Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[3.7] bpo-40791: Make compare_digest more constant-time. (GH-20444) #23438

Merged
merged 1 commit into from Nov 22, 2020

Conversation

miss-islington
Copy link
Contributor

@miss-islington miss-islington commented Nov 21, 2020

  • bpo-40791: Make compare_digest more constant-time.

The existing volatile left/right pointers guarantee that the reads will all occur, but does not guarantee that they will be used. So a compiler can still short-circuit the loop, saving e.g. the overhead of doing the xors and especially the overhead of the data dependency between result and the reads. That would change performance depending on where the first unequal byte occurs. This change removes that optimization.

(This is change GH-1 from https://bugs.python.org/issue40791 .)
(cherry picked from commit 3172936)

Co-authored-by: Devin Jeanpierre jeanpierreda@google.com

https://bugs.python.org/issue40791

* bpo-40791: Make compare_digest more constant-time.

The existing volatile `left`/`right` pointers guarantee that the reads will all occur, but does not guarantee that they will be _used_. So a compiler can still short-circuit the loop, saving e.g. the overhead of doing the xors and especially the overhead of the data dependency between `result` and the reads. That would change performance depending on where the first unequal byte occurs. This change removes that optimization.

(This is change GH-1 from https://bugs.python.org/issue40791 .)
(cherry picked from commit 3172936)

Co-authored-by: Devin Jeanpierre <jeanpierreda@google.com>
@bedevere-bot bedevere-bot added type-bug type-security labels Nov 21, 2020
@miss-islington
Copy link
Contributor Author

@miss-islington miss-islington commented Nov 21, 2020

@ssbr and @gpshead: Status check is done, and it's a success .

@miss-islington
Copy link
Contributor Author

@miss-islington miss-islington commented Nov 21, 2020

@ssbr and @gpshead: Status check is done, and it's a success .

@miss-islington
Copy link
Contributor Author

@miss-islington miss-islington commented Nov 22, 2020

Sorry, I can't merge this PR. Reason: You're not authorized to push to this branch. Visit https://docs.github.com/articles/about-protected-branches/ for more information..

@benjaminp benjaminp merged commit db95802 into python:3.7 Nov 22, 2020
11 checks passed
@miss-islington miss-islington deleted the backport-3172936-3.7 branch Nov 22, 2020
gentoo-bot pushed a commit to gentoo/cpython that referenced this issue Dec 14, 2020
The existing volatile `left`/`right` pointers guarantee that the reads will all occur, but does not guarantee that they will be _used_. So a compiler can still short-circuit the loop, saving e.g. the overhead of doing the xors and especially the overhead of the data dependency between `result` and the reads. That would change performance depending on where the first unequal byte occurs. This change removes that optimization.

(This is change GH-1 from https://bugs.python.org/issue40791 .)
(cherry picked from commit 3172936)

Co-authored-by: Devin Jeanpierre <jeanpierreda@google.com>

Rebased for Python 2.7 by Michał Górny <mgorny@gentoo.org>
@miss-islington
Copy link
Contributor Author

@miss-islington miss-islington commented Dec 14, 2020

Thanks @miss-islington for the PR, and @benjaminp for merging it 🌮🎉.. I'm working now to backport this PR to: 3.6.
🐍🍒🤖

@bedevere-bot
Copy link

@bedevere-bot bedevere-bot commented Dec 14, 2020

GH-23767 is a backport of this pull request to the 3.6 branch.

miss-islington added a commit to miss-islington/cpython that referenced this issue Dec 14, 2020
The existing volatile `left`/`right` pointers guarantee that the reads will all occur, but does not guarantee that they will be _used_. So a compiler can still short-circuit the loop, saving e.g. the overhead of doing the xors and especially the overhead of the data dependency between `result` and the reads. That would change performance depending on where the first unequal byte occurs. This change removes that optimization.

(This is change GH-1 from https://bugs.python.org/issue40791 .)
(cherry picked from commit 3172936)

Co-authored-by: Devin Jeanpierre <jeanpierreda@google.com>
(cherry picked from commit db95802)

Co-authored-by: Miss Islington (bot) <31488909+miss-islington@users.noreply.github.com>
ned-deily pushed a commit that referenced this issue Dec 14, 2020
The existing volatile `left`/`right` pointers guarantee that the reads will all occur, but does not guarantee that they will be _used_. So a compiler can still short-circuit the loop, saving e.g. the overhead of doing the xors and especially the overhead of the data dependency between `result` and the reads. That would change performance depending on where the first unequal byte occurs. This change removes that optimization.

(This is change GH-1 from https://bugs.python.org/issue40791 .)
(cherry picked from commit 3172936)

Co-authored-by: Devin Jeanpierre <jeanpierreda@google.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CLA signed type-bug type-security
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

6 participants