[illumos-Advocates] [RTI] 754 (find_elf and perl51.0), 850 (libc, etc, should just use /usr/bin/sh)

Garrett D'Amore garrett at damore.org
Sun Apr 10 20:52:49 PDT 2011


On Sun, 2011-04-10 at 23:04 -0400, Richard Lowe wrote:
> Anyone?  Gordon was eager for at least #754 to land.

RTI approved.

	- Garrett
> 
> -- Rich 
> 
> On Apr 8, 2011 8:31 PM, "Richard Lowe" <richlowe at richlowe.net> wrote:
> > outgoing:
> > comparing with /builds/illumos-clone
> > searching for changes
> > 
> > changeset: 13330:25d2c0b6a6c3
> > user: Richard Lowe <richlowe at richlowe.net>
> > date: Wed Mar 30 23:00:17 2011 -0400
> > 
> > description:
> > 873 Want a bsd-like <paths.h>
> > Reviewed by: Gordon Ross <gwr at nexenta.com>
> > Reviewed by: John Sonnenschein <johns at joyent.com>
> > 
> > modified:
> > usr/src/head/Makefile
> > usr/src/pkg/manifests/system-header.mf
> > added:
> > usr/src/head/paths.h
> > removed:
> > usr/src/lib/libresolv2/include/paths.h
> > 
> > 
> > changeset: 13331:fdb1b87ce7c8
> > user: Richard Lowe <richlowe at richlowe.net>
> > date: Tue Mar 29 22:27:20 2011 +0000
> > 
> > description:
> > 850 system(3C) and similar could just use /bin/sh
> > Reviewed by: Gordon Ross <gwr at nexenta.com>
> > Reviewed by: John Sonnenschein <johns at joyent.com>
> > 
> > modified:
> > usr/src/cmd/cron/at.c
> > usr/src/cmd/ed/ed.c
> > usr/src/lib/libc/port/gen/execvp.c
> > usr/src/lib/libc/port/stdio/popen.c
> > usr/src/lib/libc/port/stdio/system.c
> > usr/src/lib/libc/port/threads/spawn.c
> > usr/src/man/man1/at.1
> > usr/src/man/man1/ed.1
> > usr/src/man/man1/env.1
> > usr/src/man/man1/nice.1
> > usr/src/man/man3c/popen.3c
> > usr/src/man/man3c/system.3c
> > usr/src/man/man5/standards.5
> > 
> > 
> > changeset: 13332:6f6dea86cfd9
> > tag: tip
> > user: Richard Lowe <richlowe at richlowe.net>
> > date: Thu Apr 07 21:19:00 2011 -0400
> > 
> > description:
> > 754 Build of illumos-gate fails with perl-5.10.0 as default perl
> > Reviewed by: Gordon Ross <gordon.w.ross at gmail.com>
> > Reviewed by: Albert Lee <trisk at opensolaris.org>
> > 
> > modified:
> > usr/src/tools/scripts/find_elf.pl
> > 
> > pbchk:
> > Copyright check:
> > usr/src/cmd/cron/at.c: no copyright claim for current year found
> > usr/src/cmd/ed/ed.c: no copyright claim for current year found
> > usr/src/head/Makefile: no copyright claim for current year found
> > usr/src/head/paths.h: no copyright claim for current year found
> > usr/src/lib/libc/port/gen/execvp.c: no copyright claim for current
> > year found
> > usr/src/lib/libc/port/stdio/popen.c: no copyright claim for
> > current year found
> > usr/src/lib/libc/port/stdio/system.c: no copyright claim for
> > current year found
> > usr/src/lib/libc/port/threads/spawn.c: no copyright claim for
> > current year found
> > usr/src/man/man1/at.1: no copyright claim for current year found
> > usr/src/man/man1/ed.1: no copyright claim for current year found
> > usr/src/man/man1/env.1: no copyright claim for current year found
> > usr/src/man/man1/nice.1: no copyright claim for current year found
> > usr/src/man/man3c/popen.3c: no copyright claim for current year
> found
> > usr/src/man/man3c/system.3c: no copyright claim for current year
> found
> > usr/src/man/man5/standards.5: no copyright claim for current year
> found
> > usr/src/pkg/manifests/system-header.mf: no copyright claim for
> > current year found
> > usr/src/tools/scripts/find_elf.pl: no copyright claim for current
> year found
> > 
> > C style check:
> > Skipping usr/src/cmd/ed/ed.c...
> > 
> > Header format check:
> > 
> > Java style check:
> > 
> > Mapfile comment check:
> > 
> > File permission check:
> > 
> > Keywords check:
> > 
> > Comments check:
> > 
> > Checking for new tags:
> > 
> > Checking for multiple heads (or branches):
> > 
> > Checking for branch changes:
> > 
> > Checking for uncommitted changes:
> > 
> > Checking for merges:
> > 
> > ed is very far from clean, and so skipped.
> > 
> > Testing:
> > 
> > 754: Full build (as in mail_msg), and asked Gordon to run a build
> > with his perl 5.10-based system. my build succeeded, he also reports
> > success.
> > Ran find_elf by hand with '/usr/perl5/X.Y.Z/bin/perl find_elf
> > -fr $ROOT' with 5.8.4 and 5.10.0 and verified there was no closedir
> > noise
> > 
> > 850:
> > 
> > Ran both ed and /usr/xpg4/bin/ed under truss and did '!ls',
> > verified that /usr/bin/sh was used in both cases, and that ls ran.
> > Ran both at and /usr/xpg4/bin/at with $SHELL unset while
> > trussing cron, and verified that /usr/bin/sh was used to run the job
> > in both cases.
> > Ran a simple program which used system(3C), execvp(3C),
> > popen(3C) and posix_spawnp(3C) in ways which invoke the shell, and
> > verified that /usr/bin/sh was used in all cases when the program was
> > built normally, as XPG4, and as XPG6. Also verified that the jobs
> wre
> > actually run.
> > 
> > The (crappy hack) program is attached. 'script.sh' is an
> > executable shell scrip with no #! line to specify the interpretter,
> > which will cause the attempt to exec() it to fail, and us to fall
> back
> > to running it with the shell. These are the cases where execvp() and
> > posix_spawnp use the shell.
> > 
> > I have also been running this change on my machine for
> > several days at this point.
> > 
> > -- Rich
> 
> _______________________________________________
> Advocates mailing list
> Advocates at lists.illumos.org
> http://lists.illumos.org/m/listinfo/advocates





More information about the Advocates mailing list