Skip to content

Update pytorch/cpuinfo to 69e20fa2 - #32432

Closed
maxim-davgalev wants to merge 1 commit into
microsoft:mainfrom
maxim-davgalev:cpuinfo-bump
Closed

Update pytorch/cpuinfo to 69e20fa2#32432
maxim-davgalev wants to merge 1 commit into
microsoft:mainfrom
maxim-davgalev:cpuinfo-bump

Conversation

@maxim-davgalev

Copy link
Copy Markdown

Description

Update the pytorch/cpuinfo pin by ten commits, from 4628dc06 (2026-06-12) to 69e20fa2 (2026-07-07). This includes fixes and hardware detection updates relevant to ORT:

The remaining four commits contain smaller cleanups (pytorch/cpuinfo#272, pytorch/cpuinfo#391, pytorch/cpuinfo#401, and pytorch/cpuinfo#410).

69e20fa2 is the last commit before pytorch/cpuinfo#411, which removes the cpuinfo_deinitialize() API that ORT began using in #28245. ORT currently calls it from cpuid_info.cc and posix/env.cc. Updating past 69e20fa2 requires replacing those calls, which is outside the scope of this update.

The three local patches in cmake/patches/cpuinfo apply cleanly to the new source tree.

Update the pytorch/cpuinfo pin by ten commits, from 4628dc06
(2026-06-12) to 69e20fa2 (2026-07-07). This includes fixes and
hardware detection updates relevant to ORT:

- Initialize the vendor/uarch out-parameters in
  cpuinfo_arm_decode_vendor_uarch (pytorch/cpuinfo#395)
- Zero-initialize stack buffers in the ARM/Linux parsers to fix
  Valgrind warnings (pytorch/cpuinfo#376)
- Fix an undeclared syscall() error under stricter build options
  (pytorch/cpuinfo#374)
- Add Linux MIDR detection for Apple M1/M2/M3/M4
  (pytorch/cpuinfo#385)
- Add Nova Lake (Coyote Cove) uarch support
  (pytorch/cpuinfo#393)
- Normalize AMD brand strings containing the "with Radeon Graphics"
  suffix (pytorch/cpuinfo#389)

The remaining four commits contain smaller cleanups
(pytorch/cpuinfo#272, microsoft#391, microsoft#401, and microsoft#410).

69e20fa2 is the last commit before pytorch/cpuinfo#411, which removes
the cpuinfo_deinitialize() API that ORT began using in microsoft#28245. ORT
currently calls it from cpuid_info.cc and posix/env.cc. Updating past
69e20fa2 requires replacing those calls, which is outside the scope
of this update.

The three local patches in cmake/patches/cpuinfo apply cleanly to the
new source tree.
Copilot AI balanced review requested due to automatic review settings September 3, 2026 23:10

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🟡 Changes recommended

The vcpkg overlay remains pinned to the old cpuinfo revision.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Pull request overview

Updates the cpuinfo FetchContent pin to include newer hardware detection and correctness fixes.

Changes:

  • Bumps cpuinfo from 4628dc06 to 69e20fa2.
  • Updates the archive SHA1.
File summaries
File Description
cmake/deps.txt Updates the cpuinfo source revision and checksum.
Review details
  • Files reviewed: 1/1 changed files
  • Comments generated: 1
  • Review effort level: Balanced

💡 Configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Comment thread cmake/deps.txt
pthreadpool;https://github.com/google/pthreadpool/archive/dcc9f28589066af0dbd4555579281230abbf74dd.zip;533a77943203ef15ca608bcd9dbe2c94da7451d2
pybind11;https://github.com/pybind/pybind11/archive/refs/tags/v3.0.2.zip;a064e663b4d7a337ac291d1bef7337ef4e60a1ae
pytorch_cpuinfo;https://github.com/pytorch/cpuinfo/archive/4628dc060ce4e82345dc166bbac875609db4ff69.zip;e58d4b47c16a982111c897e669ae4f1821a393d7
pytorch_cpuinfo;https://github.com/pytorch/cpuinfo/archive/69e20fa2bb5e4d6950e431d09db028a21c8d9487.zip;7a4ea1f2bc55e4888d52079c6213209b03a8b7b4
@maxim-davgalev

Copy link
Copy Markdown
Author

superseded by #32300

@maxim-davgalev
maxim-davgalev deleted the cpuinfo-bump branch September 3, 2026 23:24
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants