saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-07-17 18:40:28 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Read AA64ISAR0_EL1 to check dot product support (#16082)

Browse source 
### Description

Use an assembly instruction to read the `AA64ISAR0_EL1` register for dot
product support.

### Motivation and Context

The only reliable way to check for supported instruction extensions in
ARM is to
query the instruction set attribute registers. [Dot product instructions
can
be checked using bits 47:44 in the AA64ISAR0_EL1

register](https://developer.arm.com/documentation/ddi0601/2021-12/AArch64-Registers/ID-AA64ISAR0-EL1--AArch64-Instruction-Set-Attribute-Register-0?lang=en#fieldset_0-47_44).

On `qemu-aarch64` with the `a64fx` cpu which does not support the dot
product
instructions, running a quantized BERT-Large (from MLPerf) results in
`SIGILL`.
With the change, the program continues without using the dot product
instructions. Also verified that `S8S8_SDOT` kernels are invoked when
running
on hardware that supports dot product instructions.

---------

Co-authored-by: Skand Hurkat <skhurkat@microsoft.com>
This commit is contained in:
Skand Hurkat 2023-05-25 17:05:30 -07:00 committed by GitHub
parent 0d1a8cc651
commit b28e927ca4
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 4 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
onnxruntime/core/mlas/lib/platform.cpp
View file

@ -55,7 +55,7 @@ MLASCPUIDInfo::MLASCPUIDInfo()
#if defined(BUILD_MLAS_NO_ONNXRUNTIME)
MLASCPUIDInfo::MLASCPUIDInfo()
{
has_arm_neon_dot_ = ((getauxval(AT_HWCAP) & HWCAP_ASIMDDP) != 0);
has_arm_neon_dot_ = ((getauxval(AT_HWCAP) & HWCAP_ASIMDDP) != 0);
// raw hack! Need CPUIDInfo implementation for more precise detection
has_fp16_ = has_arm_neon_dot_;
@ -447,10 +447,10 @@ Return Value:
#if defined(_WIN32)
HasDotProductInstructions = (IsProcessorFeaturePresent(PF_ARM_V82_DP_INSTRUCTIONS_AVAILABLE) != 0);
#elif defined(__linux__)
HasDotProductInstructions = MLAS_CPUIDINFO::GetCPUIDInfo().HasArmNeonDot();
#else
HasDotProductInstructions = false;
uint64_t isar0_el1;
asm("mrs %[reg], ID_AA64ISAR0_EL1\n" : [reg] "=r"(isar0_el1) : :);
HasDotProductInstructions = ((isar0_el1 >> 44) & 0xfu) == 0x1u;
#endif
if (HasDotProductInstructions) {