Skip to content

Cleanup DEPS, BUILD.gn, make build/ available - #27

Open
illwieckz wants to merge 7 commits into
masterfrom
illwieckz/cleanup
Open

illwieckz wants to merge 7 commits into
masterfrom
illwieckz/cleanup

Conversation

@illwieckz

Copy link
Copy Markdown
Member
  • Remove BUILD.gn support
  • Remove DEPS files
  • Ignore /toolchain symbolic link
    This way we can symlink the prebuilt toolchain by Google
    when we need to build with it for investigation purpose.
  • Move files from build/ to tools/build/
    Keep the build/ directory available as a build folder.
  • Ignore the build/ directory
    It is common practice with CMake to build in build/.

Move files from build/ to tools/build/.

Keep the build/ directory available as a build folder.
Ignore toolchain/ symbolic link.

This way we can symlink the prebuilt toolchain by Google
when we need to build with it for investigation purpose.
It is common practice with CMake to build in build/.
@illwieckz illwieckz changed the title Illwieckz/cleanup Cleanup DEPS, BUILD.gn, keep build/ available Jun 25, 2026
@illwieckz illwieckz mentioned this pull request Jun 25, 2026
@illwieckz
illwieckz force-pushed the illwieckz/cleanup branch 2 times, most recently from a698942 to 8ab33e7 Compare July 7, 2026 10:46
@illwieckz illwieckz changed the title Cleanup DEPS, BUILD.gn, keep build/ available Cleanup DEPS, BUILD.gn, make build/ available Jul 13, 2026
@illwieckz
illwieckz force-pushed the illwieckz/cleanup branch 2 times, most recently from d281908 to 7064df2 Compare October 6, 2026 00:21
@illwieckz

illwieckz commented Oct 6, 2026 •

Copy link
Copy Markdown
Member Author

Hmm, something broke something:

CC: /home/runner/work/native_client/native_client/sdk/sdk/bin/x86_64-nacl-clang
ASPPCOM: /home/runner/work/native_client/native_client/sdk/sdk/bin/x86_64-nacl-clang -m64 -D__STDC_LIMIT_MACROS=1 -D__STDC_FORMAT_MACROS=1 -D_GNU_SOURCE=1 -D_POSIX_C_SOURCE=199506 -D_XOPEN_SOURCE=600 -DDYNAMIC_ANNOTATIONS_ENABLED=1 -DDYNAMIC_ANNOTATIONS_PREFIX=NACL_ -DNACL_BUILD_ARCH=x86 -DNACL_BUILD_SUBARCH=64 -I/home/runner/work/native_client/native_client/scons-out/nacl_irt-x86-64/include -I/home/runner/work/native_client/native_client/include-hax -I/home/runner/work/native_client/native_client/src/untrusted/pthread -c -o
UNKNOWN COMPILER
*** BROKEN  run_seccomp_open_errno_test
*** BROKEN  run_ncval_annotate_test
*** BROKEN  run_dynamic_modify_test
*** BROKEN  run_nanosleep_test
*** BROKEN  run_sel_ldr_seccomp_test
*** BROKEN  run_timefuncs_test
*** BROKEN  run_raw_timefuncs_test
*** BROKEN  run_main_thread_pthread_exit_test2
There are 8 broken tests.

Edit: This was unrelated.

@illwieckz

illwieckz commented Oct 6, 2026 •

Copy link
Copy Markdown
Member Author

Ah, it seems to require the DEPS file:

FileNotFoundError: [Errno 2] No such file or directory: '…/native_client/native_client/DEPS'

@illwieckz

Copy link
Copy Markdown
Member Author

Ah, it seems to require the DEPS file:

FileNotFoundError: [Errno 2] No such file or directory: '…/native_client/native_client/DEPS'

That import is useless.

@illwieckz

Copy link
Copy Markdown
Member Author

OK, now the only failing test is this one (windows-amd64-mingw):

======================================================================
Stderr for nacl_newlib.run_mmap_atomicity_test:
======================================================================
[7552,784:01:09:15.180000] Native Client module will be loaded at base address 0x000006b100000000

waiting for child thread...
child thread has started.
mmap call #0
checking for write to page...
mmap call #1
checking for write to page...
mmap call #2
checking for write to page...
mmap call #3
…
mmap call #886
checking for write to page...
mmap call #887
checking for write to page...
mmap call #888
checking for write to page...
mmap call #889
checking for write to page...
Killed test due to timeout!

[  FAILED  ] nacl_newlib.run_mmap_atomicity_test (110578 ms)

I have no idea why (help welcome).

@illwieckz

illwieckz commented Oct 6, 2026 •

Copy link
Copy Markdown
Member Author
# * This test is extremely slow on the memcheck bot, which eventually
#   times out, so it is disabled there.  This may be a consequence of
#   running under a VM, since running the test on memcheck locally is
#   fast enough.

Maybe I'm just unlucky?

Failing after 8m

It's supposed to last 10 minutes.

@slipher

slipher commented Oct 6, 2026

Copy link
Copy Markdown
Member

run_mmap_atomicity_test does a sort of hand-off between 2 threads 1000 times. So maybe the progress could be slow with unlucky thread scheduling, if it never manages to get both threads running at once. A possible solution is to add is_flaky=True to the test definition so it doesn't run on CI.

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