https://github.com/spall closed https://github.com/llvm/llvm-project/pull/119820
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/llvm-beanz approved this pull request.
https://github.com/llvm/llvm-project/pull/119820
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
spall wrote:
Waiting on #120630
https://github.com/llvm/llvm-project/pull/119820
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
bogner wrote:
Note: To see the bitcode writer crash, apply this patch and set up
https://github.com/llvm-beanz/HLSLTest - Basic/simple.test hits an assert:
```
Assertion failed: ((Val & ~(~0U >> (32-NumBits))) == 0 && "High bits set!"),
function Emit, file BitstreamWriter.h, line 284.
...
#7 0
bogner wrote:
> Two questions:
>
> 1. Do we foresee any issue adding `afn` (I suspect not)
Adding `afn` should be fine - we don't have library functions per se anyway so
this shouldn't make any difference one way or the other.
> 2. Do we also need to apply `contract`, which I believe was also
llvm-beanz wrote:
Quick sanity check question for @bogner:
In DXC we apply the `fast` flag, which in LLVM 3.7 implies: `nnan ninf nsz
arcp` and includes the modern interpretation of `reassoc`
With this change we are applying: `reassoc nnan ninf nsz arcp afn` (no `fast`
because its meaning is
https://github.com/bogner approved this pull request.
https://github.com/llvm/llvm-project/pull/119820
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/spall edited https://github.com/llvm/llvm-project/pull/119820
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
llvm-beanz wrote:
With how `RUN` lines tend to get copied from test to test, I think it is
probably best to limit how many tests explicitly disable fast math otherwise
they're likely to proliferate across future tests.
https://github.com/llvm/llvm-project/pull/119820
__
spall wrote:
> This looks correct, but how much work is it to update all of these tests to
> have fast math enabled rather than explicitly disabling it? I worry that (1)
> folks looking at these tests in the future will think that disabling fast
> math is a meaningful part of the test and (2)
https://github.com/bogner commented:
This looks correct, but how much work is it to update all of these tests to
have fast math enabled rather than explicitly disabling it? I worry that (1)
folks looking at these tests in the future will think that disabling fast math
is a meaningful part of t
https://github.com/spall edited https://github.com/llvm/llvm-project/pull/119820
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Sarah Spall (spall)
Changes
Make fast math the default for HLSL
Closes #108597
---
Patch is 53.45 KiB, truncated to 20.00 KiB below, full version:
https://github.com/llvm/llvm-project/pull/119820.diff
52 Files Affected:
- (modified)
llvmbot wrote:
@llvm/pr-subscribers-hlsl
Author: Sarah Spall (spall)
Changes
Make fast math the default for HLSL
Closes #108597
---
Patch is 53.45 KiB, truncated to 20.00 KiB below, full version:
https://github.com/llvm/llvm-project/pull/119820.diff
52 Files Affected:
- (modified) c
https://github.com/spall ready_for_review
https://github.com/llvm/llvm-project/pull/119820
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/spall created
https://github.com/llvm/llvm-project/pull/119820
Make fast math the default for HLSL
Closes #108597
>From d167cd92875f7aec8ed3ab15c3321ee9000e8481 Mon Sep 17 00:00:00 2001
From: Sarah Spall
Date: Thu, 12 Dec 2024 20:35:47 +
Subject: [PATCH 1/2] make fast m
16 matches
Mail list logo
