GNU bug report logs - #78837
[PATCH ruby-team 0/3] Almost there!

Previous Next

To reply to this bug, email your comments to 78837 AT debbugs.gnu.org.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Nicolas Graves <ngraves <at> ngraves.fr>
To: guix-patches <at> gnu.org
Cc: Nicolas Graves <ngraves <at> ngraves.fr>
Subject: [PATCH ruby-team 0/3] Almost there!
Date: Thu, 19 Jun 2025 10:32:21 +0200

Now that I don't have locally failing tests, let's try and address tests that fail on CI.

Nicolas Graves (3):
  gnu: ruby-puma: Skip additional failing test.
  gnu: ruby-minitest-stub-const: Add minitest native-input.
  gnu: ruby-spy: Skip problematic test.

 gnu/packages/ruby-xyz.scm | 9 +++++++++
 1 file changed, 9 insertions(+)

-- 
2.49.0

Message #8 received at 78837 <at> debbugs.gnu.org (full text, mbox):

From: Nicolas Graves <ngraves <at> ngraves.fr>
To: 78837 <at> debbugs.gnu.org
Cc: Nicolas Graves <ngraves <at> ngraves.fr>
Subject: [PATCH 1/3] gnu: ruby-puma: Skip additional failing test.
Date: Thu, 19 Jun 2025 10:39:06 +0200

* gnu/packages/ruby.scm (ruby-puma)[arguments]{phases}: Skip
additional test in phase 'disable-problematic-tests.
---
 gnu/packages/ruby-xyz.scm | 1 +
 1 file changed, 1 insertion(+)

diff --git a/gnu/packages/ruby-xyz.scm b/gnu/packages/ruby-xyz.scm
index d727dd06a3..9c3fb4b167 100644
--- a/gnu/packages/ruby-xyz.scm
+++ b/gnu/packages/ruby-xyz.scm
@@ -11714,6 +11714,7 @@ (define-public ruby-puma
                 (skip-tests "test/test_web_concurrency_auto.rb" "\
 test_web_concurrency_with_concurrent_ruby_unavailable")
                 (skip-tests "test/helpers/integration.rb"
+                            "test_hot_restart_does_not_drop_connections_threads"
                             "test_puma_started_log_writing"
                             "test_require_dependencies"))))
           (add-before 'check 'relax-test-case-timeout
-- 
2.49.0

Message #11 received at 78837 <at> debbugs.gnu.org (full text, mbox):

From: Nicolas Graves <ngraves <at> ngraves.fr>
To: 78837 <at> debbugs.gnu.org
Cc: Nicolas Graves <ngraves <at> ngraves.fr>
Subject: [PATCH 2/3] gnu: ruby-minitest-stub-const: Add minitest native-input.
Date: Thu, 19 Jun 2025 10:39:07 +0200

For some reason, I'm able to run tests locally without ruby-minitest,
but it's failing on CI as if minitest was missing.

* gnu/packages/ruby-xyz.scm (ruby-minitest-stub-const)[native-inputs]:
  Add ruby-minitest.
---
 gnu/packages/ruby-xyz.scm | 1 +
 1 file changed, 1 insertion(+)

diff --git a/gnu/packages/ruby-xyz.scm b/gnu/packages/ruby-xyz.scm
index 9c3fb4b167..228ec15c63 100644
--- a/gnu/packages/ruby-xyz.scm
+++ b/gnu/packages/ruby-xyz.scm
@@ -5460,6 +5460,7 @@ (define-public ruby-minitest-stub-const
          (sha256
           (base32 "08avrpvk8yj1f6x1m8bcq7gdf08x1zcx9iamjysxnalrpwnw5lh6"))))
       (build-system ruby-build-system)
+      (native-inputs (list ruby-minitest))
       (synopsis "Stub constants for the duration of a block in MiniTest")
       (description
        "This package provides a MiniTest extension to stub constants for the
-- 
2.49.0

Message #14 received at 78837 <at> debbugs.gnu.org (full text, mbox):

From: Nicolas Graves <ngraves <at> ngraves.fr>
To: 78837 <at> debbugs.gnu.org
Cc: Nicolas Graves <ngraves <at> ngraves.fr>
Subject: [PATCH 3/3] gnu: ruby-spy: Skip problematic test.
Date: Thu, 19 Jun 2025 10:39:08 +0200

* gnu/packages/ruby-xyz.scm (ruby-spy)[arguments]{phases}: Add phase
'delete-problematic-tests.
---
 gnu/packages/ruby-xyz.scm | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/gnu/packages/ruby-xyz.scm b/gnu/packages/ruby-xyz.scm
index 228ec15c63..144e07484b 100644
--- a/gnu/packages/ruby-xyz.scm
+++ b/gnu/packages/ruby-xyz.scm
@@ -14122,6 +14122,13 @@ (define-public ruby-spy
      (list
       #:phases
       #~(modify-phases %standard-phases
+          (add-after 'unpack 'delete-problematic-tests
+            (lambda _
+              ;; XXX: Flaky test? Fails on CI.
+              (substitute* "test/integration/test_subroutine_spying.rb"
+                ((".*test_spy_on_hooks_and_saves_spy_with_array.*" all)
+                 (string-append all
+                                "      skip('fails on guix')\n")))))
           (add-after 'extract-gemspec 'relax-requirements
             (lambda _
               (substitute* "spy.gemspec"
-- 
2.49.0

Message #17 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Christopher Baines <mail <at> cbaines.net>
To: Nicolas Graves via Guix-patches via <guix-patches <at> gnu.org>
Cc: 78837 <at> debbugs.gnu.org, Nicolas Graves <ngraves <at> ngraves.fr>
Subject: Re: [bug#78837] [PATCH ruby-team 0/3] Almost there!
Date: Thu, 19 Jun 2025 10:45:12 +0100

[Message part 1 (text/plain, inline)]

Nicolas Graves via Guix-patches via <guix-patches <at> gnu.org> writes:

> Now that I don't have locally failing tests, let's try and address tests that fail on CI.
>
> Nicolas Graves (3):
>   gnu: ruby-puma: Skip additional failing test.
>   gnu: ruby-minitest-stub-const: Add minitest native-input.
>   gnu: ruby-spy: Skip problematic test.
>
>  gnu/packages/ruby-xyz.scm | 9 +++++++++
>  1 file changed, 9 insertions(+)

Hmm, I can build all of these locally. Maybe some of the tests are flaky
but I'm tempted to wait until QA has caught up.

I'm most unsure about the ruby-minitest-stub-const change, that seems to
be building on CI as well https://ci.guix.gnu.org/build/11337230/details

[signature.asc (application/pgp-signature, inline)]

Message #23 received at 78837 <at> debbugs.gnu.org (full text, mbox):

From: Nicolas Graves <ngraves <at> ngraves.fr>
To: Christopher Baines <mail <at> cbaines.net>
Cc: 78837 <at> debbugs.gnu.org
Subject: Re: [bug#78837] [PATCH ruby-team 0/3] Almost there!
Date: Thu, 19 Jun 2025 14:17:43 +0200

On 2025-06-19 10:45, Christopher Baines wrote:

> Nicolas Graves via Guix-patches via <guix-patches <at> gnu.org> writes:
>
>> Now that I don't have locally failing tests, let's try and address tests that fail on CI.
>>
> Hmm, I can build all of these locally. Maybe some of the tests are flaky
> but I'm tempted to wait until QA has caught up.
>
> I'm most unsure about the ruby-minitest-stub-const change, that seems to
> be building on CI as well
> https://ci.guix.gnu.org/build/11337230/details

Maybe I'm not looking at the right place? Usually what I do is I look at
the dashboard of the last evaluation of the branch :
https://ci.guix.gnu.org/eval/2065235/dashboard

then I type in "ruby-" to focus on those, and I still see ruby-spy,
ruby-puma failing for some reason.

What am I doing wrong?

-- 
Best regards,
Nicolas Graves

Message #26 received at 78837 <at> debbugs.gnu.org (full text, mbox):

From: Christopher Baines <mail <at> cbaines.net>
To: Nicolas Graves <ngraves <at> ngraves.fr>
Cc: 78837 <at> debbugs.gnu.org
Subject: Re: [bug#78837] [PATCH ruby-team 0/3] Almost there!
Date: Fri, 20 Jun 2025 11:23:19 +0100

[Message part 1 (text/plain, inline)]

Nicolas Graves <ngraves <at> ngraves.fr> writes:

> On 2025-06-19 10:45, Christopher Baines wrote:
>
>> Nicolas Graves via Guix-patches via <guix-patches <at> gnu.org> writes:
>>
>>> Now that I don't have locally failing tests, let's try and address tests that fail on CI.
>>>
>> Hmm, I can build all of these locally. Maybe some of the tests are flaky
>> but I'm tempted to wait until QA has caught up.
>>
>> I'm most unsure about the ruby-minitest-stub-const change, that seems to
>> be building on CI as well
>> https://ci.guix.gnu.org/build/11337230/details
>
> Maybe I'm not looking at the right place? Usually what I do is I look at
> the dashboard of the last evaluation of the branch :
> https://ci.guix.gnu.org/eval/2065235/dashboard
>
> then I type in "ruby-" to focus on those, and I still see ruby-spy,
> ruby-puma failing for some reason.
>
> What am I doing wrong?

I don't think you're doing anything wrong, although it can be difficult
to reason about these flaky packages.

These are the links to the relevant packages:

ruby-spy                 https://ci.guix.gnu.org/build/11336972/details
ruby-puma                https://ci.guix.gnu.org/build/11336817/details
ruby-minitest-stub-const https://ci.guix.gnu.org/build/11337230/details

QA has now pretty much built ruby-team and the changes in this issue
https://qa.guix.gnu.org/issue/78837

From the package changes
https://qa.guix.gnu.org/issue/78837/package-changes it looks like
ruby-puma can still fail to build (and failed 3 times
https://data.qa.guix.gnu.org/gnu/store/vsvjkzkbhwpmifazbqgdrq0n8vbj1nly-ruby-puma-6.6.0.drv
 )

For ruby-spy and ruby-minitest-stub-const, all the builds before and
after have succeeded, which doesn't really say much and is a shortcoming
in our testing approach as since the first build succeeded, there have
been no other attempts.

[signature.asc (application/pgp-signature, inline)]

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.