simon-git: putty (main): Simon Tatham

[Commits to Tartarus hosted VCS]

TL;DR:
  07caba27 Buildscr.cv: stop using 32-bit Coverity.
  aaa5fc51 Remove pointless check in timing.c.
  f27bf441 Post-0.84 release checklist updates.

Repository:     https://git.tartarus.org/simon/putty.git
On the web:     https://git.tartarus.org/?p=simon/putty.git
Branch updated: main
Committer:      Simon Tatham <anakin at pobox.com>
Date:           2026-05-22 12:01:25

commit 07caba277d9a8489a438e81af55d95ab08ea34f1
web diff https://git.tartarus.org/?p=simon/putty.git;a=commitdiff;h=07caba277d9a8489a438e81af55d95ab08ea34f1;hp=47b1abdd1a053914d6cbaa5c810912302ce54654
Author: Simon Tatham <anakin at pobox.com>
Date:   Thu Apr 16 16:37:38 2026 +0100

    Buildscr.cv: stop using 32-bit Coverity.
    
    It was useful to do one 32-bit and one 64-bit build, so as to get some
    checks that apply in each case. But Coverity has updated its tools and
    now doesn't support 32-bit any more at all, so building both Linux and
    Windows 64-bit is the best we can manage.
    
    However, even that isn't very useful, and I may stop running Coverity
    entirely, because in the most recent update, it seems to have lost the
    ability to understand that strbuf_free() frees a strbuf, so it reports
    hundreds of resource leaks anywhere a strbuf is used at all, because
    from its point of view, nothing ever frees one. That's too much
    nonsense to wade through!

 Buildscr.cv | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

commit aaa5fc51cc1aa4302f32633c1e739b620d9ae986
web diff https://git.tartarus.org/?p=simon/putty.git;a=commitdiff;h=aaa5fc51cc1aa4302f32633c1e739b620d9ae986;hp=07caba277d9a8489a438e81af55d95ab08ea34f1
Author: Simon Tatham <anakin at pobox.com>
Date:   Thu Apr 16 16:40:43 2026 +0100

    Remove pointless check in timing.c.
    
    Coverity spotted this (as one of the few reports that was actually
    useful this time round): when our ticks type was a signed long, it
    made sense (if you don't count integer overflow UB) to check if
    when - now <= 0, but now that it's an unsigned long, it doesn't.
    
    We've been using unsigned long for some releases now, and nobody has
    reported any weird behaviour arising from this check not doing
    anything. And it was always a last-ditch defence after several more
    sensible things have failed. So I think it can just be removed, rather
    than trying to fix it.

 timing.c | 8 --------
 1 file changed, 8 deletions(-)

commit f27bf4415996ee4f707f7dcae84f0eadfaa07757
web diff https://git.tartarus.org/?p=simon/putty.git;a=commitdiff;h=f27bf4415996ee4f707f7dcae84f0eadfaa07757;hp=aaa5fc51cc1aa4302f32633c1e739b620d9ae986
Author: Simon Tatham <anakin at pobox.com>
Date:   Fri May 22 11:55:39 2026 +0100

    Post-0.84 release checklist updates.
    
    I had a better idea about how to find the URLs for Windows Store
    submission, which doesn't require manually modifying anything. Also,
    my first attempt to send the putty-announce mailing list message was
    rejected by my own mail filter, because I'd forgotten that I'd put a
    safety catch in front of it to block spam attempts.
    
    I'm also considering removing Coverity completely from the list: their
    latest version seems to have added a new checker which has a huge
    number of false positives. But they've been good in the past, so I'll
    give them one more chance and see if things have improved next time.

 CHECKLST.txt | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)