Skip to content

fix: pager EOF lockup bug#380

Merged
ethanpailes merged 1 commit into
masterfrom
fix-pager-eof-lockup
Jun 8, 2026
Merged

fix: pager EOF lockup bug#380
ethanpailes merged 1 commit into
masterfrom
fix-pager-eof-lockup

Conversation

@ethanpailes

Copy link
Copy Markdown
Contributor

Issue Link

n/a

AI Policy Ack

Ack

This PR was:

  • mostly or completely vibe coded
  • mostly or completely meat coded
  • bit of both

Meat coded the investigation and fix, vibed a regression test.

Description

51609f8 introduced a regression where shpool would lock up when displaying the message of the day in a pager process like less because it would return an error instead of just terminating the pager display process by returning the current tty size as it ought to have.

This is arguably the first slop bug shpool has had since gemini helped me find the issue that 51609f8 was trying to address, though I wrote the fix that caused the regression. The original CPU pinning EOF bug was real though, it was just my fix that was bad.

I yanked versions 0.10.0 an 0.10.1 over this issue, which might be too extreme now that I have discovered the problem (I think I might be the only person who uses this feature), but at the time I didn't have the cycles to investigate.

@ethanpailes ethanpailes requested a review from maxhbooth June 8, 2026 16:24
@ethanpailes ethanpailes force-pushed the fix-pager-eof-lockup branch 2 times, most recently from 788816c to 660bdcf Compare June 8, 2026 17:30
51609f8 introduced a regression where shpool would lock up
when displaying the message of the day in a pager process like
less because it would return an error instead of just terminating
the pager display process by returning the current tty size as it
ought to have.

This is arguably the first slop bug shpool has had since gemini
helped me find the issue that 51609f8 was trying to address,
though I wrote the fix that caused the regression. The original
CPU pinning EOF bug was real though, it was just my fix that
was bad.

I yanked versions 0.10.0 an 0.10.1 over this issue, which might
be too extreme now that I have discovered the problem (I think I
might be the only person who uses this feature), but at the time
I didn't have the cycles to investigate.

There are some unrelated test tweaks to deflake some tests.
@ethanpailes ethanpailes force-pushed the fix-pager-eof-lockup branch from 660bdcf to 73db3e9 Compare June 8, 2026 17:37
@ethanpailes ethanpailes merged commit 3703dfb into master Jun 8, 2026
11 checks passed
@ethanpailes ethanpailes deleted the fix-pager-eof-lockup branch June 8, 2026 18:27
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