{"thread":{"id":"66276","subject":"[PATCH] ci: bump debian-11 job to debian-12","startedAt":"2026-09-05T13:58:24Z","lastAt":"2026-09-11T15:42:40Z","messageCount":8,"participants":["Jeff King","Junio C Hamano","Patrick Steinhardt"],"isPatch":true,"patchVersion":1,"patchTotal":null},"messages":[{"id":"552029","messageId":"20260905135822.GA3914811@coredump.intra.peff.net","threadId":"66276","inReplyTo":null,"subject":"[PATCH] ci: bump debian-11 job to debian-12","fromName":"Jeff King","fromEmail":"peff@peff.net","sentAt":"2026-09-05T13:58:22Z","receivedAt":"2026-09-05T13:58:24Z","isPatch":true,"body":"Debian 11 just recently went out of its LTS period, and is unmaintained\nby the project (there is \"Extended LTS\", but it is a paid service\nprovided by a third party).\n\nThe point of the debian-11 job was to cover older releases in the LTS\nstate, per ac112fd4f0 (Add additional CI jobs to avoid accidental\nbreakage, 2024-10-31). Bumping to debian-12 will cover us there for the\nnext 2 years.\n\nSigned-off-by: Jeff King <peff@peff.net>\n---\nI started looking at this because I got an apt failure on a debian-11\njob today. It might have just been a transient mirror failure (although\nit reproduced for several minutes afterwards). But this seems like the\nright step forward anyway.\n\nI tested the GitHub job, but not the GitLab one. They should be pulling\nthe same docker images, though, so I would expect it to Just Work.\n\n .github/workflows/main.yml | 6 +++---\n .gitlab-ci.yml             | 6 +++---\n 2 files changed, 6 insertions(+), 6 deletions(-)\n\ndiff --git a/.github/workflows/main.yml b/.github/workflows/main.yml\nindex 205325eb33..a0c3f53c6d 100644\n--- a/.github/workflows/main.yml\n+++ b/.github/workflows/main.yml\n@@ -411,9 +411,9 @@ jobs:\n         # A RHEL 8 compatible distro.  Supported until 2029-05-31.\n         - jobname: almalinux-8\n           image: almalinux:8\n-        # Supported until 2026-08-31.\n-        - jobname: debian-11\n-          image: debian:11\n+        # Supported until 2028-06-30.\n+        - jobname: debian-12\n+          image: debian:12\n     env:\n       jobname: ${{matrix.vector.jobname}}\n       CC: ${{matrix.vector.cc}}\ndiff --git a/.gitlab-ci.yml b/.gitlab-ci.yml\nindex 0242283c3c..cd6fd4a504 100644\n--- a/.gitlab-ci.yml\n+++ b/.gitlab-ci.yml\n@@ -71,9 +71,9 @@ test:linux:\n       # A RHEL 8 compatible distro.  Supported until 2029-05-31.\n       - jobname: almalinux-8\n         image: almalinux:8\n-      # Supported until 2026-08-31.\n-      - jobname: debian-11\n-        image: debian:11\n+      # Supported until 2028-06-30.\n+      - jobname: debian-12\n+        image: debian:12\n   artifacts:\n     paths:\n       - t/failed-test-artifacts\n-- \n2.55.0.1127.g25100ff258\n"},{"id":"552037","messageId":"xmqqecf764lu.fsf@gitster.g","threadId":"66276","inReplyTo":"20260905135822.GA3914811@coredump.intra.peff.net","subject":"Re: [PATCH] ci: bump debian-11 job to debian-12","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2026-09-05T16:32:45Z","receivedAt":"2026-09-05T16:32:48Z","isPatch":true,"body":"Jeff King <peff@peff.net> writes:\n\n> Debian 11 just recently went out of its LTS period, and is unmaintained\n> by the project (there is \"Extended LTS\", but it is a paid service\n> provided by a third party).\n>\n> The point of the debian-11 job was to cover older releases in the LTS\n> state, per ac112fd4f0 (Add additional CI jobs to avoid accidental\n> breakage, 2024-10-31). Bumping to debian-12 will cover us there for the\n> next 2 years.\n>\n> Signed-off-by: Jeff King <peff@peff.net>\n> ---\n> I started looking at this because I got an apt failure on a debian-11\n> job today. It might have just been a transient mirror failure (although\n> it reproduced for several minutes afterwards). But this seems like the\n> right step forward anyway.\n>\n> I tested the GitHub job, but not the GitLab one. They should be pulling\n> the same docker images, though, so I would expect it to Just Work.\n\nThanks.  I also noticed debian-11 job failing.  Will queue.\n\n\n\n>\n>  .github/workflows/main.yml | 6 +++---\n>  .gitlab-ci.yml             | 6 +++---\n>  2 files changed, 6 insertions(+), 6 deletions(-)\n>\n> diff --git a/.github/workflows/main.yml b/.github/workflows/main.yml\n> index 205325eb33..a0c3f53c6d 100644\n> --- a/.github/workflows/main.yml\n> +++ b/.github/workflows/main.yml\n> @@ -411,9 +411,9 @@ jobs:\n>          # A RHEL 8 compatible distro.  Supported until 2029-05-31.\n>          - jobname: almalinux-8\n>            image: almalinux:8\n> -        # Supported until 2026-08-31.\n> -        - jobname: debian-11\n> -          image: debian:11\n> +        # Supported until 2028-06-30.\n> +        - jobname: debian-12\n> +          image: debian:12\n>      env:\n>        jobname: ${{matrix.vector.jobname}}\n>        CC: ${{matrix.vector.cc}}\n> diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml\n> index 0242283c3c..cd6fd4a504 100644\n> --- a/.gitlab-ci.yml\n> +++ b/.gitlab-ci.yml\n> @@ -71,9 +71,9 @@ test:linux:\n>        # A RHEL 8 compatible distro.  Supported until 2029-05-31.\n>        - jobname: almalinux-8\n>          image: almalinux:8\n> -      # Supported until 2026-08-31.\n> -      - jobname: debian-11\n> -        image: debian:11\n> +      # Supported until 2028-06-30.\n> +      - jobname: debian-12\n> +        image: debian:12\n>    artifacts:\n>      paths:\n>        - t/failed-test-artifacts\n"},{"id":"552074","messageId":"20260906151137.GA328152@coredump.intra.peff.net","threadId":"66276","inReplyTo":"20260905135822.GA3914811@coredump.intra.peff.net","subject":"Re: [PATCH] ci: bump debian-11 job to debian-12","fromName":"Jeff King","fromEmail":"peff@peff.net","sentAt":"2026-09-06T15:11:37Z","receivedAt":"2026-09-06T15:11:45Z","isPatch":true,"body":"On Sat, Sep 05, 2026 at 09:58:22AM -0400, Jeff King wrote:\n\n> I started looking at this because I got an apt failure on a debian-11\n> job today. It might have just been a transient mirror failure (although\n> it reproduced for several minutes afterwards). But this seems like the\n> right step forward anyway.\n\nBTW, I noticed that the linux32 build is using ubuntu 20.04, which has\nbeen out of LTS for a year. But bumping isn't really an option; they\ndropped i386 platform support, and so has Debian.\n\nI'm mostly inclined to leave it unless/until it starts creating\nheadaches. To some degree, if we cannot even find an image to test\nagain, it might not be an important enough platform to care about. But I\ncan also imagine there is a long tail of oddball 32-bit platforms that\nGit does run on (like small ARM chips), and it's nice to at least have\nsome coverage. Possibly there's an ARM image we could use (looks like\narmhf?).\n\nWe also seem to use 20.04 for linux-TEST-vars. On the surface there's no\nreason it couldn't be using ubuntu-latest, though I think this may be\none of those cases where it's doing double duty as \"test exotic configs\"\nand \"test on an older platform\". But might be worth bumping to the\noldest in-scope LTS.\n\nAll out of scope for this patch, and mostly I'm inclined to ignore it\nfor now until we hit problems (and then decide if it's worth\naccommodating or if old systems are too old).\n\n-Peff\n"},{"id":"552088","messageId":"ap5TrxRXrYZzwMN2@pks.im","threadId":"66276","inReplyTo":"20260906151137.GA328152@coredump.intra.peff.net","subject":"Re: [PATCH] ci: bump debian-11 job to debian-12","fromName":"Patrick Steinhardt","fromEmail":"ps@pks.im","sentAt":"2026-09-07T06:03:27Z","receivedAt":"2026-09-07T06:03:34Z","isPatch":true,"body":"On Sun, Sep 06, 2026 at 11:11:37AM -0400, Jeff King wrote:\n> On Sat, Sep 05, 2026 at 09:58:22AM -0400, Jeff King wrote:\n> \n> > I started looking at this because I got an apt failure on a debian-11\n> > job today. It might have just been a transient mirror failure (although\n> > it reproduced for several minutes afterwards). But this seems like the\n> > right step forward anyway.\n> \n> BTW, I noticed that the linux32 build is using ubuntu 20.04, which has\n> been out of LTS for a year. But bumping isn't really an option; they\n> dropped i386 platform support, and so has Debian.\n> \n> I'm mostly inclined to leave it unless/until it starts creating\n> headaches. To some degree, if we cannot even find an image to test\n> again, it might not be an important enough platform to care about. But I\n> can also imagine there is a long tail of oddball 32-bit platforms that\n> Git does run on (like small ARM chips), and it's nice to at least have\n> some coverage. Possibly there's an ARM image we could use (looks like\n> armhf?).\n\nI agree that in the best case we'd find an alternative that works on 32\nbit. But as you say, the list of x86 distros that still have support for\n32 bit is shrinking quite a bit these days.\n\nOne of the bigger distros that still supports it properly is Alpine\nLinux. We already support it via linux-musl-meson, so it should be easy\nenough to extend. In theory, we could even merge these two jobs into a\nsingle job. We'd of course lose 32 bit glibc support, but that might be\nfine.\n\n> We also seem to use 20.04 for linux-TEST-vars. On the surface there's no\n> reason it couldn't be using ubuntu-latest, though I think this may be\n> one of those cases where it's doing double duty as \"test exotic configs\"\n> and \"test on an older platform\". But might be worth bumping to the\n> oldest in-scope LTS.\n\nYeah, might be worth it. It would've been great if they had a tag for\nthis so we odn't have to bump ourselves, but on the other hand it's only\na small bit of churn.\n\nPatrick\n"},{"id":"552089","messageId":"ap5Ttt-2NmM5dRZl@pks.im","threadId":"66276","inReplyTo":"20260905135822.GA3914811@coredump.intra.peff.net","subject":"Re: [PATCH] ci: bump debian-11 job to debian-12","fromName":"Patrick Steinhardt","fromEmail":"ps@pks.im","sentAt":"2026-09-07T06:03:34Z","receivedAt":"2026-09-07T06:03:38Z","isPatch":true,"body":"On Sat, Sep 05, 2026 at 09:58:22AM -0400, Jeff King wrote:\n> Debian 11 just recently went out of its LTS period, and is unmaintained\n> by the project (there is \"Extended LTS\", but it is a paid service\n> provided by a third party).\n> \n> The point of the debian-11 job was to cover older releases in the LTS\n> state, per ac112fd4f0 (Add additional CI jobs to avoid accidental\n> breakage, 2024-10-31). Bumping to debian-12 will cover us there for the\n> next 2 years.\n> \n> Signed-off-by: Jeff King <peff@peff.net>\n> ---\n> I started looking at this because I got an apt failure on a debian-11\n> job today. It might have just been a transient mirror failure (although\n> it reproduced for several minutes afterwards). But this seems like the\n> right step forward anyway.\n\nI noticed that the Debian jobs became flaky lately, too.\n\n> I tested the GitHub job, but not the GitLab one. They should be pulling\n> the same docker images, though, so I would expect it to Just Work.\n\nYeah, same.\n\n> diff --git a/.github/workflows/main.yml b/.github/workflows/main.yml\n> index 205325eb33..a0c3f53c6d 100644\n> --- a/.github/workflows/main.yml\n> +++ b/.github/workflows/main.yml\n> @@ -411,9 +411,9 @@ jobs:\n>          # A RHEL 8 compatible distro.  Supported until 2029-05-31.\n>          - jobname: almalinux-8\n>            image: almalinux:8\n> -        # Supported until 2026-08-31.\n> -        - jobname: debian-11\n> -          image: debian:11\n> +        # Supported until 2028-06-30.\n> +        - jobname: debian-12\n> +          image: debian:12\n>      env:\n>        jobname: ${{matrix.vector.jobname}}\n>        CC: ${{matrix.vector.cc}}\n> diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml\n> index 0242283c3c..cd6fd4a504 100644\n> --- a/.gitlab-ci.yml\n> +++ b/.gitlab-ci.yml\n> @@ -71,9 +71,9 @@ test:linux:\n>        # A RHEL 8 compatible distro.  Supported until 2029-05-31.\n>        - jobname: almalinux-8\n>          image: almalinux:8\n> -      # Supported until 2026-08-31.\n> -      - jobname: debian-11\n> -        image: debian:11\n> +      # Supported until 2028-06-30.\n> +      - jobname: debian-12\n> +        image: debian:12\n>    artifacts:\n>      paths:\n>        - t/failed-test-artifacts\n\nThe changes look good, but I wonder whether we want to swap to the\n\"oldstable\" tag instead. That points to debian-12, too, and might result\nin less churn going forward.\n\nPatrick\n"},{"id":"552496","messageId":"20260911021933.GA948291@coredump.intra.peff.net","threadId":"66276","inReplyTo":"ap5Ttt-2NmM5dRZl@pks.im","subject":"Re: [PATCH] ci: bump debian-11 job to debian-12","fromName":"Jeff King","fromEmail":"peff@peff.net","sentAt":"2026-09-11T02:19:33Z","receivedAt":"2026-09-11T02:19:35Z","isPatch":true,"body":"On Mon, Sep 07, 2026 at 08:03:34AM +0200, Patrick Steinhardt wrote:\n\n> > +      # Supported until 2028-06-30.\n> > +      - jobname: debian-12\n> > +        image: debian:12\n> >    artifacts:\n> >      paths:\n> >        - t/failed-test-artifacts\n> \n> The changes look good, but I wonder whether we want to swap to the\n> \"oldstable\" tag instead. That points to debian-12, too, and might result\n> in less churn going forward.\n\nI was all ready to agree and propose a patch, but I think it might be\nmore complicated. The \"old\" flags flip when a new stable version is\nreleased, but what we really care about is things leaving LTS. And the\ntwo are not really aligned.\n\nSo if we wanted the oldest supported Debian release, that would have\nbeen \"oldoldstable\" (debian-11) until August 31st of this year. But\nafter that, we want \"oldstable\" (debian-12), which is now the oldest\nthing under LTS. And that will be true until probably next summer, when\ndebian-14 is released, at which point we'd want to stay on it by\nswitching to \"oldoldstable\" again for another year.\n\nThe offset comes from the fact that Debian releases approximately every\n2 years, but with a 5-year LTS, and they never align.\n\nIf we track oldstable, then every other year we're not using the oldest\nsupported release (because the prior release is still in LTS). Or we\ntrack oldoldstable, which means every other year we're using a release\nthat's no longer supported. Or we flip-flop between them once a year,\nwhen oldoldstable goes out of support.\n\nWhich makes me inclined to just keep the current scheme: use the numeric\ncodes and just bump them every 2 years or so when they go out of\nsupport. And we can even put it off until the out-of-support release\nstarts causing problems like it did here.\n\n-Peff\n"},{"id":"552498","messageId":"aqOND8cPTRtV1Io8@pks.im","threadId":"66276","inReplyTo":"20260911021933.GA948291@coredump.intra.peff.net","subject":"Re: [PATCH] ci: bump debian-11 job to debian-12","fromName":"Patrick Steinhardt","fromEmail":"ps@pks.im","sentAt":"2026-09-11T05:09:35Z","receivedAt":"2026-09-11T05:09:43Z","isPatch":true,"body":"On Thu, Sep 10, 2026 at 10:19:33PM -0400, Jeff King wrote:\n> On Mon, Sep 07, 2026 at 08:03:34AM +0200, Patrick Steinhardt wrote:\n> \n> > > +      # Supported until 2028-06-30.\n> > > +      - jobname: debian-12\n> > > +        image: debian:12\n> > >    artifacts:\n> > >      paths:\n> > >        - t/failed-test-artifacts\n> > \n> > The changes look good, but I wonder whether we want to swap to the\n> > \"oldstable\" tag instead. That points to debian-12, too, and might result\n> > in less churn going forward.\n> \n> I was all ready to agree and propose a patch, but I think it might be\n> more complicated. The \"old\" flags flip when a new stable version is\n> released, but what we really care about is things leaving LTS. And the\n> two are not really aligned.\n> \n> So if we wanted the oldest supported Debian release, that would have\n> been \"oldoldstable\" (debian-11) until August 31st of this year. But\n> after that, we want \"oldstable\" (debian-12), which is now the oldest\n> thing under LTS. And that will be true until probably next summer, when\n> debian-14 is released, at which point we'd want to stay on it by\n> switching to \"oldoldstable\" again for another year.\n> \n> The offset comes from the fact that Debian releases approximately every\n> 2 years, but with a 5-year LTS, and they never align.\n> \n> If we track oldstable, then every other year we're not using the oldest\n> supported release (because the prior release is still in LTS). Or we\n> track oldoldstable, which means every other year we're using a release\n> that's no longer supported. Or we flip-flop between them once a year,\n> when oldoldstable goes out of support.\n> \n> Which makes me inclined to just keep the current scheme: use the numeric\n> codes and just bump them every 2 years or so when they go out of\n> support. And we can even put it off until the out-of-support release\n> starts causing problems like it did here.\n\nSounds good, let's stick with the current patch then. Thanks!\n\nPatrick\n"},{"id":"552564","messageId":"xmqqpkyjdcb6.fsf@gitster.g","threadId":"66276","inReplyTo":"20260911021933.GA948291@coredump.intra.peff.net","subject":"Re: [PATCH] ci: bump debian-11 job to debian-12","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2026-09-11T15:42:37Z","receivedAt":"2026-09-11T15:42:40Z","isPatch":true,"body":"Jeff King <peff@peff.net> writes:\n\n> If we track oldstable, then every other year we're not using the oldest\n> supported release (because the prior release is still in LTS). Or we\n> track oldoldstable, which means every other year we're using a release\n> that's no longer supported. Or we flip-flop between them once a year,\n> when oldoldstable goes out of support.\n>\n> Which makes me inclined to just keep the current scheme: use the numeric\n> codes and just bump them every 2 years or so when they go out of\n> support. And we can even put it off until the out-of-support release\n> starts causing problems like it did here.\n\nSounds sensible to me.  Thanks for thinking this through.\n"}]}