to see? -->
The `1`, `2`, and `3` widgets laid out RTL.
**Actual results:** <!-- what did you see? -->
<details>
<summary>Logs</summary>
<!-- Finally, paste the output of running `flutter doctor -v` here. -->
```
[√] Flutter (Channel stable, v1.12.13+hotfix.9, on Microsoft Windows [Version 10.0.18362.720], locale en-US)
• Flutter version 1.12.13+hotfix.9 at C:\src\flutter
• Framework revision f139b11009 (10 days ago), 2020-03-30 13:57:30 -0700
• Engine revision af51afceb8
• Dart version 2.7.2
[√] Android toolchain - develop for Android devices (Android SDK version 29.0.3)
• Android SDK at C:\Users\mimbr\AppData\Local\Android\Sdk
• Android NDK location not configured (optional; useful for native profiling support)
• Platform android-29, build-tools 29.0.3
• ANDROID_HOME = C:\Users\mimbr\AppData\Local\Android\Sdk
• Java binary at: C:\Program Files\Android\Android Studio\jre\bin\java
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b03)
• All Android licenses accepted.
[√] Android Studio (version 3.5)
• Android Studio at C:\Program Files\Android\Android Studio
• Flutter plugin version 44.0.1
• Dart plugin version 191.8593
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b03)
[√] VS Code (version 1.44.0)
• VS Code at C:\Users\mimbr\AppData\Local\Programs\Microsoft VS Code
• Flutter extension version 3.9.1
[√] Connected device (1 available)
• sdk gphone x86 • emulator-5554 • android-x86 • Android R (API 29) (emulator)
• No issues found!
```
</details>
|
framework,a: internationalization,a: quality,a: typography,has reproducible steps,P2,found in release: 3.7,found in release: 3.10,team-framework,triaged-framework
|
low
|
Critical
|
597,612,670 |
TypeScript
|
Find all references on overridden subclass method includes references to the parent class
|
<!-- 🚨 STOP 🚨 𝗦𝗧𝗢𝗣 🚨 𝑺𝑻𝑶𝑷 🚨
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
-->
https://github.com/microsoft/vscode/issues/93654
**TypeScript Version:** 3.9.0-dev.20200409
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
- Find all references
- inheritance
**Code**
For the TS
```ts
class Foo {
bar() { }
}
class SubFoo extends Foo {
bar() { }
}
new SubFoo().bar();
```
1. Run `find all references` on `bar` in `class SubFoo`
**Expected behavior:**
A reference to `new SubFoo().bar();` is returned
**Actual behavior:**
Two references are returned, including one to the parent class's `bar`
|
Suggestion,Awaiting More Feedback,Domain: Symbol Navigation
|
low
|
Critical
|
597,621,135 |
go
|
cmd/go: make package lookups in vendor directories case-insensitive on case-sensitive filesystems
|
### What version of Go are you using (`go version`)?
<pre>
$ go version
1.13.7
</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/username/Library/Caches/go-build"
GOENV="/Users/username/Library/Application Support/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="darwin"
GOOS="darwin"
GOPATH="/Users/username/Documents/src/go"
GOPROXY="https://proxy.golang.org,direct"
GOROOT="/usr/local/Cellar/go/1.13.7/libexec"
GOSUMDB="sum.golang.org"
GOTMPDIR=""
GOTOOLDIR="/usr/local/Cellar/go/1.13.7/libexec/pkg/tool/darwin_amd64"
GCCGO="gccgo"
AR="ar"
CC="clang"
CXX="clang++"
CGO_ENABLED="1"
GOMOD="/dev/null"
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/77/06148cr15lvb1jbd9slc58v80000gq/T/go-build264165664=/tmp/go-build -gno-record-gcc-switches -fno-common"
</pre></details>
### What did you do?
`go mod vendor`
### What happened?
When vendoring on MacOS, files are written with case sensitivity to a case insensitive filesystem. This can cause two issues, but is only present for vendoring, not the module cache, since modules use `!` prefixes to make the module cache case insensitive.
#### Collision
Because import paths are case sensitive, in the event that two packages share case insensitive , but not case sensitive, prefixes collision can occur. This is infrequent due to `alllowercaseoneword` package name conventions, and github username and repo name fields being case insensitive (but with case metadata), however the it is a bug and I can provide code samples to produce compilation failures.
#### Portability
This issue was discovered when sharing the vendor directory between systems that have different case sensitivity (MacOS to Linux) and does not require package collision to reproduce.
Given two modules share a case insensitive prefix, say `github.com/Golang/exp` and `github.com/golang/tools`. On MacOS, they will both be written to `github.com/[Gg]olang` and the first writer will pick the case. Builds work fine because both `github.com/Golang` and `github.com/golang` exist, but when you mount the `vendor` directory in a docker container `github.com/Golang` stops existing. This also happens for other data sharing mechasisms between filesystems like `cp`, `tar`, `git`, etc.
|
help wanted,Vendoring,NeedsInvestigation,FeatureRequest
|
medium
|
Critical
|
597,621,766 |
PowerToys
|
[Run] To decide and remove the hardcoded pattern and maxCount values
|
https://github.com/microsoft/Launcher/blob/ca916deda91181ce335374b21e7f195e77861e59/src/modules/launcher/Plugins/Wox.Plugin.Indexer/SearchHelper/WindowsSearchAPI.cs#L91
Right now we have hardcoded values for the parameter and the maxSearchResults.
* To decide how we want to support regex based search and allow that pattern to be passed.
* To decide on the maximum number of results that we want to populate.
|
Product-PowerToys Run,Area-Quality
|
low
|
Major
|
597,622,043 |
PowerToys
|
[Run] Implement Logical search in Windows Indexer Plugin similar to Everything
|
# Summary of the new feature/enhancement
Everything has some special operators which allow users to query, such as -
Operator | Operation
-- | --
space | AND
\| | OR
! | NOT
< > | Grouping
" " | Search for an exact phrase
<!--
A clear and concise description of what the problem is that the new feature would solve.
Describe why and how a user would use this new functionality (if applicable).
-->
# Proposed technical implementation details (optional)
To modify the DB search query that we are creating by adding these features to help in searching for a query.
<!--
A clear and concise description of what you want to happen.
-->
|
Idea-Enhancement,Product-PowerToys Run
|
low
|
Major
|
597,627,419 |
go
|
x/playground: send SIGQUIT on timeout to provide useful context
|
The playground could kill processes in a way that provided the user helpful context instead of `Process.Kill()`.
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.14 linux/amd64 (playground)
</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
</pre></details>
### What did you do?
```
package main
func main() {
for {
}
}
```
### What did you expect to see?
```
timeout running program
SIGQUIT: quit
PC=0x458050 m=0 sigcode=0
goroutine 1 [running]:
main.main()
/tmp/sandbox300849141/prog.go:4 fp=0xc000034788 sp=0xc000034780 pc=0x458050
runtime.main()
/usr/local/go-faketime/src/runtime/proc.go:203 +0x20e fp=0xc0000347e0 sp=0xc000034788 pc=0x42bf4e
runtime.goexit()
/usr/local/go-faketime/src/runtime/asm_amd64.s:1373 +0x1 fp=0xc0000347e8 sp=0xc0000347e0 pc=0x453831
rax 0x458050
rbx 0x0
rcx 0x4d0000
rdx 0x47a788
rdi 0xc000060010
rsi 0x0
rbp 0xc0000347d0
rsp 0xc000034780
r8 0x0
r9 0x1
r10 0xc000060000
r11 0x1
r12 0x0
r13 0x40
r14 0x483ecc
r15 0x0
rip 0x458050
rflags 0x246
cs 0x33
fs 0x0
gs 0x0
Program exited: status 2.
```
### What did you see instead?
```timeout running program```
/cc @bcmills @cagedmantis @dmitshur
|
NeedsDecision
|
low
|
Minor
|
597,637,010 |
pytorch
|
Memory leak issue still exists in CI
|
## 🐛 Bug
I checked several CI from one of today's commits, see
cuda 10.1 nogpu: https://circleci.com/gh/pytorch/pytorch/5100845?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link
cuda 10.1: https://circleci.com/gh/pytorch/pytorch/5101775?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link
cuda 10.2: https://circleci.com/gh/pytorch/pytorch/5100839?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link
They all **displayed passed**. However, if you open "Test" log and search "Leak Summary", you can find those memory leaks from
```
valgrind --suppressions=/var/lib/jenkins/workspace/aten/tools/valgrind.sup --error-exitcode=1 ./basic '--gtest_filter=-*CUDA'
```
e.g.
```
Apr 09 21:22:34 ==73453== HEAP SUMMARY:
Apr 09 21:22:34 ==73453== in use at exit: 707,522,005 bytes in 616,626 blocks
Apr 09 21:22:34 ==73453== total heap usage: 1,620,959 allocs, 1,004,333 frees, 1,764,490,084 bytes allocated
Apr 09 21:22:34 ==73453==
Apr 09 21:22:38 ==73453== LEAK SUMMARY:
Apr 09 21:22:38 ==73453== definitely lost: 72,888 bytes in 6 blocks
Apr 09 21:22:38 ==73453== indirectly lost: 72 bytes in 3 blocks
Apr 09 21:22:38 ==73453== possibly lost: 3,279,992 bytes in 24,437 blocks
Apr 09 21:22:38 ==73453== still reachable: 704,169,053 bytes in 592,180 blocks
Apr 09 21:22:38 ==73453== suppressed: 0 bytes in 0 blocks
Apr 09 21:22:38 ==73453== Rerun with --leak-check=full to see details of leaked memory
```
## To Reproduce
I checked the latest version of this file. https://github.com/pytorch/pytorch/blob/6d3783a6bccf7054e793e5e617e46747c1d54ce6/aten/src/ATen/test/basic.cpp
It has three tests `BasicTestCPU`, `BasicTestCUDA`, and `FactoryMethodsTest`.
The first contains cpu code, the second contains cuda code and the third contains both.
So I did tests like this for **CPU**:
```sh
cd aten/tools
valgrind --suppressions=./valgrind.sup --error-exitcode=1 ../../build/bin/basic --gtest_filter=BasicTest.BasicTestCPU
```
I got
```
==3843385==
==3843385== HEAP SUMMARY:
==3843385== in use at exit: 1,419,660 bytes in 15,506 blocks
==3843385== total heap usage: 684,285 allocs, 668,779 frees, 83,394,360 bytes allocated
==3843385==
==3843385== LEAK SUMMARY:
==3843385== definitely lost: 16 bytes in 2 blocks
==3843385== indirectly lost: 4 bytes in 1 blocks
==3843385== possibly lost: 25,864 bytes in 64 blocks
==3843385== still reachable: 1,393,776 bytes in 15,439 blocks
==3843385== suppressed: 0 bytes in 0 blocks
==3843385== Rerun with --leak-check=full to see details of leaked memory
==3843385==
==3843385== For lists of detected and suppressed errors, rerun with: -s
==3843385== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 16 from 2)
```
Then I did this for **GPU**:
```sh
valgrind --suppressions=./valgrind.sup --error-exitcode=1 ../../build/bin/basic --gtest_filter=BasicTest.BasicTestGPU
```
I got
```
==3843732==
==3843732== HEAP SUMMARY:
==3843732== in use at exit: 999,399 bytes in 15,369 blocks
==3843732== total heap usage: 280,321 allocs, 264,952 frees, 40,447,163 bytes allocated
==3843732==
==3843732== LEAK SUMMARY:
==3843732== definitely lost: 0 bytes in 0 blocks
==3843732== indirectly lost: 0 bytes in 0 blocks
==3843732== possibly lost: 0 bytes in 0 blocks
==3843732== still reachable: 999,399 bytes in 15,369 blocks
==3843732== suppressed: 0 bytes in 0 blocks
==3843732== Rerun with --leak-check=full to see details of leaked memory
==3843732==
==3843732== For lists of detected and suppressed errors, rerun with: -s
==3843732== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)
```
`FactoryMethodsTest` took too long on my machine so I stopped it.
The two tests give similar leak/no leak results on my machine with **cuda 10.1 and 10.2**
## Expected behavior
No memory leak and correct CI return code.
## Environment
- PyTorch Version (e.g., 1.0): 1.6.0a0+1ffc2d9
- OS (e.g., Linux): Fedora 29
- How you installed PyTorch (`conda`, `pip`, source): source
- Build command you used (if compiling from source): python setup.py develop --user
- Python version: 3.7.6
- CUDA/cuDNN version: 10.1 and 10.2
- GPU models and configuration: RTX 2070 Super
- Any other relevant information:
## Additional context
This issue may be related to #33471 #34241
For internal reference: 2870960
cc. @ptrblck @ezyang @seemethere
cc @ezyang @seemethere
|
module: build,module: ci,triaged
|
low
|
Critical
|
597,676,145 |
deno
|
Proposal: file system map
|
## Motivation
When develop CLI app with multiple input and output files, I often find myself writing these file names twice in the command line, for example:
```sh
deno run \
--allow-read=input1.txt,input2.txt \
--allow-write=output1.txt,output2.txt \
my-cli-app \
--input1=input1.txt --input2=input2.txt \
--output1=output1.txt --output2=output2.txt
```
## Proposal
Inspired by `--mapdir` in wasmtime and wasmer, and `--mount` in Docker, I suggest we introduce a `--map-fs`, `--map-fs-read`, and `--map-fs-write` flag. The above code snippet can now be rewritten as:
```sh
# my-cli-app will only ever read from /input1 and /input2 and write to /output1 and /output2
# regardless of flags.
# It is user's responsibility to map /input1, /input2, /output1, and /output2 to desired location
deno run \
--map-fs-read=/input1::input1.txt --map-fs-read=/input2::input2.txt \
--map-fs-write=/output1::output1.txt --map-fs-write=/output2::output2.txt \
my-cli-app
```
### Restrictions
* Mount points must not collide with or include module paths.
* Guess mount points (/input1, /input2, /output1, and /output2) must always be absolute.
|
cli,suggestion
|
low
|
Minor
|
597,683,544 |
ant-design
|
Responsive Sider should have an option to collapse / expand on mobile devices without squashing the main content
|
- [x] I have searched the [issues](https://github.com/ant-design/ant-design/issues) of this repository and believe that this is not a duplicate.
### What problem does this feature solve?
Currently when using Responsive Sider on a smaller screen, when the sider appears it would have to squash the main content (and all contents within it). It is not a big problem on bigger mobile devices like an iPad but on cellphones it creats chaos and unnecessary performance expense.
I propose a property called "expandBehavior" which let the user to choose when the sider appears should it be "over the top" of the main content, or "At the side" of the main content.
### What does the proposed API look like?
```
<Sider width={300} breakpoint="lg" collapsedWidth="0" expandBehaviour="overTheTop">
{children}
</Sider>
```
The above setting should allow the Sider to be `on top of the main content` with absolute positioning, 100% height of the screen and a dark overlay to cover the content when Sider appears.
```
<Sider width={300} breakpoint="lg" collapsedWidth="0" expandBehaviour="sideBySide">
{children}
</Sider>
```
The above setting should allow the Sider to be `at the side of the content`. This is the current behaviour.
<!-- generated by ant-design-issue-helper. DO NOT REMOVE -->
|
💡 Feature Request,Inactive
|
low
|
Major
|
597,694,507 |
TypeScript
|
tsconfig.json should support multiple configurations
|
## Search Terms
multiple configurations, transitive closure
## Suggestion
When using multiple packages in a monorepo, it is common to have a `tsconfig.shared.json` that all packages extend in their individual `tsconfig.json`. The current system makes it difficult to support various build configurations across packages (such as "dev" vs. "prod").
Because a property named `"config"` seems redundant in a file named `tsconfig.json`, I'll borrow the term [select](https://docs.bazel.build/versions/master/configurable-attributes.html) from Bazel and propose the following new option in a `tsconfig.json` file for defining configurations:
```js
// tsconfig.shared.json
{
// Via "select", we define two configurations: "universal" and "modern".
"select": {
"universal": {
"compilerOptions": {
"target": "es3",
"downlevelIteration": true,
},
},
"modern": {
"compilerOptions": {
"target": "es2019",
},
},
},
// These compilerOptions are common to both configurations.
"compilerOptions": {
"sourceMap": true,
"module": "esnext",
"moduleResolution": "node",
"lib": ["dom", "es5", "es2019"],
"strict": true,
},
}
```
To complement this, `tsc` would have to support a new `--select` flag:
```
# Assume tsconfig.json extends tsconfig.json and defines "include" and "references".
# This would build the modern configuration.
$ tsc --build --select modern -p tsconfig.json
```
Ideally, it would also be possible to parameterize paths such as `outDir` so that you could also define this on the base `compilerOptions` in `tsconfig.shared.json`:
```
"outDir": "./dist/${select}/",
```
though I haven't been able to get `"outDir"` to do what I want in `tsconfig.shared.json` in my own project, so there might be more work to do on that front.
## Use Cases
I want to be able to have a shared `tsconfig.json` that defines multiple configurations in *one* place that can be shared across *multiple* modules, as opposed to the current situation where we need *O*(M ⋅ C) `tsconfig.json` files where M is the number of modules and C is the number of configurations. A detailed example is below.
## Examples
I have a Yarn workspace where each package is under the `modules/` folder, so my root `package.json` contains the following:
```json
{
"workspaces": [
"modules/*"
],
"scripts": {
"compile": "tsc --build modules/*"
}
}
```
I also have a `shared.tsconfig.json` in the root of my Yarn workspace where I define the `"compilerOptions"` that I want all packages in the workspace to use. (For example, I set `"target": "es2019"` in `shared.tsconfig.json`.) Each folder under `modules/` contains its own `tsconfig.json` file that contains the line:
```json
"extends": "../../tsconfig.shared.json",
```
and there is also a `"references"` property with the appropriate list of relative paths to other packages in the workspace.
Ideally, each such `tsconfig.json` would contain only `"extends"` and `"references"`, but it seems I have to redeclare `"compilerOptions.outDir"` and `"include"` in each of these files even though the values are the same in each one (`["src/**/*"]` and `"./dist"`, respectively).
OK, so far, so good, but now I want to be able to build my entire Yarn workspace with a different set of `compilerOptions`, specifically:
```
"downlevelIteration": true,
"target": "es3",
```
Ideally, I would be able to specify this in *one* place and leverage the `"references"` I already had to define so I could compile one `.ts` file or package and all of its transitive deps with these compiler options. (I happen to be using `ts-loader` in Webpack, so my ultimate output is a single `.js` file, which perhaps biases my expectations here.)
Unfortunately, there does not appear to be any clean way to do that. I could go through and define a `tsconfig.es3.json` in each of my packages that looks like this:
```
{
"extends": "./tsconfig.json",
"compilerOptions": {
"downlevelIteration": true,
"target": "es3",
},
}
```
though even if I did that, I'm not sure I could build all of my modules as ES3 with a one-liner as I did in my original `package.json` file. Now I'm faced with the additional incidental complexity of:
* Maintaining *O*(M) `tsconfig.es3.json` files and ensuring all of them are in sync.
* Introducing Lerna or some other tool to "walk my DAG" and run `tsc -b -p tsconfig.es3.json` or something like that.
With the `--select` proposal, I could avoid the extra `tsconfig.json` files and still build via a one-liner:
```
tsc --build --select universal modules/*
```
I would expect if any of the `tsconfig.json` files under `modules/*` did not define a `"universal"` configuration, the build command would fail (or at least the strictness should be configurable).
## 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
|
Major
|
597,704,818 |
go
|
cmd/go: unclear error when SumDB tree note is corrupted
|
<!--
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 go1.14.2 windows/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
set GO111MODULE=on
set GOARCH=amd64
set GOBIN=
set GOCACHE=C:\Users\Nicolas\AppData\Local\go-build
set GOENV=C:\Users\Nicolas\AppData\Roaming\go\env
set GOEXE=.exe
set GOFLAGS= -mod=
set GOHOSTARCH=amd64
set GOHOSTOS=windows
set GOINSECURE=
set GONOPROXY=
set GONOSUMDB=
set GOOS=windows
set GOPATH=D:\Mis Documentos\Programacion\Go
set GOPRIVATE=
set GOPROXY=direct
set GOROOT=C:\Go
set GOSUMDB=sum.golang.org
set GOTMPDIR=
set GOTOOLDIR=C:\Go\pkg\tool\windows_amd64
set GCCGO=gccgo
set AR=ar
set CC=gcc
set CXX=g++
set CGO_ENABLED=1
set GOMOD=NUL
set CGO_CFLAGS=-g -O2
set CGO_CPPFLAGS=
set CGO_CXXFLAGS=-g -O2
set CGO_FFLAGS=-g -O2
set CGO_LDFLAGS=-g -O2
set PKG_CONFIG=pkg-config
set GOGCCFLAGS=-m64 -mthreads -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -fdebug-prefix-map=C:\Users\Nicolas\AppData\Local\Temp\go-build045229398=/tmp/go-build -gno-record-gcc-switches
</pre></details>
### What did you do?
<!--
If possible, provide a recipe for reproducing the error.
A complete runnable program is good.
A link on play.golang.org is best.
-->
I'm trying to use go mod tidy command. Since my computer turned off because of a blackout it never worked again
### What did you expect to see?
Succesful go.mod file
### What did you see instead?
For every project and import I'm seeing this error:
```
github.com/sirupsen/logrus: github.com/sirupsen/[email protected]: verifying module: github.com/sirupsen/[email protected]: initializing sumdb.Client: reading tree note: malformed note
note:
```
```
github.com/dgrijalva/jwt-go: github.com/dgrijalva/[email protected]: verifying module: github.com/dgrijalva/[email protected]: initializing sumdb.Client: reading tree note: malformed note
note:
```
I know I just copied 2 libraries, but It's happening with every one
|
NeedsInvestigation,modules
|
medium
|
Critical
|
597,708,897 |
neovim
|
Unable to use `:w !sudo tee %` command in NeoVim
|
Hello! I already know about https://github.com/neovim/neovim/issues/1716 and https://github.com/neovim/neovim/issues/1496, but I wanted to open this issue to help people find it more easily in search engines.
I am unable to run `:w !sudo tee %` as I get the following error, but searching for the error along with the words "neovim" or "vim" led to [zero results on Google](https://www.google.com/search?q=neovim+%22sudo%3A+a+terminal+is+required+to+read+the+password%3B+either+use+the+-S+option+to+read+from+standard+input+or+configure+an+askpass+helper%22) (at time of opening this issue):
```
sudo: a terminal is required to read the password; either use the -S option to read from standard input or configure an askpass helper
```
So I hope that by pasting the full output here, which is not listed in the other two issues, other people may be able to land here more easily and see the relevant issues https://github.com/neovim/neovim/issues/1716 and https://github.com/neovim/neovim/issues/1496. (Plus sometimes it helps me re-stumble on old issues in the future).
Please feel free to close this.
|
bug,job-control,tui,channels-rpc,has:workaround,closed:duplicate
|
high
|
Critical
|
597,727,371 |
excalidraw
|
Rotate in discrete steps on mobile phone
|
I love the rotate in discrete steps feature (by constraining the rotation by holding the SHIFT key) but I would like to use that on my phone too. Is it possible to add some kind for setting for that so that I can define the degrees which must be used when rotating? Or I would be happy with a setting for 22.5, 30 or 45 degrees steps.
A solution without a special UI setting would be great although I don't know what would be possible. Perhaps rotating closer tot the center (as mentioned as a response to my tweet). Or always use small discrete steps when using a touch interface, e.g. small steps of 7.5 degrees would also be acceptable.
|
mobile,UX/UI
|
low
|
Major
|
597,728,297 |
pytorch
|
Jit doesn't match schema like eager mode does
|
```
import torch
def fn(a, b: float, out):
return torch.div(a, b, out=out)
t = torch.tensor((2.,))
divisor = 2.
out = torch.empty((0,))
print(fn(t, divisor, out))
scripted = torch.jit.script(fn)
print(scripted(t, divisor, out))
```
This will print:
```
tensor([1.])
Traceback (most recent call last):
File "../../issues.py", line 12, in <module>
scripted = torch.jit.script(fn)
File "/private/home/mruberry/git/pytorch/torch/jit/__init__.py", line 1289, in script
fn = torch._C._jit_script_compile(qualified_name, ast, _rcb, get_default_args(obj))
RuntimeError:
Arguments for call are not valid.
The following variants are available:
aten::div.Tensor(Tensor self, Tensor other) -> (Tensor):
Expected a value of type 'Tensor' for argument 'other' but instead found type 'float'.
aten::div.out(Tensor self, Tensor other, *, Tensor(a!) out) -> (Tensor(a!)):
Expected a value of type 'Tensor' for argument 'other' but instead found type 'float'.
aten::div.Scalar(Tensor self, Scalar other) -> (Tensor):
Keyword argument out unknown.
aten::div.int(int a, int b) -> (float):
Expected a value of type 'int' for argument 'b' but instead found type 'float'.
aten::div.float(float a, float b) -> (float):
Keyword argument out unknown.
aten::div(Scalar a, Scalar b) -> (float):
Keyword argument out unknown.
div(float a, Tensor b) -> (Tensor):
Expected a value of type 'Tensor' for argument 'b' but instead found type 'float'.
div(int a, Tensor b) -> (Tensor):
Expected a value of type 'Tensor' for argument 'b' but instead found type 'float'.
The original call is:
File "../../issues.py", line 4
def fn(a, b: float, out):
return torch.div(a, b, out=out)
~~~~~~~~~ <--- HERE
```
The jit is correct that there is no schema in native functions for div(Tensor, Scalar, Out), but eager mode performs the operation without issue.
@bhosmer
cc @suo
|
oncall: jit,triaged
|
low
|
Critical
|
597,732,228 |
godot
|
Editor crashes when navigating through call stack in debug mode "Index p_index is out of bounds"
|
**Godot version:**
3.2.1.stable.official
**OS/device including version:**
Windows 10 x64 Home Edition
**Issue description:**
While running my game, the following exception was thrown because I call .instance() without providing the required variables for _init()
I began to navigate through the call stack in the debugger. When I clicked on line 3, the editor hung for a second and then crashed, but the console window stayed open (showing the output below). This is the line that clicking on call stack line "3" _should_ have brought me to:
```
func process_actions(delta: float) -> void:
emit_signal('fire')
```
**Console Output**
```
bug 127.0.0.1:6007 --allow_focus_steal_pid 7700 --position -1472,300
Godot Engine v3.2.1.stable.official - https://godotengine.org
OpenGL ES 3.0 Renderer: GeForce GTX 1060 3GB/PCIe/SSE2
ERROR: _create_instance: Condition "r_error.error != Variant::CallError::CALL_OK" is true. Returned: __null
At: modules/gdscript/gdscript.cpp:127
ERROR: unfold_line: Index p_line = -1 is out of bounds (text.size() = 28).
At: scene/gui/text_edit.cpp:6009
ERROR: set_executing_line: Index p_line = -1 is out of bounds (text.size() = 28).
At: scene/gui/text_edit.cpp:5649
ERROR: unfold_line: Index p_line = -1 is out of bounds (text.size() = 5).
```
ABOVE THIS LINE IS GENERATED WHEN CODE EXCEPTION THROWN. BELOW THIS LINE GENERATED WHILE BROWSING STACK HISTORY IN DEBUGGER
```
At: scene/gui/text_edit.cpp:6009
ERROR: set_executing_line: Index p_line = -1 is out of bounds (text.size() = 5).
At: scene/gui/text_edit.cpp:5649
ERROR: get: FATAL: Index p_index = 29 is out of bounds (size() = 29).
At: ./core/cowdata.h:152
```
**Steps to reproduce:**
Open project at the linked and run it. Navigate to call stack item 3 in the debugger window when the exception is thrown.
**Minimal reproduction project:**
[Debug.zip](https://github.com/godotengine/godot/files/4460284/Debug.zip)
|
bug,topic:editor,confirmed,crash
|
low
|
Critical
|
597,772,568 |
youtube-dl
|
Please add space.bilibili.com
|
## Checklist
- [x ] I'm reporting a new site support request
- [x ] I've verified that I'm running youtube-dl version **2020.03.24**
- [x ] I've checked that all provided URLs are alive and playable in a browser
- [x ] I've checked that none of provided URLs violate any copyrights
- [x ] I've searched the bugtracker for similar site support requests including closed ones
## Example URLs
- Single video: https://space.bilibili.com/178781504
- Single video: https://space.bilibili.com/178781504/video
- Playlist: https://space.bilibili.com/178781504
## Description
[root@144 ~]# youtube-dl -v -F -i https://space.bilibili.com/178781504
[debug] System config: []
[debug] User config: []
[debug] Custom config: []
[debug] Command-line args: [u'-v', u'-F', u'-i', u'https://space.bilibili.com/178781504']
[debug] Encodings: locale UTF-8, fs UTF-8, out UTF-8, pref UTF-8
[debug] youtube-dl version 2020.03.24
[debug] Python version 2.7.5 (CPython) - Linux-4.14.129-bbrplus-x86_64-with-centos-7.7.1908-Core
[debug] exe versions: ffmpeg 3.1, ffprobe 3.1
[debug] Proxy map: {}
[generic] 178781504: Requesting header
WARNING: Falling back on generic information extractor.
[generic] 178781504: Downloading webpage
[generic] 178781504: Extracting information
ERROR: Unsupported URL: https://space.bilibili.com/178781504
Traceback (most recent call last):
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/generic.py", line 2375, in _real_extract
doc = compat_etree_fromstring(webpage.encode('utf-8'))
File "/usr/local/bin/youtube-dl/youtube_dl/compat.py", line 2551, in compat_etree_fromstring
doc = _XML(text, parser=etree.XMLParser(target=_TreeBuilder(element_factory=_element_factory)))
File "/usr/local/bin/youtube-dl/youtube_dl/compat.py", line 2540, in _XML
parser.feed(text)
File "/usr/lib64/python2.7/xml/etree/ElementTree.py", line 1642, in feed
self._raiseerror(v)
File "/usr/lib64/python2.7/xml/etree/ElementTree.py", line 1506, in _raiseerror
raise err
ParseError: not well-formed (invalid token): line 1, column 613
Traceback (most recent call last):
File "/usr/local/bin/youtube-dl/youtube_dl/YoutubeDL.py", line 797, in extract_info
ie_result = ie.extract(url)
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/common.py", line 530, in extract
ie_result = self._real_extract(url)
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/generic.py", line 3352, in _real_extract
raise UnsupportedError(url)
UnsupportedError: Unsupported URL: https://space.bilibili.com/178781504
[root@144 ~]# youtube-dl -v -F -i https://space.bilibili.com/178781504/video
[debug] System config: []
[debug] User config: []
[debug] Custom config: []
[debug] Command-line args: [u'-v', u'-F', u'-i', u'https://space.bilibili.com/178781504/video']
[debug] Encodings: locale UTF-8, fs UTF-8, out UTF-8, pref UTF-8
[debug] youtube-dl version 2020.03.24
[debug] Python version 2.7.5 (CPython) - Linux-4.14.129-bbrplus-x86_64-with-centos-7.7.1908-Core
[debug] exe versions: ffmpeg 3.1, ffprobe 3.1
[debug] Proxy map: {}
[generic] video: Requesting header
WARNING: Falling back on generic information extractor.
[generic] video: Downloading webpage
[generic] video: Extracting information
ERROR: Unsupported URL: https://space.bilibili.com/178781504/video
Traceback (most recent call last):
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/generic.py", line 2375, in _real_extract
doc = compat_etree_fromstring(webpage.encode('utf-8'))
File "/usr/local/bin/youtube-dl/youtube_dl/compat.py", line 2551, in compat_etree_fromstring
doc = _XML(text, parser=etree.XMLParser(target=_TreeBuilder(element_factory=_element_factory)))
File "/usr/local/bin/youtube-dl/youtube_dl/compat.py", line 2540, in _XML
parser.feed(text)
File "/usr/lib64/python2.7/xml/etree/ElementTree.py", line 1642, in feed
self._raiseerror(v)
File "/usr/lib64/python2.7/xml/etree/ElementTree.py", line 1506, in _raiseerror
raise err
ParseError: not well-formed (invalid token): line 1, column 613
Traceback (most recent call last):
File "/usr/local/bin/youtube-dl/youtube_dl/YoutubeDL.py", line 797, in extract_info
ie_result = ie.extract(url)
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/common.py", line 530, in extract
ie_result = self._real_extract(url)
File "/usr/local/bin/youtube-dl/youtube_dl/extractor/generic.py", line 3352, in _real_extract
raise UnsupportedError(url)
UnsupportedError: Unsupported URL: https://space.bilibili.com/178781504/video
|
site-support-request
|
low
|
Critical
|
597,782,132 |
pytorch
|
CMake targets wrongly forward unknown options to NVCC (v1.5+)
|
## 🐛 Bug
It seems that versions 1.5+ have introduced two cmake targets (torch_cpu and torch_cuda), but these forward unknown options to nvcc (`nvcc fatal : Unknown option 'Wall'`).
Also see this torchvision issue: https://github.com/pytorch/vision/issues/2001
## To Reproduce
Steps to reproduce the behavior:
1. Download a nightly version of libtorch from the website
2. Use it to build torchvision (eg. using the command
```cmake -DCMAKE_PREFIX_PATH=/home/user/libtorch1.6 -DWITH_CUDA=ON ..```)
3. Try to compile (e.g. `make`)
It will result in `nvcc fatal : Unknown option 'Wall'`
## Expected behavior
Compiles successfully like with current stable (1.4)
## Environment
Collecting environment information...
PyTorch version: 1.6.0.dev20200409+cu101
Is debug build: No
CUDA used to build PyTorch: 10.1
OS: Ubuntu 18.04.4 LTS
GCC version: (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0
CMake version: version 3.17.1
Python version: 3.6
Is CUDA available: Yes
CUDA runtime version: 10.1.243
GPU models and configuration: GPU 0: GeForce GTX 1060 6GB
Nvidia driver version: 440.33.01
cuDNN version: /usr/lib/x86_64-linux-gnu/libcudnn.so.7.6.5
## Additional context
It doesn't seem to be a cmake problem, as versions 3.10, 3.13 and 3.17 are all affected.
Pytorch v1.5.0a0 is also affected according to the torchvision issue.
Most likely the newly created targets are adding the compiler flags without taking into account the cuda compiler.
cc @peterjc123
|
module: build,triaged
|
low
|
Critical
|
597,797,864 |
go
|
net: clarify safe way to do URL deep-copying
|
It sometimes happens to have a need to duplicate (aka "deep copy") a `net.URL` object.
My understanding is that `net.URL` can be safely duplicated by simply copying the structure:
```go
newUrl := *url
```
Since `net.URL` is usually handled by pointer, it's normal to check the documentation to see whether the object contains internal pointers, in which case duplication might get hairy. `net.URL` has the following contents:
```go
type URL struct {
Scheme string
Opaque string // encoded opaque data
User *Userinfo // username and password information
Host string // host or host:port
Path string // path (relative paths may omit leading slash)
RawPath string // encoded path hint (see EscapedPath method); added in Go 1.5
ForceQuery bool // append a query ('?') even if RawQuery is empty; added in Go 1.7
RawQuery string // encoded query values, without '?'
Fragment string // fragment for references, without '#'
}
```
So there is indeed a pointer in there. This brought people to come up with alternative ways of duplicating a URL such as a `String`/`Parse` roundtrip:
https://medium.com/@nrxr/quick-dirty-way-to-deep-copy-a-url-url-on-go-464312241de4
The trick here is that `User *Userinfo` is actually "immutable". This is documented in `Userinfo`: `The Userinfo type is an immutable encapsulation of username and password [...]`. So it looks like that, given this guarantee of immutability, it is indeed safe to simply copy a URL instance to duplicate it. Copying a URL instance is also performed by the standard library, for instance:
https://github.com/golang/go/blob/245409ea86f20fd9f4167223c2339fb238f9e4b6/src/net/http/request.go#L361
Hence the subject of this issue:
* Can I get a confirmation that copying a URL is the correct way and not just an accident of its current implementation (so, it's guaranteed by Go 1.0 compatibility)?
* Should we document it a little bit better, since it tends to be confusing to newcomers to this library? Either explicitly stating the correct way of copying it, or moving the "immutable" word in the inline documentation of the `User *Userinfo` field (or both?).
|
Documentation,NeedsInvestigation
|
medium
|
Major
|
597,800,866 |
pytorch
|
Population Count Op
|
## 🚀 Feature
It would be great to have the bitwise operation 'Population Count' that counts the number of 1 bits to be exposed.
Not sure if this exists in the back end of torch? It is exposed in [tensorflow](https://github.com/tensorflow/tensorflow/issues/1592)
## Motivation
This is a key operation in implementing binarised neural nets, for example in [XNOR-Net](https://arxiv.org/pdf/1603.05279.pdf). Binarised linear and conv layers can be performed quickly using XNOR and population counts. XNOR is already exposed in the pytorch API.
cc @albanD
|
feature,triaged,actionable,module: python frontend
|
low
|
Major
|
597,806,655 |
TypeScript
|
Constants + Generic causing TS4023 ("but cannot be named")
|
**TypeScript Version:** 3.8, but also the 3.9 beta
**Search Terms:** ts4023 cannot be named constant
**Code**
Create a tsconfig with `declaration: true`. Then, in `main.ts`:
```ts
import {StructType} from './lib';
declare function foo<T>(): T;
export const exportedFooFunc = foo<StructType>();
```
In `lib.ts`:
```ts
export const SOME_CONSTANT = 'fieldKey';
export type MakeStruct<S> = S;
export type StructType = MakeStruct<{
readonly [SOME_CONSTANT]: any;
}>;
```
---
<details>
<summary>
Old instructions (don't repro anymore)
</summary>
Create a tsconfig with `declaration: true`. Then, in `main.ts`:
```ts
import {
ReturnType1,
ReturnType2,
} from './lib';
export const impl1 = (): ReturnType1 => ({
});
export const impl2 = (): ReturnType2 => ({
});
```
In `lib.ts`:
```ts
export const SOME_CONSTANT = 'myConstant';
export type WrapType<S> = S;
export type ReturnType1 = WrapType<{
[SOME_CONSTANT]?: {};
}>;
export type ReturnType2 = {
[SOME_CONSTANT]?: {};
};
```
</details>
**Expected behavior:**
The file should be properly compiled.
**Actual behavior:**
TypeScript reports a diagnostic:
```
Exported variable 'exportedFooFunc' has or is using name 'SOME_CONSTANT' from external module "/.../lib" but cannot be named.
```
**Playground Link:** n/a
**Related Issues:** n/a
|
Needs Investigation,Rescheduled
|
low
|
Major
|
597,808,803 |
three.js
|
TrackballControls are not mouse event to rAF ratio independent.
|
##### Description of the problem
The TrackballControls seem to be designed that they take input from mouseMove and use it in update which is expected be called from requestAnimationFrame? They seem to expect there to be a 1x1 relationship, one mouseMove followed by 1 rAF. But if there are more mousemove events per rAF then they stop working correctly (run really slow)
Here's a repo. Not sure this happens on all machines or OSes or browser. I'm in Chrome 80 on Windows 10. I also have the Oculus Rift software installed (no idea what that does to my system)
https://jsfiddle.net/greggman/qz7jug2v/
Here's what I see.
With the devtools closed
With the devtools open
Adding some code I found with the devtools closed there is 1 mouse event per rAF but open I saw as many as 16 events per rAF
Adding calls to `_this.update()` in `mousemove`, `touchmove`, and `wheel` seems to fix it since then each event is handled individually?
https://jsfiddle.net/greggman/k8fotepb/
Note I didn't test touchmove but I'm guessing it can't hurt?
##### Three.js version
- [x] Dev
- [x] r115
- [ ] ...
##### Browser
- [x] All of them
- [x] Chrome
- [?] Firefox
- [?] Internet Explorer
##### OS
- [x] All of them
- [x] Windows
- [?] macOS
- [?] Linux
- [ ] Android
- [ ] iOS
##### Hardware Requirements (graphics card, VR Device, ...)
|
Browser Issue
|
low
|
Major
|
597,835,182 |
godot
|
Visual Script "make Tool" Fail
|
**Godot version:**
Stable 3.2.1. official
**OS/device including version:**
Windows 10
**Issue description:**
When you leave the project, all VS that had tool on turn it off, causes issues with plugins because of the Tool visual scripts that just don't properly save their Tool status
**Steps to reproduce:**
Make Visual Script, Toggle Make Tool on (save ofc), Go to project manager, load project, load visual script --> make Tool will be off
**Minimal reproduction project:**
No point in sending as it'll just be an empty script with "make tool" off even if i turned it on
|
bug,topic:editor,topic:visualscript
|
low
|
Minor
|
597,838,581 |
terminal
|
Vim in WSL scrolls 3 lines when gaining focus with mouse click
|
# Environment
```none
Windows build number: 10.0.18363.752
Windows Terminal version (if applicable): I am using native WSL terminal
Any other software? Vim
```
# Steps to reproduce
I use Vim inside the default WSL terminal. When I click inside Vim to regain focus on the terminal, then it scrolls down 3 lines. This doesn't happen if gain focus using control tab or if I click somewhere inside Vim when I already have the terminal focused.
# Expected behavior
Vim should not scroll.
# Actual behavior
Vim scrolls down 3 lines.
Gif for clarity: https://i.redd.it/vxyk7h9bpyr41.gif
|
Product-Conhost,Area-Input,Area-VT,Issue-Bug,Priority-2
|
low
|
Major
|
597,852,507 |
godot
|
Text in EditorHelp can overflow horizontally
|
**Godot version:**
c52beb6adb3b7af6f9ca79b6dfad1783a1fae351
**Issue description:**
This is in Camera2D.
|
bug,topic:editor
|
low
|
Minor
|
597,853,959 |
terminal
|
Feature Request: Multiple Cloud Shells for Multiple accounts
|
I have multiple accounts into the same Azure Tenant, different accounts for different environments Dev/QA/etc, but I don't see a way of easily switching accounts in the same tenant. Is that something that could be added?
_@DHowett note: this will also help with people who cannot log in for one or another reason but need to hardcode their tenant IDs_
|
Issue-Bug,Area-Settings,Product-Terminal,Issue-Task,Area-AzureShell
|
low
|
Major
|
597,872,919 |
PowerToys
|
limit maximum CPU usage of certain process/program
|
Summary:
Add the capability of being able to limit maximum CPU usage of certain process.
Scenario:
When using a laptop/tablet you don't want it to get overheated, despite decreasing performance at expense. Windows can already do this by setting maximum processor state.
But this isn't practical at all as the computer would be crazily slow if the value is too low while setting it higher makes no effect on cooling down.
By applying the same settings but to certain processes only, users can watch YouTube unzipping large files without heat or lag, or play games while encoding videos background with minor to no frame rate loss.
Technical details:
A utility called [Battle Encoder Shirase](http://mion.faireal.net/BES/) (or BES in short) achieved this, by quickly suspending & resuming target processes at certain time ratios. It works on all modern processors and requires access to suspend/resume only which means no need to run as administrator.
BES is open source in GNU license so you may migrate it into PowerToys.
Implement:
BES lists all processes in which you can set specific ones to be limited. It is also capable of limiting all processes with same path.
In additional to that PowerToys may also integrate with #142 so when an application is minimized its CPU time gets limited.
There should be a default blocklist of which processes shouldn't be limited such as svchost.exe (users get warned when trying to limit one of those)
|
Idea-New PowerToy,Product-Power management
|
low
|
Major
|
597,873,021 |
flutter
|
appbundle apk size inconsistent with "flutter build apk --split-per-abi" apk
|
Issue: I have a problem with my Flutter application
I wanted to use appbundle for this release of my since they are the recommendation now. So I decided to test it on internal app sharing first before I release to my users. After I uploaded the appbundle and started to download, the apk size shown to me during download was 18.68MB while when I did split 'flutter build apk --split-per-abi' 8.7MB.
I thought appbundle apks are supposed to be smaller not bigger than the normal apks. Please How do I resolve this issue. I have tried in on the latest stable channel and the Latest beta channel its still the same results.
|
platform-android,tool,c: performance,a: size,perf: app size,P2,team-android,triaged-android
|
low
|
Major
|
597,898,515 |
pytorch
|
Docs of distributed module do not include the full documentation for torch.distributed.launch
|
## 📚 Documentation
https://pytorch.org/docs/stable/distributed.html#launch-utility does not include the full documentation from https://github.com/pytorch/pytorch/blob/f326045/torch/distributed/launch.py#L4-L140. An earlier version of the docs at https://pytorch.org/docs/1.1.0/distributed_deprecated.html#launch-utility does not have this issue.
|
module: docs,triaged
|
low
|
Minor
|
597,899,008 |
scrcpy
|
Does it burden mobile device CPU if I increase bitrate and resolution?
|
If I stream high-resolution video from my mobile, will it also increase stress on mobile device? Or this will only increase load on PC?
Please highlight this. Does bumping quality adversely affect mobile performance or not?
|
question,performance
|
low
|
Major
|
597,903,941 |
storybook
|
Custom icon in the tree
|
**Is your feature request related to a problem? Please describe.**
I'm making a MDX docs-only story that will act as the homepage of our Storybook instance, and it would be nice to give that item a homepage-worthy icon in the tree.
**Describe the solution you'd like**
Perhaps in the `<Meta>` MDX element, add a prop that accepts an SVG path and a colour, that will be passed on to the tree.
**Describe alternatives you've considered**
I don't think a workaround is possible, because icons are hardcoded in [ListItem.tsx](https://github.com/storybookjs/storybook/blob/c9f702dbf45df6c026a1fb28ae2f7aaff1a95678/lib/ui/src/components/sidebar/Tree/ListItem.tsx).
I've tried to add an emoji into the title of a story, but that doesn't look quite right.
**Are you able to assist bring the feature to reality?**
Maybe. Depends on the solution we're heading for, should this feature request be approved.
**Additional context**
Not just homepages, you know. It might be helpful to easily distinguish different component by different icons, or different colours of icons.
|
feature request,ui
|
low
|
Major
|
597,931,675 |
terminal
|
Feature Request: Fuzzy Search
|
An innate feature to search through the Terminal text is quite limited and basic:
- it doesn't highlight several entries in the window;
- it doesn't support using regular expressions;
- it can't filter output to a fewer number of lines (much like grep does).
- it only supports strict match;
With the power of fuzzy search, users gain a boost to their productivity of being able to find a text through history very fast. It's implemented in many text editors (JetBrains IDEA, Visual Studio Code) for their project's files search. There's a solution specifically for the command line (bash/zsh) as well - [junegunn/fzf](https://github.com/junegunn/fzf).
If we could integrate the feature into the terminal (via its UI) without relying on the bash settings, it would be a great capability!
|
Area-UserInterface,Area-TerminalControl,Product-Terminal,Issue-Task
|
low
|
Minor
|
597,966,040 |
flutter
|
Consider using compositing triggers for better performance on the Web
|
Flutter [avoids 3D transforms](https://github.com/flutter/engine/pull/15649) on the Web as much as possible because it causes [text blurriness](https://github.com/flutter/flutter/issues/32274). The Chrome team is now working fixing it ([Chrome issue](https://bugs.chromium.org/p/chromium/issues/detail?id=642885)). Eliminating 3D completely can lose some layerization optimizations that Blink can do for us.
We could strategically add a 3D component (e.g. `translateZ(0)`) to trigger Chrome to create a layer so during animation we could skip browser-side repaints.
`will-change: transform` (we can probably also use it for opacity) also triggers a layer, although the text fix will not apply.
Leaving this issue here as a possible future exploration after we eliminate the heavier bottlenecks in the system.
Hat tip to @progers for providing the info.
/cc @ferhatb
|
framework,engine,c: performance,a: typography,platform-web,c: rendering,perf: speed,e: web_html,P3,team-web,triaged-web
|
low
|
Critical
|
598,017,098 |
vscode
|
Truncate and copy large values from the debug console
|
<!-- ⚠️⚠️ Do Not Delete This! feature_request_template ⚠️⚠️ -->
In js-debug, we truncate large data inside the debug adapter before giving back to the UI. However, this means that if the user logs a very large string, they have no way to view it. The Chrome devtools solve this by adding a "copy" and "show all" button to very large strings.
We don't currently have a way to show buttons in the output. I believe the just having a "copy" button in the UI is sufficient. An approach for this is to use the `evaluateName` in the output's `variablesReference` with an `evaluate` call, in the `clipboard` context if supported. It's notable that the UI can use this even for values that aren't truncated to allow the user to copy complex expressions.
The question of who does the truncation is orthogonal to the ability to copy. If the debug adapter continues to do the truncation, then it should have a DAP flag to hint to the UI that the value can be expanded (using the presence of an `evaluateName` alone would prevent thatgeneralized copy functionality).
If the UI does the truncation, it means more traffic over the protocol, and it might also be tricky to figure out how to show a truncated state for complex objects. But then the UI can also implement "show all" without any debug adapter changes. Overall I think UI-side truncation is preferable.
cc @isidorn @weinand @roblourens
|
feature-request,debug
|
medium
|
Critical
|
598,023,570 |
flutter
|
iOS text selection toolbar animation
|
On the native iOS text selection menu/toolbar, tapping Select All causes the menu to fade out and then back in with new children. In Flutter it's an unanimated sudden change.
| Native iOS | Flutter |
| ---------- | ------- |
| <video src="https://github.com/flutter/flutter/assets/948037/30b05cb4-6649-4b9c-9aad-1087568d1e3c" /> |  |
|
a: text input,platform-ios,framework,a: fidelity,f: cupertino,P3,team-design,triaged-design
|
low
|
Major
|
598,027,548 |
godot
|
Tools Configuration not created by default in .csproj
|
**Godot version:** 3.2.1.stable.mono.official
**OS/device including version:** Windows
**Issue description:** Default .csproj created by Godot does not have a Tools configuration, preventing creating plugins in C#
**Steps to reproduce:**
Create a brand new Godot project. Create a C# plugin. Try to activate the plugin. It will not activate. Error dialog says "Unable to load addon script from path: 'res://addons/my_plugin/my_plugin.cs' There seems to be an error in the code, please check the syntax."
This can be solved by manually creating a Tools configuration in the .csproj. I have not been able to find any documentation stating it is necessary to do this.
**Minimal reproduction project:**
|
topic:dotnet
|
low
|
Critical
|
598,050,768 |
godot
|
Index p_bone = -1 is out of bounds (bones.size() = 0) when PhysicalBone is child of Skeleton3D
|
**Godot version:**
4.0.dev.custom_build. ae42cb7b0 but it also happens with Godot 3.2
**OS/device including version:**
Ubuntu 19.10
**Issue description:**
Godot spams
```
scene/3d/skeleton_3d.cpp:694 - Index p_bone = -1 is out of bounds (bones.size() = 0).
```
when PhysicalBone is child of Skeleton
In constructor physical bone, id is set to - 1.
Probably it needs to use `PhysicalBone3D::update_bone_id` after changing of parent to Skeleton
**Steps to reproduce:**
1. Create Skeleton(3D)
2. Create PhysicalBone(3D) as child of this Skeleton
|
bug,topic:core
|
low
|
Minor
|
598,062,250 |
go
|
build: prefix baked-in environment variables with GO_ or similar
|
Per request here: https://github.com/golang/go/issues/38361#issuecomment-612184174
Currently a `go` binary is built with the value of the `PKG_CONFIG` variable as-defined at build-time, and uses that variable to find the `pkg-config` binary:
https://github.com/golang/go/blob/master/src/make.bash#L53
A [coincidental change in Homebrew](https://github.com/Homebrew/brew/pull/7277/files#diff-b7261a4b3eaff4e69f0871ecfd15ca54R115) broke cgo builds for anyone using Homebrew to manage their Go installation.
Given that `pkg-config` is critical for building Go packages with cgo, maybe it makes sense to rename the variable to `GO_PKG_CONFIG` so it’s unlikely to be clobbered by upstream build systems. Or, going further, prefix any other environment variables baked into the Go binary with `GO`?
|
NeedsInvestigation
|
low
|
Major
|
598,062,769 |
pytorch
|
Unable to build wheel from RC3
|
## 🐛 Bug
I am trying to build a wheel using RC3. Because certain files(like third_party/gloo/CMakeLists.txt) are missing in the .zip file, I am trying to build from source using the commit hash (build wheel torch-1.5.0a0+b58f89b). However, I am unable to build a wheel (output below). On the other hand, I am able to building wheel torch-1.6.0a0+c5662dd, from the latest pull rebase from source in my local repo. Can someone help me understand this behavior and what I can do to successfully build a 1.5.0rc3 wheel? Thanks.
## To Reproduce
Steps to reproduce the behavior:
```python setup.py bdist_wheel --universal --plat-name=manylinux1_x86_64```
Output:
```
Building wheel torch-1.5.0a0+b58f89b
-- Building version 1.5.0a0+b58f89b
cmake --build . --target install --config Release -- -j 8
No such file or directory
CMake Error: Generator: execution of make failed. Make command was: //anaconda3/bin/ninja -j 8 install &&
Traceback (most recent call last):
File "setup.py", line 745, in <module>
build_deps()
File "setup.py", line 316, in build_deps
cmake=cmake)
File "/Users/choidong/pytorch/tools/build_pytorch_libs.py", line 62, in build_caffe2
cmake.build(my_env)
File "/Users/choidong/pytorch/tools/setup_helpers/cmake.py", line 339, in build
self.run(build_args, my_env)
File "/Users/choidong/pytorch/tools/setup_helpers/cmake.py", line 141, in run
check_call(command, cwd=self.build_dir, env=env)
File "//anaconda3/lib/python3.7/subprocess.py", line 347, in check_call
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['cmake', '--build', '.', '--target', 'install', '--config', 'Release', '--', '-j', '8']' returned non-zero exit status 1.
```
## Expected behavior
Successfully build a wheel.
## Environment
```
PyTorch version: N/A
Is debug build: N/A
CUDA used to build PyTorch: N/A
OS: Mac OSX 10.14.6
GCC version: Could not collect
CMake version: version 3.15.3
Python version: 3.7
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:
[pip3] numpy==1.18.1
[pip3] numpydoc==0.9.2
[pip3] torch==1.3.0a0+d53d9a4
[pip3] torchvision==0.4.2
[conda] blas 1.0 mkl
[conda] mkl 2019.4 233
[conda] mkl-service 2.3.0 py37hfbe908c_0
[conda] mkl_fft 1.0.15 py37h5e564d8_0
[conda] mkl_random 1.1.0 py37ha771720_0
[conda] numpy 1.18.1 py37h7241aed_0
[conda] numpy-base 1.18.1 py37h6575580_1
[conda] numpydoc 0.9.2 py_0
[conda] torch 1.3.0a0+d53d9a4 dev_0 <develop>
[conda] torchvision 0.4.2 pypi_0 pypi
```
|
module: build,triaged
|
low
|
Critical
|
598,076,150 |
terminal
|
Feature Request: Reduce cursor blinking to save battery life
|
# Description of the new feature/enhancement
In several more modern UI experiences in Windows, the blinking cursor effect has been changed to reduce battery impact - I'm proposing Terminal should do the same. The cursor stops blinking after a period of inactivity to save on the amount of CPU/GPU cycles spent running the animation, updating the graphics, updating the screen, stuff like that.
To see an example of this in action, use Outlook. Start writing an e-mail, type a little bit, then pause your typing for several seconds. The cursor will go solid and stop blinking.
It sounds crazy that this could have a meaningful battery impact - but it certainly can. I implemented this feature in Silverlight for Windows Phone, and my team did it for XAML in UWP. In both cases the battery impact far exceeded my expectations - and this is no doubt why Outlook did it too.
# Proposed technical implementation details (optional)
The way this works in XAML, Outlook, etc. is that a 5 second timer is started after a key is let up. At the end of that 5 seconds the cursor is set to a solid 'on' state. Note that the cursor stays solid while you type too, so it stays in this 'on' state until you pause, then blinks for a few seconds, then goes back to solid.
This may have extra complexity in terminal due to block cursors and stuff like that, but it would be the rough sketch of what to do.
|
Area-Performance,Area-TerminalControl,Product-Terminal,Issue-Task
|
low
|
Major
|
598,084,305 |
go
|
net/http/httptest: (*Server).Close races with HTTP/2 request handler
|
```
~/go/src$ go version
go version devel +24ea77559f Thu Apr 9 11:09:00 2020 -0400 linux/amd64
~/go/src$ go doc httptest.Server.Close
package httptest // import "net/http/httptest"
func (s *Server) Close()
Close shuts down the server and blocks until all outstanding requests on
this server have completed.
```
While investigating #37669, I saw a reported race that led me to believe that, contrary to its documentation, `(*httptest.Server).Close` was not blocking for requests to complete.
To check that hypothesis, I added the following test function:
```diff
diff --git c/src/net/http/httptest/server_test.go w/src/net/http/httptest/server_test.go
index 0aad15c5ed..8b0e0f137f 100644
--- c/src/net/http/httptest/server_test.go
+++ w/src/net/http/httptest/server_test.go
@@ -6,9 +6,12 @@ package httptest
import (
"bufio"
+ "context"
+ "io"
"io/ioutil"
"net"
"net/http"
+ "strings"
"testing"
)
@@ -238,3 +241,51 @@ func TestTLSServerWithHTTP2(t *testing.T) {
})
}
}
+
+func TestCloseBlocksUntilRequestsCompleted(t *testing.T) {
+ t.Parallel()
+
+ modes := []string{
+ "http1",
+ "http2",
+ }
+
+ for _, mode := range modes {
+ mode := mode
+ t.Run(mode, func(t *testing.T) {
+ t.Parallel()
+
+ ctx, cancel := context.WithCancel(context.Background())
+ closed := false
+
+ cst := NewUnstartedServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+ cancel()
+
+ io.Copy(w, r.Body)
+
+ <-r.Context().Done()
+ if closed {
+ panic("Close did not wait for handler to return")
+ }
+ }))
+
+ switch mode {
+ case "http1":
+ cst.Start()
+ case "http2":
+ cst.EnableHTTP2 = true
+ cst.StartTLS()
+ }
+ defer func() {
+ cst.Close()
+ closed = true
+ }()
+
+ req, _ := http.NewRequestWithContext(ctx, "POST", cst.URL, strings.NewReader("Hello, server!"))
+ resp, err := cst.Client().Do(req)
+ if err == nil {
+ resp.Body.Close()
+ }
+ })
+ }
+}
```
It reliably fails under the race detector with even a very modest count, indicating that the server is not waiting for the handler to complete when HTTP/2 is enabled:
```
~/go/src$ go test -race net/http/httptest -run=TestCloseBlocksUntilRequestsCompleted -count=10
==================
WARNING: DATA RACE
Write at 0x00c000202018 by goroutine 10:
net/http/httptest.TestCloseBlocksUntilRequestsCompleted.func1.2()
/usr/local/google/home/bcmills/go/src/net/http/httptest/server_test.go:281 +0x46
net/http/httptest.TestCloseBlocksUntilRequestsCompleted.func1()
/usr/local/google/home/bcmills/go/src/net/http/httptest/server_test.go:289 +0x508
testing.tRunner()
/usr/local/google/home/bcmills/go/src/testing/testing.go:1022 +0x1eb
Previous read at 0x00c000202018 by goroutine 31:
net/http/httptest.TestCloseBlocksUntilRequestsCompleted.func1.1()
/usr/local/google/home/bcmills/go/src/net/http/httptest/server_test.go:267 +0x18b
net/http.HandlerFunc.ServeHTTP()
/usr/local/google/home/bcmills/go/src/net/http/server.go:2012 +0x51
net/http.serverHandler.ServeHTTP()
/usr/local/google/home/bcmills/go/src/net/http/server.go:2808 +0xc4
net/http.initALPNRequest.ServeHTTP()
/usr/local/google/home/bcmills/go/src/net/http/server.go:3380 +0xfc
net/http.(*initALPNRequest).ServeHTTP()
<autogenerated>:1 +0xa6
net/http.Handler.ServeHTTP-fm()
/usr/local/google/home/bcmills/go/src/net/http/server.go:87 +0x64
net/http.(*http2serverConn).runHandler()
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:5712 +0xc0
Goroutine 10 (running) created at:
testing.(*T).Run()
/usr/local/google/home/bcmills/go/src/testing/testing.go:1073 +0x6a3
net/http/httptest.TestCloseBlocksUntilRequestsCompleted()
/usr/local/google/home/bcmills/go/src/net/http/httptest/server_test.go:255 +0x13e
testing.tRunner()
/usr/local/google/home/bcmills/go/src/testing/testing.go:1022 +0x1eb
Goroutine 31 (finished) created at:
net/http.(*http2serverConn).processHeaders()
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:5446 +0x8fb
net/http.(*http2serverConn).processFrame()
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:4975 +0x44a
net/http.(*http2serverConn).processFrameFromReader()
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:4933 +0x750
net/http.(*http2serverConn).serve()
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:4432 +0x1485
net/http.(*http2Server).ServeConn()
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:4033 +0xd88
net/http.http2ConfigureServer.func1()
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:3859 +0x117
net/http.(*conn).serve()
/usr/local/google/home/bcmills/go/src/net/http/server.go:1805 +0x1c77
==================
--- FAIL: TestCloseBlocksUntilRequestsCompleted (0.00s)
--- FAIL: TestCloseBlocksUntilRequestsCompleted/http2 (0.02s)
testing.go:937: race detected during execution of test
2020/04/10 16:35:14 http2: panic serving 127.0.0.1:46044: Close did not wait for handler to return
goroutine 132 [running]:
net/http.(*http2serverConn).runHandler.func1(0xc0000ae068, 0xc00022af4f, 0xc00031a180)
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:5705 +0x231
panic(0x8b5220, 0x9c14f0)
/usr/local/google/home/bcmills/go/src/runtime/panic.go:975 +0x3b3
net/http/httptest.TestCloseBlocksUntilRequestsCompleted.func1.1(0x9cd0c0, 0xc0000ae068, 0xc0000ac600)
/usr/local/google/home/bcmills/go/src/net/http/httptest/server_test.go:268 +0x1e2
net/http.HandlerFunc.ServeHTTP(0xc000096740, 0x9cd0c0, 0xc0000ae068, 0xc0000ac600)
/usr/local/google/home/bcmills/go/src/net/http/server.go:2012 +0x52
net/http.serverHandler.ServeHTTP(0xc0000a8460, 0x9cd0c0, 0xc0000ae068, 0xc0000ac600)
/usr/local/google/home/bcmills/go/src/net/http/server.go:2808 +0xc5
net/http.initALPNRequest.ServeHTTP(0x9cd940, 0xc000282b70, 0xc00029a380, 0xc0000a8460, 0x9cd0c0, 0xc0000ae068, 0xc0000ac600)
/usr/local/google/home/bcmills/go/src/net/http/server.go:3380 +0xfd
net/http.(*http2serverConn).runHandler(0xc00031a180, 0xc0000ae068, 0xc0000ac600, 0xc0000977c0)
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:5712 +0xc1
created by net/http.(*http2serverConn).processHeaders
/usr/local/google/home/bcmills/go/src/net/http/h2_bundle.go:5446 +0x8fc
FAIL
FAIL net/http/httptest 0.202s
FAIL
```
CC @bradfitz @tombergan @nerd2
|
NeedsInvestigation
|
low
|
Major
|
598,084,648 |
godot
|
Removing noise from noise texture doesn't update it to empty in the editor
|
**Godot version:**
3.2.1
**Issue description:**
It shows correctly after running game or reloading scene, so it's minor thing.
|
bug,topic:core,topic:editor,confirmed
|
low
|
Minor
|
598,105,690 |
excalidraw
|
Modify Sentry error logging
|
- [ ] don't log known non-errors
- [ ] don't log same errors multiple times per session
Rationale:
- We're getting flooded by invalid errors either from extensions or some arcane browser bugs that we can't do anything about (e.g. the `window.__pad.performLoop` Safari error).
Inspiratiation:
- https://github.com/codesandbox/codesandbox-client/blob/master/packages/common/src/utils/analytics/sentry.ts
- https://github.com/expo/expo/blob/master/docs/common/sentry-utilities.js
|
enhancement
|
low
|
Critical
|
598,117,550 |
godot
|
Invalid script extension is also invalid path
|
**Godot version:**
3.2.1
**Issue description:**
Shouldn't it be only the latter? Both paths are technically valid.
|
discussion,topic:editor,usability
|
low
|
Minor
|
598,123,062 |
rust
|
De-stabilize target spec JSON
|
In a recent [Zulip discussion](https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/Stability.20guarantees.20of.20custom.20target.20specifications), it was noted that we currently consider JSON target specifications unstable, in the sense that options are relatively freely added and removed.
Currently, the compiler will load target specifications from three places.
If a triple is passed without trailing ".json":
* a built-in list, specified as Rust structs
* `$RUST_TARGET_PATH`, split like PATH on the system we're running on, or, if not set, the current directory.
* we then attempt to load `$triple.json` from the directories listed
If a triple is passed with .json, then we directly attempt to load that file path (relative or absolute).
This is the history as best as I can tell:
* Original `--target` support, including RUST_TARGET_PATH, was implemented in #16156.
* Trailing .json leading to a file lookup was added by @phil-opp in #49019.
I cannot currently find any explicit discussion around stability, but presumably that discussion did happen at some time, happy to receive links to it. I would like to propose that we de-stabilize the RUST_TARGET_PATH and file-path loading, making only built-in targets usable. To my knowledge, this essentially is already the status quo: you need to build core/std to actually use a target, and that's not possible on beta/stable without intentional stability holes like RUSTC_BOOTSTRAP.
|
A-stability,T-compiler,A-target-specs
|
medium
|
Major
|
598,129,606 |
flutter
|
Enable EqualsHashCode check in engine java files
|
Both the framework as well as Google3 enforce requiring override of both equals and hashCode if either is overridden. We should enable this in the engine as well to prevent Google3 roll breakages such as:
```
flutter_engine/shell/platform/android/io/flutter/plugin/editing/InputConnectionAdaptor.java:64: error: [EqualsHashCode] Classes that override equals should also override hashCode.
public boolean equals(Object o) {
^
```
|
team,engine,dependency: android,P2,team-engine,triaged-engine
|
low
|
Critical
|
598,129,790 |
rust
|
Bad error message using shared borrow of non-sync type across await point
|
<!--
Thank you for filing a bug report! 🐛 Please provide a short summary of the bug,
along with any information you feel relevant to replicating the bug.
-->
I tried this code:
```rust
struct NonSync(*const ());
unsafe impl Send for NonSync {}
impl NonSync {
async fn foo(&self) {}
async fn bar(&mut self) {}
}
fn require_send(_: impl Send) {}
fn main() {
require_send(async {
let mut foo = NonSync(std::ptr::null());
foo.foo().await;
});
require_send(async {
let foo = NonSync(std::ptr::null());
foo.bar().await;
});
}
```
I expected to see this happen: one of three things:
1. Both branches compile
2. Both branches don't compile
3. `foo` branch doesn't compile, but `bar` branch does, and error message explains that `&T` is non-sync while `&mut T` is sync `where T: !Sync`
Instead, this happened: *explanation*
`foo` branch doesn't compile, but `bar` branch does, error message implies that the error is a lifetime issue rather than related to shared vs exclusive borrowing.
```
error: future cannot be sent between threads safely
--> src/main.rs:12:5
|
9 | fn require_send(_: impl Send) {}
| ------------ ---- required by this bound in `require_send`
...
12 | require_send(async {
| ^^^^^^^^^^^^ future returned by `main` is not `Send`
|
= help: within `NonSync`, the trait `std::marker::Sync` is not implemented for `*const ()`
note: future is not `Send` as this value is used across an await
--> src/main.rs:14:9
|
14 | foo.foo().await;
| ---^^^^^^^^^^^^- `foo` is later dropped here
| |
| await occurs here, with `foo` maybe used later
| has type `&NonSync`
help: consider moving this into a `let` binding to create a shorter lived borrow
--> src/main.rs:14:9
|
14 | foo.foo().await;
| ^^^^^^^^^
```
### Meta
I ran this code using the latest stable (1.42.0) and nightly (1.44.0-nightly (2020-04-09 94d346360da50f159e0d)) versions of the compiler. Same error message occurs both times.
|
A-diagnostics,A-trait-system,T-compiler,C-bug,A-async-await,AsyncAwait-Triaged,WG-async
|
low
|
Critical
|
598,136,172 |
godot
|
Signal list isn't updated immediately after adding script
|
**Godot version:**
3.2.1
**Steps to reproduce:**
1. Create a script with any signal
2. Select a node and go to signals tab
3. Attach the script to node
4. The signal won't appear until you re-select the node
**Minimal reproduction project:**
[ReproductionProject.zip](https://github.com/godotengine/godot/files/4463686/ReproductionProject.zip)
|
bug,topic:editor,confirmed,usability
|
low
|
Minor
|
598,148,551 |
godot
|
New class not highlighted in previously opened scripts
|
**Godot version:**
3.2.1
**Steps to reproduce:**
1. Create a few scripts (or use reproduction project)
2. Open all of them except 1
3. Add `class_name` to one of the scripts (e.g. `class_name MyCoolClass`)
4. Refer to this class in other scripts (e.g. `MyCoolClass.new()`
You will see that `MyCoolClass` is not colored as custom class. But when you open the remaining script, it will be colored there. This is due to old scripts not refreshing class list upon adding class.
**Minimal reproduction project:**
[ReproductionProject.zip](https://github.com/godotengine/godot/files/4463753/ReproductionProject.zip)
|
bug,topic:gdscript,topic:editor
|
low
|
Minor
|
598,154,078 |
rust
|
Implement interpolation methods in std
|
Currently, linear interpolation (aka `lerp`) and `midpoint` don't exist in std.
But exists in C++20 forwards. It would be nice to have them.
<!-- TRIAGEBOT_START -->
<!-- TRIAGEBOT_ASSIGN_START -->
<!-- TRIAGEBOT_ASSIGN_DATA_START$${"user":null}$$TRIAGEBOT_ASSIGN_DATA_END -->
<!-- TRIAGEBOT_ASSIGN_END -->
<!-- TRIAGEBOT_END -->
|
T-libs-api,C-feature-request
|
low
|
Major
|
598,157,856 |
terminal
|
Support HTM (headless terminal multiplexer) for remote pane/tab management
|
<!--
🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
I ACKNOWLEDGE THE FOLLOWING BEFORE PROCEEDING:
1. If I delete this entire template and go my own path, the core team may close my issue without further explanation or engagement.
2. If I list multiple bugs/concerns in this one issue, the core team may close my issue without further explanation or engagement.
3. If I write an issue that has many duplicates, the core team may close my issue without further explanation or engagement (and without necessarily spending time to find the exact duplicate ID number).
4. If I leave the title incomplete when filing the issue, the core team may close my issue without further explanation or engagement.
5. If I file something completely blank in the body, the core team may close my issue without further explanation or engagement.
All good? Then proceed!
-->
# Description of the new feature/enhancement
<!--
A clear and concise description of what the problem is that the new feature would solve.
Describe why and how a user would use this new functionality (if applicable).
-->
As mentioned in #3656, it would be great to have `tmux` control mode support for Windows. More abstractly, it would be especially great if there was remote pane/tab management available at all, through any terminal multiplexer (not just `tmux`). In fact, @MisterTea's [EternalTerminal](https://github.com/MisterTea/EternalTerminal) has such a terminal multiplexer called "HTM" (headless terminal multiplexer). It is being adopted by Hyper.js (see https://github.com/zeit/hyper/pull/2988) and would seem to have a simpler client-side implementation than `tmux` control mode.
To be extra clear: I really want to be able to open a connection to a remote machine, then use Windows Terminal to split a pane in two, revealing two terminal inputs for the same remote machine - or use a new tab hotkey do to the same for a tab instead of a pane. It seems like HTM might be an easy way to go forward.
# Proposed technical implementation details (optional)
I'm not sure. Maybe @MisterTea would have some ideas. But from [a quick glance through the HTM source code,](https://github.com/MisterTea/EternalTerminal/blob/6fb4daf82845c4dca4f609694a3ffb921fea25c3/src/htm/HtmHeaderCodes.hpp) it would appear that one might start by parsing and interpreting its header codes.
<!--
A clear and concise description of what you want to happen.
-->
|
Issue-Feature,Area-UserInterface,Area-Extensibility,Product-Terminal
|
low
|
Critical
|
598,170,079 |
go
|
cmd/go: "fatal: git fetch-pack: expected shallow list" when retrieving earlier version of already installed package on CentOS 7.
|
### What version of Go are you using (`go version`)?
<pre>
$ go version
go version go1.13.6 linux/amd64
</pre>
### Does this issue reproduce with the latest release?
Not sure, but this is the latest supported version for CentOS 7.
### What operating system and processor architecture are you using (`go env`)?
CentOS 7.
<details><summary><code>go env</code> Output</summary><br><pre>
$ go env
GO111MODULE=""
GOARCH="amd64"
GOBIN=""
GOCACHE="/users/d00u3151/.cache/go-build"
GOENV="/users/d00u3151/.config/go/env"
GOEXE=""
GOFLAGS=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GONOPROXY=""
GONOSUMDB=""
GOOS="linux"
GOPATH="/users/d00u3151/go"
GOPRIVATE=""
GOPROXY="direct"
GOROOT="/usr/lib/golang"
GOSUMDB="off"
GOTMPDIR=""
GOTOOLDIR="/usr/lib/golang/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=/users/d00u3151/tmp/go-build945266341=/tmp/go-build -gno-record-gcc-switches"
[d00u3151@pde-bsp3hq2 dan]$
</pre></details>
### What did you do?
Tried to build the latest unreleased version of reposurgeon (problem does not occur with reposurgeon release 4.5).
<pre>
$ git clone https://gitlab.com/esr/reposurgeon.git
Cloning into 'reposurgeon'...
remote: Enumerating objects: 1203, done.
remote: Counting objects: 100% (1203/1203), done.
remote: Compressing objects: 100% (456/456), done.
remote: Total 33416 (delta 890), reused 1009 (delta 739), pack-reused 32213
Receiving objects: 100% (33416/33416), 550.06 MiB | 11.28 MiB/s, done.
Resolving deltas: 100% (22980/22980), done.
$ cd reposurgeon
$ git checkout -q b392d176f30825d150b828c4df3d2d8e2b3166f7 # version as of April 10, 2020
$ make
asciidoctor -a nofooter -b manpage reposurgeon.adoc
asciidoctor -a nofooter -b manpage repocutter.adoc
asciidoctor -a nofooter -b manpage repomapper.adoc
asciidoctor -a nofooter -b manpage repotool.adoc
asciidoctor -a nofooter -b manpage repobench.adoc
asciidoctor -a webfonts! reposurgeon.adoc
asciidoctor -a webfonts! repocutter.adoc
asciidoctor -a webfonts! repomapper.adoc
asciidoctor -a webfonts! repotool.adoc
asciidoctor -a webfonts! repobench.adoc
asciidoctor -a webfonts! repository-editing.adoc
go build -gcflags '-N -l' -o repocutter ./cutter
go: github.com/go-delve/[email protected] requires
github.com/sirupsen/[email protected]: invalid version: git fetch --unshallow -f origin in /users/d00u3151/go/pkg/mod/cache/vcs/020616345f7c7f88438c217f9d0e26744bce721c80e4a28f93399a8a4cd2acf1: exit status 128:
fatal: git fetch-pack: expected shallow list
make: *** [build] Error 1
$ make
go build -gcflags '-N -l' -o repocutter ./cutter
go: github.com/go-delve/[email protected] requires
golang.org/x/[email protected]: invalid version: git fetch --unshallow -f https://go.googlesource.com/arch in /users/d00u3151/go/pkg/mod/cache/vcs/a260a67b53c91dca287c34ff2bdee130be447c2ea411a64bf4489a69e886411b: exit status 128:
fatal: git fetch-pack: expected shallow list
make: *** [build] Error 1
</pre>
### What did you expect to see?
A successful build.
### What did you see instead?
A cryptic error message.
This seems to occur when multiple versions of the same package are required. In this case, the go.mod contains <code>golang.org/x/arch v0.0.0-20200312215426-ff8b605520f4</code> and <code>github.com/go-delve/delve v1.4.0</code>, but delve v1.4.0 has a dependency on: <code>golang.org/x/[email protected]</code>. It seems that trying to get the earlier version of arch when the later version has already been retrieved causes this error message.
In this case I was able to work around and successfully build by removing the reference to the later version of arch from go.mod.
|
NeedsInvestigation,modules
|
low
|
Critical
|
598,178,615 |
flutter
|
[google_maps_flutter] Updating marker position with change in Lat-Lng position of marker
|
<!-- Thank you for using Flutter!
Please check out our documentation first:
* https://flutter.dev/
* https://api.flutter.dev/
If you can't find the answer there, please consider asking a question on
the Stack Overflow Web site:
* https://stackoverflow.com/questions/tagged/flutter?sort=frequent
Please don't file a GitHub issue for support requests. GitHub issues are
for tracking defects in the product. If you file a bug asking for help, we
will consider this a request for a documentation update.
-->
I am making a vehicle tracking administration app using flutter. I have implemented the maps in the app using the ```google-maps-flutter``` plugin. **The part I am stuck is how to update the marker position when there is a change in the location of the vehicle**. My location data of the vehicles are stored in the firestore database which I can get into the app using the ```StreamBuilder ```.
I found a ```updateMarker``` function inside the plugin code but don't know how to implement it.
If I can not ask this type of here: I have asked the same problem in StackOverflow too the link is here:[https://stackoverflow.com/questions/61136968/updating-marker-position-with-change-in-lat-lng-position-of-marker-in-google-map]( https://stackoverflow.com/questions/61136968/updating-marker-position-with-change-in-lat-lng-position-of-marker-in-google-map)
|
c: new feature,p: maps,package,c: proposal,team-ecosystem,P3,triaged-ecosystem
|
low
|
Critical
|
598,189,562 |
rust
|
Consider skipping 'unused import' lint for traits when errors have occured
|
The following code:
```rust
trait Foo {
fn dummy(&self) {}
}
impl Foo for bool {}
mod bar {
use super::Foo;
fn contains_error() {
let a = missing_method();
a.dummy();
}
}
```
gives the following output:
```
error[E0425]: cannot find function `missing_method` in this scope
--> src/lib.rs:9:17
|
9 | let a = missing_method();
| ^^^^^^^^^^^^^^ not found in this scope
warning: unused import: `super::Foo`
--> src/lib.rs:7:9
|
7 | use super::Foo;
| ^^^^^^^^^^
|
= note: `#[warn(unused_imports)]` on by default
```
I think the 'unused import' lint firing is questionable. If the trait in question is being imported only for its methods, then the user will never actually write down the trait name in their code. As a result, introducing a seemingly unrelated compilation error during development may cause a method to no longer be resolved (`a.dummy()` in this example), resulting in the trait appearing unused.
However, the fact that the trait appears unused here is only because we weren't able to perform method resolution. If the user decides to remove the import, they'll end up with a *new* compilation error after fixing the original one, requiring them to revert their change.
I think we should suppress the `unused_imports` for traits if we encountered any errors before method resolution (or maybe before type-checking). The lint will still fire if there are other errors (e.g. borrowcheck) or no errors, but we won't give users this kind of false positive.
|
C-enhancement,A-lints,T-compiler
|
low
|
Critical
|
598,191,598 |
pytorch
|
Any reference to LPPool2d
|
## 📚 Documentation
<!-- A clear and concise description of what content in https://pytorch.org/docs is an issue. If this has to do with the general https://pytorch.org website, please file an issue at https://github.com/pytorch/pytorch.github.io/issues/new/choose instead. If this has to do with https://pytorch.org/tutorials, please file an issue at https://github.com/pytorch/tutorials/issues/new -->
Hi all,
I found the LPPool2d & LPPool1d in the pytorch implementation, but there is no reference provided in the documents. And I didn't find the related paper explaining why it works. Any related paper would be helpful, thanks!
|
module: docs,triaged
|
low
|
Minor
|
598,215,518 |
TypeScript
|
Type guards can't deal with unknown type as default
|
We can't know the failures of type inference in type guards because of this problem. Maybe `a is any[]` should also return unknown[].
<!--
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.x-dev.20200410
<!-- Search terms you tried before logging this (so others can find this issue more easily) -->
**Search Terms:**
**Code**
```ts
declare function f(a: any): a is any[];
declare function g(a: any): a is unknown[];
declare const a: readonly number[];
f(a) && a; // any[]
g(a) && a; // not unknown[]
```
**Expected behavior:**
a's type narrowed by g is `unknown[]`.
**Actual behavior:**
a's type narrowed by g is `readonly number[] & unknown[]`.
**Playground Link:** https://www.typescriptlang.org/play/?ts=3.9.0-dev.20200410&ssl=1&ssc=1&pln=6&pc=11#code/CYUwxgNghgTiAEAzArgOzAFwJYHtVIAooAueKVATwEpSp4sBnMygbQF0BuAKFElgRTpseeAHMitSjTL0maANaocAd1TtuvaHHhg8DDGVJwowPBArxUyALYAjEDHVcuiIlXgAyD2W7io7rx8gA
**Related Issues:** #17002
|
Suggestion,Needs Proposal
|
low
|
Critical
|
598,226,758 |
vue-element-admin
|
support /icons/svg/index to defined child folder of icons
|
## Feature request（新功能建议）
in router we could use subfolder project like follow
```js
meta:{
title:'dingding.settingpage',
icon:'dingding.setting'
}
```
|
feature,in plan
|
low
|
Minor
|
598,262,213 |
godot
|
ProjectSettings.load_resource_pack won't replace files.
|
**Godot version:**
3.2.1.stable
**OS/device including version:**
Windows10 (latest)
**Issue description:**
Been trying a solution for a least a few hours on this issue. ProjectSettings.load_resource_pack doesn't replace files if replace_files is true . I made one pck file using export with a scene that had a red color_rect. the pck file loads properly without errors but wont have the red square.
pls don't fix this because its the only way to make pcks to work but turning on 'convert text resources to binary on export' does make it work but filenames are like res://folder/scene.tscn.converted.res instead of res://folder/scene.tscn and they do load the red color_rect
**Steps to reproduce:**
pretty much add a red color_rect, export as pck or zip, and try to load that pck file. the pck file will load properly (will return true) it won't have the red color_rect square.
**Minimal reproduction project:**
same as above.
|
bug,topic:core
|
low
|
Critical
|
598,270,379 |
rust
|
Sub-optimal codegen: Unnecessarily dumping AVX registers to stack
|
<!--
Thank you for filing a bug report! 🐛 Please provide a short summary of the bug,
along with any information you feel relevant to replicating the bug.
-->
I tried this code ([example 1](https://rust.godbolt.org/z/86FfSi)), in which we have a public function `mutate_array` that internally calls `mutate_chunk`:
```rust
use std::arch::x86_64::*;
#[inline(always)]
pub unsafe fn mutate_chunk(rows: [__m256d; 4]) -> [__m256d; 4] {
[
_mm256_permute2f128_pd(rows[0], rows[1], 0x20),
_mm256_permute2f128_pd(rows[2], rows[3], 0x20),
_mm256_permute2f128_pd(rows[0], rows[1], 0x31),
_mm256_permute2f128_pd(rows[2], rows[3], 0x31),
]
}
#[target_feature(enable = "avx")]
pub unsafe fn mutate_array(input: *const f64, output: *mut f64) {
let mut input_data = [_mm256_setzero_pd(); 4];
for i in 0..4 {
input_data[i] = _mm256_loadu_pd(input.add(4*i));
}
let output_data = mutate_chunk(input_data);
for i in 0..4 {
_mm256_storeu_pd(output.add(4*i), output_data[i]);
}
}
```
This is a very stripped-down example of code that appears all over my project. We load data into AVX registers, do some sort of operation on the loaded data, then store it back to memory. The (more or less) optimal assembly for this example code is:
```
example::mutate_array:
vmovups ymm0, ymmword ptr [rdi]
vmovups ymm1, ymmword ptr [rdi + 32]
vmovups ymm2, ymmword ptr [rdi + 64]
vmovups ymm3, ymmword ptr [rdi + 96]
vperm2f128 ymm4, ymm0, ymm1, 32
vperm2f128 ymm5, ymm2, ymm3, 49
vperm2f128 ymm0, ymm0, ymm1, 49
vperm2f128 ymm1, ymm2, ymm3, 32
vmovups ymmword ptr [rsi], ymm4
vmovups ymmword ptr [rsi + 32], ymm1
vmovups ymmword ptr [rsi + 64], ymm0
vmovups ymmword ptr [rsi + 96], ymm5
vzeroupper
ret
```
4 loads, 4 permutes, 4 stores.
As you can see from the godbolt link, the actual generated assembly is quite a bit longer:
```
example::mutate_array:
push rbp
mov rbp, rsp
and rsp, -32
sub rsp, 288
vmovups ymm0, ymmword ptr [rdi]
vmovups ymm1, ymmword ptr [rdi + 32]
vmovups ymm2, ymmword ptr [rdi + 64]
vmovups ymm3, ymmword ptr [rdi + 96]
vmovaps ymmword ptr [rsp + 96], ymm3
vmovaps ymmword ptr [rsp + 64], ymm2
vmovaps ymmword ptr [rsp + 32], ymm1
vmovaps ymmword ptr [rsp], ymm0
vmovaps ymm0, ymmword ptr [rsp]
vmovaps ymm1, ymmword ptr [rsp + 32]
vmovaps ymm2, ymmword ptr [rsp + 64]
vmovaps ymm3, ymmword ptr [rsp + 96]
vperm2f128 ymm4, ymm0, ymm1, 32
vmovaps ymmword ptr [rsp + 128], ymm4
vperm2f128 ymm4, ymm2, ymm3, 32
vperm2f128 ymm0, ymm0, ymm1, 49
vperm2f128 ymm1, ymm2, ymm3, 49
vmovaps ymmword ptr [rsp + 160], ymm4
vmovaps ymmword ptr [rsp + 192], ymm0
vmovaps ymmword ptr [rsp + 224], ymm1
vmovaps ymm0, ymmword ptr [rsp + 224]
vmovups ymmword ptr [rsi + 96], ymm0
vmovaps ymm0, ymmword ptr [rsp + 192]
vmovups ymmword ptr [rsi + 64], ymm0
vmovaps ymm0, ymmword ptr [rsp + 160]
vmovups ymmword ptr [rsi + 32], ymm0
vmovaps ymm0, ymmword ptr [rsp + 128]
vmovups ymmword ptr [rsi], ymm0
mov rsp, rbp
pop rbp
vzeroupper
ret
```
The second assembly block is the same as the first, except for the addition of reads/writes to the `rsp` (ie the stack). It loads the 4 values from memory fine -- but before running the permutes, it stores the values to `rsp`, then immediately reads them back. Same thing after the permutes: Before writing the data to the output, it stores it to `rsp`, then immediately reads it back.
It's possible to nudge the compiler into generating the correct output by partially unrolling the input and output loops.
By changing the input loop
```rust
for i in 0..4 {
input_data[i] = _mm256_loadu_pd(input.add(4*i));
}
```
to
```rust
for i in 0..2 {
input_data[i*2] = _mm256_loadu_pd(input.add(8*i));
input_data[i*2+1] = _mm256_loadu_pd(input.add(8*i + 4));
}
```
we can see that the loop is functionally identical, but the compiler no longer writes the inputs to the stack ([example 2](https://rust.godbolt.org/z/8Ct7Uu)).
We can apply the same treatment to the output loop, completely eliminating the stack reads and writes: [example 3](https://rust.godbolt.org/z/PHhARc).
Without knowing anything about the internals of the compiler, I can imagine two possibilities here:
1. Example 1 demonstrates trivial missed optimization: the compiler is unrolling the loop, but fails to determine that it can eliminate the array. As a result, it more than doubles the instruction count of the function, tanking performance.
2. Alternatively, something in the Rust standard requires all arrays to have an in-memory representation, and they aren't allowed to be completely optimized away to register storage, even entirely within a function. If this is the case, then examples 2 and 3 demonstrate a code generation bug, because we can clearly see that the storage to the array was completely optimized away.
### Meta
<!--
If you're using the stable version of the compiler, you should also check if the
bug also exists in the beta or nightly versions.
-->
`rustc --version --verbose`:
```
rustc 1.44.0-nightly (42abbd887 2020-04-07)
binary: rustc
commit-hash: 42abbd8878d3b67238f3611b0587c704ba94f39c
commit-date: 2020-04-07
host: x86_64-pc-windows-msvc
release: 1.44.0-nightly
LLVM version: 9.0
```
|
I-slow,C-enhancement,A-codegen,T-compiler,A-SIMD,C-optimization
|
low
|
Critical
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.