> On 15 Feb 2023, at 13:19, Adhemerval Zanella Netto via Libc-alpha wrote: > > > > On 14/02/23 22:41, Joan Bruguera wrote: >> While cleaning up old libc version support, the deprecated libc4 code was >> accidentally kept in `implicit_soname`, instead of the libc6 code. >> >> This causes additional symlinks to be created by `ldconfig` for libraries >> without a soname, e.g. a library `libsomething.123.456.789` without a soname >> will create a `libsomething.123` -> `libsomething.123.456.789` symlink. >> >> As the libc6 version of the `implicit_soname` code is just a trivial xstrdup, >> just inline it and remove `implicit_soname` altogether. >> >> Some further simplification looks possible (e.g. the call to `create_links` >> looks like a no-op if `soname == NULL`, other than the verbose printfs), but >> logic is kept as-is for now. >> >> Fixes: BZ #30125 >> Fixes: 8ee878592c4a ("Assume only FLAG_ELF_LIBC6 suport") > > Patch looks ok in general, some comments on the test. You used an old revision > and then patch does not apply cleanly[1], could you rebase the patch and send > a v2? > > [1] https://patchwork.sourceware.org/project/glibc/patch/20230215014124.4144451-1-joanbrugueram@gmail.com/ > >> Signed-off-by: Joan Bruguera > > We do not signed-off-by on glibc, so no need for it. > https://sourceware.org/glibc/wiki/Contribution%20checklist#Developer_Certificate_of_Origin says we do? >> [snip] >> +gcc /dev/null -o $testroot/lib/libsomething.so.1.2.3 -shared > > We can not assume gcc will be presented in PATH, you need to use the CC used > on build. So pass the $(CC) on tst-ldconfig-no-soname.out rule and use it on > the script. (Indeed, GCC may not be the used compiler either, or specifically 'gcc' but some prefixed or versioned gcc).