aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2013-10-28Benchmark inputs for powSiddhesh Poyarekar
These inputs cover all normal processing paths for pow including all its slow paths.
2013-10-28New inputs for expSiddhesh Poyarekar
A more comprehensive set of inputs for exp, including all slow paths. The inputs have been shuffled so that they don't give a false-positive due to a hot cache.
2013-10-28Consolidate conditionals in mp sin/cos functionsSiddhesh Poyarekar
Consolidate conditionals in multiple precision sin and cos functions to prepare the code for addition of probe points.
2013-10-28Fix ChangeLog date.Adhemerval Zanella
2013-10-26Add some more NEWS items.Joseph Myers
2013-10-25PowerPC: strcpy/stpcpy optimization for PPC64/POWER7Adhemerval Zanella
This patch intends to unify both strcpy and stpcpy implementationsi for PPC64 and PPC64/POWER7. The idead default powerpc64 implementation is to provide both doubleword and word aligned memory access. For PPC64/POWER7 is also provide doubleword and word memory access, remove the branch hints, use the cmpb instruction for compare doubleword/words, and add an optimization for inputs of same alignment.
2013-10-25Fix gethostbyname_r example. Fixes bug 2801.Ondřej Bílka
2013-10-25Changelog for last commit.Ondřej Bílka
2013-10-25Make strptime %Z consistent between doc and code. Fixes bug 14876Ondřej Bílka
2013-10-25Acknowledge that fnmatch can fail. Fixes bug 14029.Ondřej Bílka
2013-10-25Document that mmap() returns MAP_FAILED on error, as per the POSIX standard.Fabrice Bauzac
2013-10-25Fix stack overflow due to large AF_INET6 requestsSiddhesh Poyarekar
Resolves #16072 (CVE-2013-4458). This patch fixes another stack overflow in getaddrinfo when it is called with AF_INET6. The AF_UNSPEC case was fixed as CVE-2013-1914, but the AF_INET6 case went undetected back then.
2013-10-25Fix incorrect getaddrinfo assertion triggerAllan McRae
[BZ #9954] With the following /etc/hosts: 127.0.0.1 www.my-domain.es 127.0.1.1 www.my-domain.es 192.168.0.1 www.my-domain.es Using getaddrinfo() on www.my-domain.es, trigger the following assertion: ../sysdeps/posix/getaddrinfo.c:1473: rfc3484_sort: Assertion `src->results[i].native == -1 || src->results[i].native == a1_native' failed. This is due to two different bugs: - In rfc3484_sort() rule 7, src->results[i].native is assigned even if src->results[i].index is -1, meaning that no interface is associated. - In getaddrinfo() the source IP address used with the lo interface needs a special case, as it can be any IP within 127.X.Y.Z.
2013-10-24Adjust language-code fields of LC_ADDRESS.Chris Leonard
2013-10-21Document rpcgen -5. Fixes bug 15825Ondřej Bílka
2013-10-20Print the reason why preloading failed in do_preload()Michael Stahl
2013-10-20When glob pattern contains a trailing slash match only directories. Fixes ↵Ondřej Bílka
bug 10278.
2013-10-20Replace alloca in __tzfile_read by malloc. Fixes bug 15670Ondřej Bílka
2013-10-19Mention FIPS 140-2 compliance and Sun RPC.Carlos O'Donell
The Secure RPC implementation in glibc uses DES encryption during authentication of the user. This use of DES means that anyone using Sun RPC will likely not be compliant with FIPS 140-2 which forbids the use of DES. One solution to the compliance issue is to disable AUTH_DES and AUTH_KERB, both use DES, when in FIPS compliance mode. This is not a good idea because it disables all of the even mildly secure methods of authentication allowing only plain text methods. Instead we leave AUTH_DES and AUTH_KERB enabled in FIPS compliance mode and document the compliance issue in the manual. FIPS allows this, that is to say that if you can't fix it you must document the non-compliance. This commit adds documentation to that effect in the "DES encryption and password handling" section of the manual.
2013-10-18Enhance localedef --list-archive option.Carlos O'Donell
The localedef --list-archive option claims that it can accept a [file] argument and list the contents of that archive. The support was never implemented. This patch adds that support and allows --list-archive to work as expected. You can now use localedef to list the contents of arbitrary locale archives by using: ./localedef --list-archive file
2013-10-18Move entries to correct port ChangeLog files.Joseph Myers
2013-10-18Add e500 port.Joseph Myers
2013-10-18Remove duplicate bug numbers from NEWS.Joseph Myers
2013-10-18Fix localedef collation handling of <U0000> (bug 15948).Richard Sandiford
2013-10-18Don't include tls.h in test casesSiddhesh Poyarekar
Remove tls.h includes where they are not needed.
2013-10-18Remove assert in malloc statistic. Fixes bug 12486.Ondřej Bílka
2013-10-17Fix inet_network("1 bar"). Fixes bug 15277.Ondřej Bílka
2013-10-17Don't use gethostbyaddr to determine canonical nameAndreas Schwab
2013-10-17Format floating routines.Ondřej Bílka
2013-10-17soft-fp: make extensions quiet signaling NaNs (bug 16041).Joseph Myers
2013-10-16soft-fp: fix horizontal whitespace.Joseph Myers
2013-10-15soft-fp: remove unused macros.Joseph Myers
2013-10-15Clear initfini list after freeing. Fixes bug 15308.Ondřej Bílka
2013-10-15soft-fp: fix vertical whitespace and indentation.Joseph Myers
2013-10-14Fix error_tail overflow in allocation calculation.Ondřej Bílka
2013-10-14Correctly copy resolver address. Fixes bug #13028.Ondřej Bílka
2013-10-14ldd: make try_trace more robust and portablePatrick 'P. J.' McDermott
It was noted in 2005 (BZ #832), 2006 (BZ #3266), and 2007 [1] that ldd fails on shells other than Bash >= 3.0 because of the pipefail option around try_trace (added on 2004-12-08). EGLIBC was patched in 2008 [2] (r6912) to make the pipefail check run only on shells that support it, but RTLD output would still be lost on other shells with certain SELinux policies. This patch rewrites try_trace to work on any POSIX-conformant shell in such a way as to also work with such SELinux policies. It also obviates one difference between glibc and EGLIBC. URL: https://sourceware.org/ml/libc-alpha/2007-01/msg00041.html URL: http://www.eglibc.org/archives/patches/msg00526.html 2013-09-11 P. J. McDermott <pj@pehjota.net> [BZ #832] * elf/ldd.bash.in (try_trace): More robustly and portably work around SELinux terminal write permissions by using a command substitution instead of a pipeline and pipefail option.
2013-10-13Adjust language-code fields of LC_ADDRESS.Chris Leonard
2013-10-12soft-fp: fix preprocessor indentation.Joseph Myers
2013-10-12Fix typos.Yuri Chornoivan
2013-10-12Fix typo in setlocale.c. Fixes BZ #15764Reuben Thomas
2013-10-12soft-fp: make __unord* raise "invalid" for signaling NaNs (bug 16036).Joseph Myers
2013-10-12soft-fp: make ordered comparisons raise "invalid" for quiet NaNs (bug 14910).Joseph Myers
2013-10-12soft-fp: add missing FP_INIT_EXCEPTIONS and FP_INIT_ROUNDMODE calls.Joseph Myers
2013-10-12soft-fp: add macro FP_NO_EXCEPTIONS.Joseph Myers
2013-10-12soft-fp: fix _FP_DIV_MEAT_* returning results with wrong exponent (bug 16032).Joseph Myers
2013-10-12soft-fp: fix floating-point to integer unsigned saturation.Joseph Myers
2013-10-11Add systemtap markers to math function slow pathsSiddhesh Poyarekar
Add systemtap probes to various slow paths in libm so that application developers may use systemtap to find out if their applications are hitting these slow paths. We have added probes for pow, exp, log, tan, atan and atan2.
2013-10-11Fix fwrite() reading beyond end of buffer in error pathEric Biggers
Partially revert commits 2b766585f9b4ffabeef2f36200c275976b93f2c7 and de2fd463b1c0310d75084b6d774fb974075a4ad9, which were intended to fix BZ#11741 but caused another, likely worse bug, namely that fwrite() and fputs() could, in an error path, read data beyond the end of the specified buffer, and potentially even write this data to the file. Fix BZ#11741 properly by checking the return value from _IO_padn() in stdio-common/vfprintf.c.
2013-10-10Fix readdir regressions on sparc 32-bit.David S. Miller
* sysdeps/posix/dirstream.h (struct __dirstream): Fix alignment of directory block.