git/list[1] front-page[2] threads[3] people[4] search[5] about
 

[PATCH v2 02/10] t7422: fix flaky test caused by buffered stdout

From
Patrick Steinhardt <ps@pks.im>
Date
Jan 6, 2025, 11:16 UTC
Message-ID
<20250106-b4-pks-ci-fixes-v2-2-06ae540771b7@pks.im>
In-Reply-To
<20250106-b4-pks-ci-fixes-v2-0-06ae540771b7@pks.im>

One test in t7422 asserts that `git submodule status --recursive` properly handles SIGPIPE. This test is flaky though and may sometimes not see a SIGPIPE at all:

    expecting success of 7422.18 'git submodule status --recursive propagates SIGPIPE':
            { git submodule status --recursive 2>err; echo $?>status; } |
                    grep -q X/S &&
            test_must_be_empty err &&
            test_match_signal 13 "$(cat status)"
    ++ git submodule status --recursive
    ++ grep -q X/S
    ++ echo 0
    ++ test_must_be_empty err
    ++ test 1 -ne 1
    ++ test_path_is_file err
    ++ test 1 -ne 1
    ++ test -f err
    ++ test -s err
    +++ cat status
    ++ test_match_signal 13 0
    ++ test 0 = 141
    ++ test 0 = 269
    ++ return 1
    error: last command exited with $?=1
    not ok 18 - git submodule status --recursive propagates SIGPIPE

The issue is caused by us using grep(1) to terminate the pipe on the first matching line in the recursing git-submodule(1) process. Standard streams are typically buffered though, so this condition is racy and may cause us to terminate the pipe after git-submodule(1) has already exited, and in that case we wouldn't see the expected signal.

Fix the issue by making the writer fill the pipe buffer before we execute git-submodule(1). Ideally, it would be git-submodule(1) itself that does produce all that data, but it would require us to create a large amount of submodules, which is inefficient. Instead, we use Perl to print gibberish until the buffer is filled.

To verify that this works as expected one can apply the following patch to the preimage of this commit, which used to reliably trigger the race:

    diff --git a/t/t7422-submodule-output.sh b/t/t7422-submodule-output.sh
    index f21e920367..9338c75626 100755
    --- a/t/t7422-submodule-output.sh
    +++ b/t/t7422-submodule-output.sh
    @@ -168,7 +168,7 @@ done
     test_expect_success !MINGW 'git submodule status --recursive propagates SIGPIPE' '
            { git submodule status --recursive 2>err; echo $?>status; } |
    -		grep -q X/S &&
    +		{ sleep 1 && grep -q X/S; } &&
            test_must_be_empty err &&
            test_match_signal 13 "$(cat status)"
     '
With the pipe-stuffing workaround the test runs successfully.
Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 t/t7422-submodule-output.sh | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/t/t7422-submodule-output.sh b/t/t7422-submodule-output.sh
index f21e9203678b94701281d5339ae8bfe53d5de0ed..976f91b0ebd9d82daee3802a212dd3f4031fe86b 100755
--- a/t/t7422-submodule-output.sh
+++ b/t/t7422-submodule-output.sh
@@ -167,8 +167,14 @@ do
 done
 
 test_expect_success !MINGW 'git submodule status --recursive propagates SIGPIPE' '
-	{ git submodule status --recursive 2>err; echo $?>status; } |
-		grep -q X/S &&
+	{
+		# Stuff pipe buffer full of input so that `git submodule
+		# status` will block on write; this script will write over
+		# 128kb.
+		perl -le "print q{foo} for (1..33000)" &&
+		git submodule status --recursive 2>err
+		echo $?>status
+	} | grep -q X/S &&
 	test_must_be_empty err &&
 	test_match_signal 13 "$(cat status)"
 '
-- 
2.48.0.rc1.245.gb3e6e7acbc.dirty
Previous: Patrick SteinhardtNext: Jeff King
Message 31 of 57 in “A couple of CI improvements”
  1. 00/10 A couple of CI improvementsPatrick Steinhardt, Jan 3, 2025
  2. 01/10 t0060: fix EBUSY in MinGW when setting up runtime prefixPatrick Steinhardt, Jan 3, 2025
  3. 02/10 t7422: fix flaky test caused by buffered stdoutPatrick Steinhardt, Jan 3, 2025
  4. Jeff KingJan 3, 2025
  5. Patrick SteinhardtJan 6, 2025
  6. Jeff KingJan 7, 2025
  7. Patrick SteinhardtJan 7, 2025
  8. Patrick SteinhardtJan 7, 2025
  9. Jeff KingJan 9, 2025
  10. Junio C HamanoJan 9, 2025
  11. Jeff KingJan 7, 2025
  12. Junio C HamanoJan 7, 2025
  13. 03/10 github: adapt containerized jobs to be rootlessPatrick Steinhardt, Jan 3, 2025
  14. 05/10 github: simplify computation of the job's distroPatrick Steinhardt, Jan 3, 2025
  15. Junio C HamanoJan 3, 2025
  16. 06/10 gitlab-ci: remove the "linux-old" jobPatrick Steinhardt, Jan 3, 2025
  17. Junio C HamanoJan 3, 2025
  18. 04/10 github: convert all Linux jobs to be containerizedPatrick Steinhardt, Jan 3, 2025
  19. Jeff KingJan 3, 2025
  20. Jeff KingJan 3, 2025
  21. Patrick SteinhardtJan 6, 2025
  22. Junio C HamanoJan 3, 2025
  23. 07/10 gitlab-ci: add linux32 job testing against i386Patrick Steinhardt, Jan 3, 2025
  24. 08/10 ci: stop special-casing for Ubuntu 16.04Patrick Steinhardt, Jan 3, 2025
  25. 09/10 ci: use latest Ubuntu releasePatrick Steinhardt, Jan 3, 2025
  26. 10/10 ci: remove stale code for Azure PipelinesPatrick Steinhardt, Jan 3, 2025
  27. Jeff KingJan 3, 2025
  28. 00/10 A couple of CI improvementsPatrick Steinhardt, Jan 6, 2025
  29. 01/10 t0060: fix EBUSY in MinGW when setting up runtime prefixPatrick Steinhardt, Jan 6, 2025
  30. 03/10 github: adapt containerized jobs to be rootlessPatrick Steinhardt, Jan 6, 2025
  31. 02/10 t7422: fix flaky test caused by buffered stdoutPatrick Steinhardt, Jan 6, 2025
  32. Jeff KingJan 7, 2025
  33. 04/10 github: convert all Linux jobs to be containerizedPatrick Steinhardt, Jan 6, 2025
  34. 06/10 gitlab-ci: remove the "linux-old" jobPatrick Steinhardt, Jan 6, 2025
  35. 05/10 github: simplify computation of the job's distroPatrick Steinhardt, Jan 6, 2025
  36. 07/10 gitlab-ci: add linux32 job testing against i386Patrick Steinhardt, Jan 6, 2025
  37. 09/10 ci: use latest Ubuntu releasePatrick Steinhardt, Jan 6, 2025
  38. 08/10 ci: stop special-casing for Ubuntu 16.04Patrick Steinhardt, Jan 6, 2025
  39. 10/10 ci: remove stale code for Azure PipelinesPatrick Steinhardt, Jan 6, 2025
  40. 00/10 A couple of CI improvementsPatrick Steinhardt, Jan 10, 2025
  41. 01/10 t0060: fix EBUSY in MinGW when setting up runtime prefixPatrick Steinhardt, Jan 10, 2025
  42. 02/10 t7422: fix flaky test caused by buffered stdoutPatrick Steinhardt, Jan 10, 2025
  43. Christian CouderJan 24, 2025
  44. 04/10 github: convert all Linux jobs to be containerizedPatrick Steinhardt, Jan 10, 2025
  45. 03/10 github: adapt containerized jobs to be rootlessPatrick Steinhardt, Jan 10, 2025
  46. Christian CouderJan 24, 2025
  47. Johannes SchindelinAug 28, 2025
  48. Johannes SchindelinNov 17, 2025
  49. 05/10 github: simplify computation of the job's distroPatrick Steinhardt, Jan 10, 2025
  50. 06/10 gitlab-ci: remove the "linux-old" jobPatrick Steinhardt, Jan 10, 2025
  51. 07/10 gitlab-ci: add linux32 job testing against i386Patrick Steinhardt, Jan 10, 2025
  52. 08/10 ci: stop special-casing for Ubuntu 16.04Patrick Steinhardt, Jan 10, 2025
  53. 09/10 ci: use latest Ubuntu releasePatrick Steinhardt, Jan 10, 2025
  54. 10/10 ci: remove stale code for Azure PipelinesPatrick Steinhardt, Jan 10, 2025
  55. Jeff KingJan 10, 2025
  56. Christian CouderJan 24, 2025
  57. Patrick SteinhardtJan 27, 2025

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.