[all-commits] [llvm/llvm-project] 46ee7f: [libclc] Avoid out-of-range float-to-int. (#145698)
Harald van Dijk via All-commits
all-commits at lists.llvm.org
Wed Jun 25 08:37:28 PDT 2025
Branch: refs/heads/main
Home: https://github.com/llvm/llvm-project
Commit: 46ee7f1908f467d8feb5617e097c88be1401296c
https://github.com/llvm/llvm-project/commit/46ee7f1908f467d8feb5617e097c88be1401296c
Author: Harald van Dijk <harald.vandijk at codeplay.com>
Date: 2025-06-25 (Wed, 25 Jun 2025)
Changed paths:
M libclc/clc/lib/generic/math/clc_pow.inc
M libclc/clc/lib/generic/math/clc_pown.inc
M libclc/clc/lib/generic/math/clc_powr.inc
M libclc/clc/lib/generic/math/clc_rootn.inc
Log Message:
-----------
[libclc] Avoid out-of-range float-to-int. (#145698)
For a kernel such as
kernel void foo(__global double3 *z) {
double3 x = {0.6631661088,0.6612268107,0.1513627528};
int3 y = {-1980459213,-660855407,615708204};
*z = pown(x, y);
}
we were not storing anything to z, because the implementation of pown
relied on an floating-point-to-integer conversion where the
floating-point value was outside of the integer's range. Although in
LLVM IR we permit that operation so long as we end up ignoring its
result -- that is the general rule for poison -- one thing we are not
permitted to do is have conditional branches that depend on it, and
through the call to __clc_ldexp, we did have that.
To fix this, rather than changing expv at the end to INFINITY/0, we can
change v at the start to values that we know will produce INFINITY/0
without performing such out-of-range conversions.
Tested with
clang --target=nvptx64 -S -O3 -o - test.cl \
-Xclang -mlink-builtin-bitcode \
-Xclang runtimes/runtimes-bins/libclc/nvptx64--.bc
A grep showed that this exact same code existed in three more places, so
I changed it there too, though I did not do a broader search for other
similar code that potentially has the same problem.
To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications
More information about the All-commits
mailing list