Re: [PATCH v2 0/2] Update Makefile fragments for DSO sorting tests

["H.J. Lu" <hjl.tools@gmail.com>]

On Thu, Nov 11, 2021 at 6:05 AM Adhemerval Zanella
<adhemerval.zanella@linaro.org> wrote:
>
>
>
> On 11/11/2021 10:46, H.J. Lu via Libc-alpha wrote:
> > On Mon, Nov 8, 2021 at 7:42 AM H.J. Lu <hjl.tools@gmail.com> wrote:
> >>
> >> One problem of using
> >>
> >> $(objpfx)%$o: $(objpfx)%.c $(before-compile); $$(compile-command.c)
> >> compile-command.c = $(compile.c) $(OUTPUT_OPTION) $(compile-mkdep-flags)
> >> compile.c = $(CC) $< -c $(CFLAGS) $(CPPFLAGS)
> >>
> >> to build DSO sorting relocatable objects is that $< may not be the source
> >> file with 3 "make -j 28" parallel builds on a machine with 112 cores.
> >> When generating Makefile fragments at build time, the same file may be
> >> generated more than once at the same time with parallel builds.
> >>
> >> 1. Update dso-ordering-test.py to put all sources for DSO sorting tests
> >> in a single directory, dso-sort-tests-src, and compile relocatable objects
> >> with
> >>
> >> $(objpfx)tst-dso-ordering1-dir/tst-dso-ordering1-a.os: $(objpfx)dso-sort-tests-src/tst-dso-ordering1-a.c
> >>         $(compile.c) $(OUTPUT_OPTION)
> >>
> >> 2. Generate Makefile fragments for DSO sorting tests at configure time
> >> to avoid generate them in the elf directory at build time.
> >>
> >> H.J. Lu (2):
> >>   dso-ordering-test.py: Put all sources in one directory [BZ #28550]
> >>   elf: Pre-generate Makefile fragment for DSO sorting tests [BZ #28550]
> >>
> >>  configure                    | 11 +++++++++
> >>  configure.ac                 | 12 +++++++++
> >>  elf/Makefile                 | 22 ++++-------------
> >>  scripts/dso-ordering-test.py | 47 ++++++++++++++++++++++++++----------
> >>  4 files changed, 62 insertions(+), 30 deletions(-)
> >>
> >> --
> >> 2.33.1
> >>
> >
> > Any comments on the v2 patches?  "make check -j28" has been failing for me
> > on machines with many cores.   I'd like to resolve it soon.
> >
>
> It is on my list, just sorting out Florian comments to rtld-audit patchset.

This bug has been causing false negative results with "makc check -jN"
on machines with many cores.  I will check it in on Tuesday if there are
no objections.

Thanks.

-- 
H.J.