Fix FMA4 detection#262
Merged
shibatch merged 1 commit intoshibatch:masterfrom May 16, 2019
colesbury:fma4
Merged
Conversation
FMA4 support is in bit 16 of register ECX, not EDX of the "extended processor info" (0x80000001). The mapping of registers to reg is: reg[0] = eax reg[1] = ebx reg[2] = ecx <--- reg[3] = edx Bit 16 of EDX is PAT (Page Attribute Table) on AMD CPUs, which is widely supported. Intel CPUs do not set this bit. This causes "Illegal instruction" errors on AMD CPUs that do not support FMA4. See pytorch/pytorch#12112 See #261 http://developer.amd.com/wordpress/media/2012/10/254811.pdf (Page 20)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
FMA4 support is in bit 16 of register ECX, not EDX of the "extended
processor info" (0x80000001).
The mapping of registers to reg is:
reg[0] = eax
reg[1] = ebx
reg[2] = ecx <---
reg[3] = edx
Bit 16 of EDX is PAT (Page Attribute Table) on AMD CPUs, which is widely
supported. Intel CPUs do not set this bit. This causes "Illegal instruction"
errors on AMD CPUs that do not support FMA4.
See pytorch/pytorch#12112
See #261
http://developer.amd.com/wordpress/media/2012/10/254811.pdf (Page 20)
Fixes #261