8 Commits

Author SHA1 Message Date
Arthur Eubanks
0d948827d7
[Instrumentation][X86] Limit setting large section flag to medium/large code models (#75542)
In #74514 and #74778 we marked various instrumentation-added sections as
large. This causes an extra PT_LOAD segment if using the small code
model. Since people using the small code model presumably aren't hitting
relocation limits, disable this when using the small code model to avoid
the extra segment.

This uses Module::getCodeModel() which isn't necessarily reliable since
it reads module metadata (which right now only the clang frontend sets),
but it would be nice to get to a point where we reliably put this sort
of information (e.g. PIC/code model/etc) in the IR. This requires
duplicating the existing tests since opt/llc currently don't set these
metadata. If we get to a point where they do set the code model metadata
based on command line arguments then we can deduplicate these tests.
2023-12-15 09:46:59 -08:00
Arthur Eubanks
66b919cb29 Reland [InstrProf][X86] Mark non-directly accessed globals as large (#74778)
We'd like to make various instrprof globals large to make them not
contribute to relocation pressure since there are no direct accesses
to them in the module.

Similar to what was done for asan_globals in #74514.

This affects the __llvm_prf_vals, __llvm_prf_vnds, and __llvm_prf_names
sections.

The reland fixes platform.ll.
2023-12-08 09:54:57 -08:00
Arthur Eubanks
96a5135e56 Revert "[InstrProf][X86] Mark non-directly accessed globals as large (#74778)"
This reverts commit 5507f70cc205a7ec21d264a64c703b3d314b998c.

Breaks bots, e.g. https://lab.llvm.org/buildbot/#/builders/232/builds/16374
2023-12-08 09:41:31 -08:00
Arthur Eubanks
5507f70cc2
[InstrProf][X86] Mark non-directly accessed globals as large (#74778)
We'd like to make various instrprof globals large to make them not
contribute to relocation pressure since there are no direct accesses
to them in the module.

Similar to what was done for asan_globals in #74514.

This affects the __llvm_prf_vals, __llvm_prf_vnds, and __llvm_prf_names
sections.
2023-12-08 09:33:40 -08:00
Qiongsi Wu
401f768061 [PGO] Setting ValueProfNode Array's Alignment
`instrprof` currently does not set `__llvm_prf_vnds`'s alignment after creating it. The consequence is that the alignment is set to 16 later (c0f3ac1d00/llvm/lib/IR/DataLayout.cpp (L1019)). This can lead to undefined behaviour when we calculate `NumVNodes` in `lprofGetLoadModuleSignature` (c0f3ac1d00/compiler-rt/lib/profile/InstrProfilingMerge.c (L32)). The reason is that when the `__llvm_prf_vnds` array is 16 byte aligned, `__llvm_profile_end_vnodes() - __llvm_profile_begin_vnodes()` may not be a multiple of the size of ValueProfNode (which is 24, 20 on 32 bit targets).

This patch sets `__llvm_prf_vnds`'s alignment to its ABI alignment, which always divides its size. Then `__llvm_profile_end_vnodes() - __llvm_profile_begin_vnodes()` will be a multiple of `sizeof(ValueProfNode)`.

Reviewed By: w2yehia, MaskRay

Differential Revision: https://reviews.llvm.org/D144302
2023-02-23 19:07:29 -05:00
Fangrui Song
9ff36df5a4 [PGO][InstrProf][test] Convert to opaque pointers and fix some stale tests 2022-07-16 11:42:35 -07:00
Fangrui Song
acead46a10 [test] Migrate legacy InstrProfiling tests to new PM 2022-07-16 01:10:08 -07:00
Fangrui Song
2f7ea054f3 [test] Split icall.ll into comdat/nocomdat variants
darwin/aix don't support comdat. Using IR comdat is incorrect.
2021-08-19 16:36:59 -07:00