Pull request: pasky/fixes-overdue

[Petr Baudis <pasky at suse dot cz>]

Hello!

The following changes since commit 3540d66b669af54900b2e4bfc0ab82960e84a471:

  Fix memory leak in fnmatch (2010-11-12 03:51:28 -0500)

are available in the git repository at:

  git://sources.redhat.com/git/glibc.git pasky/fixes-overdue

Please pull them to the master branch.


commit 75134e46476263ab116be341531cadb1e6ab87d6 [BZ #6812]
Author: Jan Sembera <jsembera@suse.cz>
Date:   Sun Aug 22 15:43:46 2010 +0200

    nscd/hstcache.c: Propagate TRY_AGAIN properly to the clients.
    
    When nscd host cache gets temporary error from nss, it should return
    temporary error instead of permanent error to the application.

commit 9fcff7a79b60cd9202fdacafdbd5fb369933c64b [BZ #11589]
Author: Petr Baudis <pasky@ucw.cz>
Date:   Sun Aug 22 16:44:15 2010 +0200

    Fix jn() precision problems around zero points of j0()
    
    There appears to be a really nasty bug in jn() from fdlibm, which is
    the foundation for most libm implementations (including glibc libm).
    The zeroth-order j0() and first-order j1() cylindrical Bessel functions
    are used to recursively generate the jn() value, but only the zeroth-order
    Bessel function is used to normalize it; however, each of the functions
    gets highly imprecise (approaching "bogus") near its zero point, making
    the jn() value itself bogus.
    
    But in fact, the zero points of j0() and j1() never coincide, thus j1()
    should be used in case it is more precise than j0(). (That is, simply
    when its value is further from zero.)
    
    As an example, 2.4048255576957729_8 is the first zero of j0().
    The proper value as calculated by Mathematica is 0.19899990535769...
    However, jn() returns -inf on 64-bit arch, or 0.185007 on 32-bit arch.
    With the proposed patch below, the returned value is 0.199000.
    
    The fix is based on work by Steve Kargl and Tobias Burnus.

commit 31eb5a4c4986367538ccb154dfce0c84276ba151 [BZ #11610]
Author: Michael Matz <matz@suse.de>
Date:   Sun Aug 22 16:53:24 2010 +0200

    Add proper unwind information for x86_64 _fini
    
    It is impossible to reliably unwind the stack above _fini() on x86_64 since no
    unwind information is provided for it and it modifies a stack register. This
    matters for gdb backtracing - if a process crashes within a destructor, it can
    frequently be essential to look at why the program began terminating in the
    first place.

commit b9be604a7bcffd1fa5fc0141c0dfd2ddcbf09451 [BZ #10085]
Author: Petr Baudis <pasky@ucw.cz>
Date:   Sun Aug 22 16:37:10 2010 +0200

    Fix multiple nss_compat initgroups() bugs
    
    Compat initgroups() is completely broken; the code will always set
    skip_initgroups_dyn to true, so initgroups() will never be actually
    called, but due to the nature of the code, setgrent() won't be called
    either - thus, subsequent invocations of initgroups() will not return
    the NIS group list anymore.
    
    This is a simple patch that makes sure skip_initgroups_dyn is set only
    in case initgroups is not available; it also attempts to handle the
    unavailability of other NSS interfaces better.

commit b321e863ac162595250446a3b107384dc7aecd89 [BZ #11928]
Author: Petr Baudis <pasky@ucw.cz>
Date:   Sun Aug 22 16:15:17 2010 +0200

    Make nscd load /etc/host.conf options in aicache
    
    This patch makes sure _res_hconf is initialized before resolving is being done.
    
    However, this would not be enough since nscd has its own _res_hconf due to
    nscd/res_hconf.c; _res_hconf_init() would work on different _res_hconf instance
    than the NSS routines. We just need to make sure nscd and glibc share the same
    _res_hconf instance - this should not be a problem since users should run
    matching versions of glibc and nscd anyway.

commit c8e6e9e783bc5018525e881f021333c3daa4b0f6 [BZ #11149]
Author: Petr Baudis <pasky@suse.cz>
Date:   Tue Nov 16 03:35:47 2010 +0100

    Allow aux_cache_file open()ing to fail silently even in the chroot mode.
    
    The aux_cache fix of bug 11149 introduced a new bug - normally,
    ldconfig -r never cares if the auxiliary cache is not available and
    that is not a fatal problem, however this is not the case in case
    of ldconfig -r when executed as non-root. In that case, ldconfig -r
    fails hard unless var/cache/ldconfig/ exists within the chroot. This
    patch fixes that.

commit fb8fb8464c2b5a44b69e62bcc3211b9c769416da [BZ #6693]
Author: Jan Kratochvil <jan.kratochvil@redhat.com>
Date:   Tue Nov 16 03:47:22 2010 +0100

    Fix incorrect backtrace unwinding through thread_start() on x86_64
    
    Provide CFI for the outermost clone() to ensure proper unwinding stop
    of gdb.



Summary of changes:

 ChangeLog                              |   44 ++++++++++++++++++++++++++++++++
 elf/ldconfig.c                         |   12 +++------
 nis/nss_compat/compat-initgroups.c     |   19 ++++++++------
 nscd/aicache.c                         |    3 ++
 nscd/hstcache.c                        |   32 +++++++++++++++++++----
 resolv/res_hconf.c                     |    2 +
 sysdeps/ieee754/dbl-64/e_jn.c          |   11 +++++++-
 sysdeps/ieee754/flt-32/e_jnf.c         |   11 +++++++-
 sysdeps/ieee754/ldbl-128/e_jnl.c       |   11 +++++++-
 sysdeps/ieee754/ldbl-128ibm/e_jnl.c    |   11 +++++++-
 sysdeps/ieee754/ldbl-96/e_jnl.c        |   11 +++++++-
 sysdeps/unix/sysv/linux/x86_64/clone.S |    4 ---
 sysdeps/x86_64/elf/initfini.c          |   33 +++++++++++++++++++++++-
 13 files changed, 172 insertions(+), 32 deletions(-)