This updates a bunch of locales based on CLDR v29 data:
bg_BG: changing Bulgaria to България
bo_CN: changing ཀྲུང་ཧྭ་མི་དམངས་སྤྱི་མཐུན་རྒྱལ་ཁབ། to རྒྱ་ནག
bo_IN: changing རྒྱ་གར to རྒྱ་གར་
cy_GB: changing Cymru to Y Deyrnas Unedig
dz_BT: changing འབྲུག། to འབྲུག
en_US: changing USA to United States
es_US: changing USA to Estados Unidos
gd_GB: changing Breatainn Mhòr to An Rìoghachd Aonaichte
ha_NG: changing Nigeria to Najeriya
mk_MK: changing Macedonia to Македонија
mn_MN: changing Mongolia to Монгол
sq_MK: changing Macedonia to Maqedoni
sr_RS@latin: changing Srbija i Crna Gora to Srbija
tr_CY: changing Northern Cyprus to Kıbrıs
tr_TR: changing Turkey to Türkiye
ug_CN: changing 中华人民共和国 to جۇڭگو
uz_UZ: changing O'zbekistan to Oʻzbekiston
vi_VN: changing Việt nam to Việt Nam
wae_CH: changing Switzerland to Schwiz
yi_US: changing די פֿאראײניקטע שטאַטן to פֿאַראייניגטע שטאַטן
yo_NG: changing Nigeria to Orílẹ́ède Nàìjíríà
yue_HK: changing 香港 to 中華人民共和國香港特別行政區
zu_ZA: changing Mzansi Afrika to i-South Africa
These all look largely straightforward. Many had English translations
instead of native, and a few have been updated. I can't verify some of
them as I'm not personally familiar, but the CLDR data matches.
The USA->United States seems a little odd, but that is also what the
CLDR database uses everywhere (rather than "United States of America").
We can also fill in a country name where there wasn't one before.
Many look correct to me (mostly the English ones), but there's also
many that I have no idea. But it can't be worse than leaving it
blank ? :)
ar_AE: changing to الإمارات العربية المتحدة
ar_BH: changing to البحرين
ar_DZ: changing to الجزائر
ar_EG: changing to مصر
ar_IN: changing to الهند
ar_IQ: changing to العراق
ar_JO: changing to الأردن
ar_KW: changing to الكويت
ar_LB: changing to لبنان
ar_LY: changing to ليبيا
ar_MA: changing to المغرب
ar_OM: changing to عُمان
ar_QA: changing to قطر
ar_SA: changing to المملكة العربية السعودية
ar_SD: changing to السودان
ar_SS: changing to جنوب السودان
ar_SY: changing to سوريا
ar_TN: changing to تونس
ar_YE: changing to اليمن
as_IN: changing to ভাৰত
ast_ES: changing to España
az_AZ: changing to Azərbaycan
be_BY: changing to Беларусь
bn_IN: changing to ভারত
br_FR: changing to Frañs
brx_IN: changing to भारत
bs_BA: changing to Bosna i Hercegovina
ca_AD: changing to Andorra
ca_ES: changing to Espanya
ca_FR: changing to França
ca_IT: changing to Itàlia
ce_RU: changing to Росси
da_DK: changing to Danmark
de_AT: changing to Österreich
de_BE: changing to Belgien
de_CH: changing to Schweiz
de_LU: changing to Luxemburg
el_CY: changing to Κύπρος
el_GR: changing to Ελλάδα
en_AG: changing to Antigua & Barbuda
en_AU: changing to Australia
en_BW: changing to Botswana
en_CA: changing to Canada
en_DK: changing to Denmark
en_GB: changing to United Kingdom
en_HK: changing to Hong Kong SAR China
en_IE: changing to Ireland
en_IN: changing to India
en_NZ: changing to New Zealand
en_PH: changing to Philippines
en_SG: changing to Singapore
en_ZW: changing to Zimbabwe
es_AR: changing to Argentina
es_BO: changing to Bolivia
es_CL: changing to Chile
es_CO: changing to Colombia
es_CU: changing to Cuba
es_DO: changing to República Dominicana
es_EC: changing to Ecuador
es_ES: changing to España
es_GT: changing to Guatemala
es_HN: changing to Honduras
es_MX: changing to México
es_NI: changing to Nicaragua
es_PA: changing to Panamá
es_PE: changing to Perú
es_PR: changing to Puerto Rico
es_PY: changing to Paraguay
es_SV: changing to El Salvador
es_UY: changing to Uruguay
es_VE: changing to Venezuela
eu_ES: changing to Espainia
fil_PH: changing to Pilipinas
fo_FO: changing to Føroyar
fr_BE: changing to Belgique
fr_CA: changing to Canada
fr_CH: changing to Suisse
fr_FR: changing to France
fr_LU: changing to Luxembourg
fur_IT: changing to Italie
fy_DE: changing to Dútslân
fy_NL: changing to Nederlân
ga_IE: changing to Éire
gl_ES: changing to España
gu_IN: changing to ભારત
gv_GB: changing to Rywvaneth Unys
he_IL: changing to ישראל
hi_IN: changing to भारत
hr_HR: changing to Hrvatska
hu_HU: changing to Magyarország
id_ID: changing to Indonesia
is_IS: changing to Ísland
it_CH: changing to Svizzera
it_IT: changing to Italia
ja_JP: changing to 日本
ka_GE: changing to საქართველო
kk_KZ: changing to Қазақстан
kl_GL: changing to Kalaallit Nunaat
kn_IN: changing to ಭಾರತ
kok_IN: changing to भारत
ko_KR: changing to 대한민국
ks_IN: changing to ہِنٛدوستان
ks_IN@devanagari: changing to भारत
kw_GB: changing to Rywvaneth Unys
ky_KG: changing to Кыргызстан
lt_LT: changing to Lietuva
lv_LV: changing to Latvija
mg_MG: changing to Madagasikara
ml_IN: changing to ഇന്ത്യ
mr_IN: changing to भारत
ms_MY: changing to Malaysia
mt_MT: changing to Malta
nb_NO: changing to Norge
ne_NP: changing to नेपाल
nl_AW: changing to Aruba
nl_BE: changing to België
nl_NL: changing to Nederland
nn_NO: changing to Noreg
or_IN: changing to ଭାରତ
os_RU: changing to Уӕрӕсе
pa_IN: changing to ਭਾਰਤ
pa_PK: changing to ਪਾਕਿਸਤਾਨ
pl_PL: changing to Polska
pt_BR: changing to Brasil
pt_PT: changing to Portugal
ru_RU: changing to Россия
ru_UA: changing to Украина
sd_IN@devanagari: changing to भारत
se_NO: changing to Norga
si_LK: changing to ශ්රී ලංකාව
sk_SK: changing to Slovensko
sl_SI: changing to Slovenija
sq_AL: changing to Shqipëri
sv_SE: changing to Sverige
ta_IN: changing to இந்தியா
ta_LK: changing to இலங்கை
ur_IN: changing to بھارت
ur_PK: changing to پاکستان
Mike Frysinger [Tue, 12 Apr 2016 07:15:11 +0000 (03:15 -0400)]
localedata: fix LC_ADDRESS.country_car entries
These entries have been checked mostly against Wikipedia, but also using
the sources it cites (like the UN and other treaty sources).
Fix incorrect values:
en_BW: changing RB to BW
kl_GL: changing GRO to KN
km_KH: changing LAO to KH
my_MM: changing BA to MYA
oc_FR: changing F to F
tr_CY: changing TR to CY
wae_CH: changing DH to CH
Add missing entries:
aa_DJ: changing to DJI
ak_GH: changing to GH
ar_OM: changing to OM
ar_SS: changing to SUD
ar_YE: changing to YAR
bo_CN: changing to CHN
cmn_TW: changing to RC
dv_MV: changing to MV
dz_BT: changing to BHT
en_AG: changing to AG
es_HN: changing to HN
es_PR: changing to PR
hak_TW: changing to RC
lzh_TW: changing to RC
nan_TW: changing to RC
nan_TW@latin: changing to RC
nl_AW: changing to AUA
pap_AW: changing to AUA
so_DJ: changing to DJI
the_NP: changing to NEP
ug_CN: changing to CHN
yue_HK: changing to HK
zh_CN: changing to CHN
zh_HK: changing to HK
zh_TW: changing to RC
This updates a few locales based on CLDR v29 data.
Add missing fields:
as_IN: changing to 356
dv_MV: changing to 462
kk_KZ: changing to 398
my_MM: changing to 104
rw_RW: changing to 646
tt_RU: changing to 643
Update ones that are wrong:
dz_BT: changing BHU to 064
en_PH: changing 360 to 608
km_KH: changing 418 to 116
ky_KG: changing 643 to 417
tr_CY: changing 792 to 196
wo_SN: changing 450 to 686
As a result of fixing these, I had to update country_ab[23]:
dz_BT: changing BHU to BTN
en_PH: changing ID/IDN to PH/PHL
km_KH: changing LA/LAO to KH/KHM
ky_KG: changing KY/KYR to KG/KGZ
tr_CY: changing TR/TUR to CY/CYP
wo_SN: changing MG/MDG to SN/SEN
Pad with leading zeros to match the standard and other locales:
ber_DZ: changing 12 to 012
ca_AD: changing 20 to 020
en_AG: changing 28 to 028
hy_AM: changing 51 to 051
li_BE: changing 56 to 056
wa_BE: changing 56 to 056
I hand checked the first two sets against ISO 3166-1 directly.
Mike Frysinger [Tue, 9 Feb 2016 11:04:07 +0000 (06:04 -0500)]
localedata: LC_PAPER: use copy directives everywhere
There are only two page sizes that locales use: US-Letter and A4.
For the former, move to copying the en_US locale, while for the
latter, move to copying the i18n locale. This lets us clean up
all the stray comments like FIXME.
Fix pread consolidation on ports that require argument alignment
This patch fixes the __ALIGNMENT_{ARG,COUNT} definition for ports that
define __ASSUME_ALIGNED_REGISTER_PAIRS by including the kernel-features.h
(where it is defined if the case).
Stefan Liebler [Thu, 14 Apr 2016 10:21:53 +0000 (12:21 +0200)]
Fix strfmon_l: Use specified locale for number formatting [BZ #19633]
The commit 985fc132f23dbb83de76c5af9e783ef1b5900148
"strfmon_l: Use specified locale for number formatting [BZ #19633]"
introduced an elf/check-abi-libc testfailure due to __printf_fp_l
on architectures which use sysdeps/ieee754/ldbl-opt/math_ldbl_opt.h.
This patch uses libc_hidden_def instead of ldbl_hidden_def.
The ldbl_strong_alias is removed due to the rename of ___printf_fp_l
to __printf_fp_l.
ChangeLog:
* stdio-common/printf_fp.c (__printf_fp_l):
Rename ___printf_fp_l to __printf_fp_l and
remove strong alias. Use libc_hidden_def instead
of ldbl_hidden_def macro.
The fork handler now runs so late that there is no risk anymore that
other fork handlers in the same thread use malloc, so it is no
longer necessary to install malloc hooks which made a subset
of malloc functionality available to the thread that called fork.
The overloading approach in the W* macros was incompatible with
integer expressions of a type different from int. Applications
using union wait and these macros will have to migrate to the
POSIX-specified int status type.
Mike Frysinger [Tue, 9 Feb 2016 11:25:52 +0000 (06:25 -0500)]
localedata: LC_MEASUREMENT: use copy directives everywhere
There are only two measurement systems that locales use: US and metric.
For the former, move to copying the en_US locale, while for the latter,
move to copying the i18n locale. This lets us clean up all the stray
comments like FIXME.
This updates all the territory fields based on CLDR v29 data. Many of
them were obviously incorrect where people used a two letter code and
not the English name.
aa_DJ: changing DJ to Djibouti
aa_ER@saaho: changing ER to Eritrea
aa_ER: changing ER to Eritrea
aa_ET: changing ET to Ethiopia
am_ET: changing ET to Ethiopia
ar_LY: changing Libyan Arab Jamahiriya to Libya
ar_SY: changing Syrian Arab Republic to Syria
bo_CN: changing P.R. of China to China
bs_BA: changing Bosnia and Herzegowina to Bosnia & Herzegovina
byn_ER: changing ER to Eritrea
ca_IT: changing Italy (L'Alguer) to Italy
ce_RU: changing RUSSIAN FEDERATION to Russia
cmn_TW: changing Republic of China to Taiwan
cy_GB: changing Great Britain to United Kingdom
de_LU@euro: changing Luxemburg to Luxembourg
de_LU: changing Luxemburg to Luxembourg
en_AG: changing Antigua and Barbuda to Antigua & Barbuda
en_GB: changing Great Britain to United Kingdom
en_HK: changing Hong Kong to Hong Kong SAR China
en_US: changing USA to United States
es_US: changing USA to United States
fr_LU@euro: changing Luxemburg to Luxembourg
fr_LU: changing Luxemburg to Luxembourg
fy_DE: changing DE to Germany
gd_GB: changing Great Britain to United Kingdom
gez_ER@abegede: changing ER to Eritrea
gez_ER: changing ER to Eritrea
gez_ET@abegede: changing ET to Ethiopia
gez_ET: changing ET to Ethiopia
gv_GB: changing Britain to United Kingdom
hak_TW: changing Republic of China to Taiwan
iu_CA: changing CA to Canada
ko_KR: changing Republic of Korea to South Korea
kw_GB: changing Britain to United Kingdom
li_BE: changing BE to Belgium
li_NL: changing NL to Netherlands
lzh_TW: changing Republic of China to Taiwan
my_MM: changing Myanmar to Myanmar (Burma)
nan_TW: changing Republic of China to Taiwan
nds_DE: changing DE to Germany
nds_NL: changing NL to Netherlands
om_ET: changing ET to Ethiopia
om_KE: changing KE to Kenya
pap_AW: changing AW to Aruba
pap_CW: changing CW to Curaçao
pt_BR: changing Brasil to Brazil
sid_ET: changing ET to Ethiopia
sk_SK: changing Slovak to Slovakia
so_DJ: changing DJ to Djibouti
so_ET: changing ET to Ethiopia
so_KE: changing KE to Kenya
so_SO: changing SO to Somalia
ti_ER: changing ER to Eritrea
ti_ET: changing ET to Ethiopia
tig_ER: changing ER to Eritrea
tt_RU@iqtelif: changing Tatarstan, Russian Federation to Russia
uk_UA: changing UA to Ukraine
unm_US: changing USA to United States
wal_ET: changing ET to Ethiopia
yi_US: changing USA to United States
yue_HK: changing Hong Kong to Hong Kong SAR China
zh_CN: changing P.R. of China to China
zh_HK: changing Hong Kong to Hong Kong SAR China
zh_TW: changing Taiwan R.O.C. to Taiwan
This updates all the language fields based on CLDR v29 data. Many of
them were obviously incorrect where people used a two letter code and
not the English name.
aa_DJ: changing aa to Afar
aa_ER: changing aa to Afar
aa_ER@saaho: changing aa to Afar
aa_ET: changing aa to Afar
am_ET: changing am to Amharic
az_AZ: changing Azeri to Azerbaijani
bn_BD: changing Bengali/Bangla to Bengali
byn_ER: changing byn to Blin
de_AT: changing German to Austrian German
de_CH: changing German to Swiss High German
en_AU: changing English to Australian English
en_CA: changing English to Canadian English
en_GB: changing English to British English
en_US: changing English to American English
es_ES: changing Spanish to European Spanish
es_MX: changing Spanish to Mexican Spanish
ff_SN: changing ff to Fulah
fr_CA: changing French to Canadian French
fr_CH: changing French to Swiss French
fur_IT: changing Furlan to Friulian
fy_DE: changing fy to Western Frisian
fy_NL: changing Frisian to Western Frisian
gd_GB: changing Scots Gaelic to Scottish Gaelic
gez_ER@abegede: changing gez to Geez
gez_ER: changing gez to Geez
gez_ET@abegede: changing gez to Geez
gez_ET: changing gez to Geez
gv_GB: changing Manx Gaelic to Manx
ht_HT: changing Kreyol to Haitian Creole
kl_GL: changing Greenlandic to Kalaallisut
lg_UG: changing Luganda to Ganda
li_BE: changing li to Limburgish
li_NL: changing li to Limburgish
nan_TW@latin: changing Minnan to Min Nan Chinese
nb_NO: changing Norwegian, Bokmål to Norwegian Bokmål
nds_DE: changing nds to Low German
nds_NL: changing nds to Low Saxon
niu_NU: changing Vagahau Niue (Niuean) to Niuean
niu_NZ: changing Vagahau Niue (Niuean) to Niuean
nl_BE: changing Dutch to Flemish
nn_NO: changing Norwegian, Nynorsk to Norwegian Nynorsk
nr_ZA: changing Southern Ndebele to South Ndebele
om_ET: changing om to Oromo
om_KE: changing om to Oromo
or_IN: changing Odia to Oriya
os_RU: changing Ossetian to Ossetic
pap_AW: changing pap to Papiamento
pap_CW: changing pap to Papiamento
pa_PK: changing Punjabi (Shahmukhi) to Punjabi
pt_BR: changing Portuguese to Brazilian Portuguese
pt_PT: changing Portuguese to European Portuguese
se_NO: changing Northern Saami to Northern Sami
sid_ET: changing sid to Sidamo
so_DJ: changing so to Somali
so_ET: changing so to Somali
so_KE: changing so to Somali
so_SO: changing so to Somali
st_ZA: changing Sotho to Southern Sotho
sw_KE: changing sw to Swahili
sw_TZ: changing sw to Swahili
ti_ER: changing ti to Tigrinya
ti_ET: changing ti to Tigrinya
tig_ER: changing tig to Tigre
uk_UA: changing uk to Ukrainian
wal_ET: changing wal to Wolaytta
yue_HK: changing Yue Chinese to Cantonese
Mike Frysinger [Mon, 11 Apr 2016 19:52:51 +0000 (15:52 -0400)]
localedata: LC_TIME.date_fmt: delete entries same as the default value
There's no real value in populating this field when it's the same as the
default POSIX setting, so drop it from most locales so it's clear what's
going on.
H.J. Lu [Tue, 12 Apr 2016 15:10:31 +0000 (08:10 -0700)]
X86-64: Use non-temporal store in memcpy on large data
The large memcpy micro benchmark in glibc shows that there is a
regression with large data on Haswell machine. non-temporal store in
memcpy on large data can improve performance significantly. This
patch adds a threshold to use non temporal store which is 6 times of
shared cache size. When size is above the threshold, non temporal
store will be used, but avoid non-temporal store if there is overlap
between destination and source since destination may be in cache when
source is loaded.
For size below 8 vector register width, we load all data into registers
and store them together. Only forward and backward loops, which move 4
vector registers at a time, are used to support overlapping addresses.
For forward loop, we load the last 4 vector register width of data and
the first vector register width of data into vector registers before the
loop and store them after the loop. For backward loop, we load the first
4 vector register width of data and the last vector register width of
data into vector registers before the loop and store them after the loop.
[BZ #19928]
* sysdeps/x86_64/cacheinfo.c (__x86_shared_non_temporal_threshold):
New.
(init_cacheinfo): Set __x86_shared_non_temporal_threshold to 6
times of shared cache size.
* sysdeps/x86_64/multiarch/memmove-avx-unaligned-erms.S
(VMOVNT): New.
* sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S
(VMOVNT): Likewise.
* sysdeps/x86_64/multiarch/memmove-sse2-unaligned-erms.S
(VMOVNT): Likewise.
(VMOVU): Changed to movups for smaller code sizes.
(VMOVA): Changed to movaps for smaller code sizes.
* sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S: Update
comments.
(PREFETCH): New.
(PREFETCH_SIZE): Likewise.
(PREFETCHED_LOAD_SIZE): Likewise.
(PREFETCH_ONE_SET): Likewise.
Rewrite to use forward and backward loops, which move 4 vector
registers at a time, to support overlapping addresses and use
non temporal store if size is above the threshold and there is
no overlap between destination and source.
Matthew Fortune [Tue, 12 Apr 2016 08:31:06 +0000 (09:31 +0100)]
VDSO support for MIPS
This patch adds support for using the implementations of gettimeofday()
and clock_gettime() provided by the kernel in the VDSO. The VDSO will
always provide clock_gettime() as CLOCK_{REALTIME,MONOTONIC}_COARSE can
be implemented regardless of platform. CLOCK_{REALTIME,MONOTONIC}, along
with gettimeofday(), are only implemented on platforms which make use of
either the CP0 count or GIC as their clocksource. On other platforms,
the VDSO does not provide the __vdso_gettimeofday symbol, as it is
never useful.
The VDSO functions return ENOSYS when they encounter an unsupported
request, in which case glibc should fall back to the standard syscall.
Tested with upstream kernel 4.5 and QEMU emulating Malta.
This patch consolidates all the pwrite/pwrite64 implementation for Linux
in only one (sysdeps/unix/sysv/linux/pwrite{64}.c). It also removes the
syscall from the auto-generation using assembly macros.
For pwrite{64} offset argument placement the new SYSCALL_LL{64} macro
is used. For pwrite ports that do not define __NR_pwrite will use
__NR_pwrite64 and for pwrite64 ports that dot define __NR_pwrite64 will
use __NR_pwrite for the syscall.
Checked on x86_64, x32, i386, aarch64, and ppc64le.
This patch consolidates all the pread/pread64 implementation for Linux
in only one (sysdeps/unix/sysv/linux/pread.c). It also removes the
syscall from the auto-generation using assembly macros.
For pread{64} offset argument placement the new SYSCALL_LL{64} macro
is used. For pread ports that do not define __NR_pread will use
__NR_pread64 and for pread64 ports that dot define __NR_pread64 will
use __NR_pread for the syscall.
Checked on x86_64, x32, i386, aarch64, and ppc64le.
This patch add three new macros (SYSCALL_LL, SYSCALL_LL64, and
__ASSUME_WORDSIZE64_ILP32) to use along with off_t and off64_t argument
syscalls. The rationale for this change is:
1. Remove multiple implementations for the same syscall for different
architectures (for instance, pread have 6 different implementations).
2. Also remove the requirement to use syscall wrappers for cancellable
entrypoints.
The macro usage should be used along __ALIGNMENT_ARG to follow ABI constrains
for architecture where it applies. For instance, pread can be rewritten as:
Another macro, SYSCALL_LL64, is provided for off64_t. The macro
__ASSUME_WORDSIZE64_ILP32 is used by the ABI to define is uses 64-bit register
even if ABI is ILP32 (for instance x32 and mips64-n32).
The changes itself are not currently used in any implementation, so no
code change is expected.
Define __ASSUME_ALIGNED_REGISTER_PAIRS for missing ports
This patch defines __ASSUME_ALIGNED_REGISTER_PAIRS for the missing
ports that require 64-bit value (e.g., long long) to be aligned to
an even register pair in argument passing.
No code change is expected, tested with builds for powerpc32,
mips-o32, and armhf.
Samuel Thibault [Sun, 10 Apr 2016 21:58:43 +0000 (23:58 +0200)]
Fix crash on getauxval call without HAVE_AUX_VECTOR
* sysdeps/generic/ldsodefs.h (struct rtld_global_ro)
[!HAVE_AUX_VECTOR]: Do not define _dl_auxv field.
* misc/getauxval.c (__getauxval) [!HAVE_AUX_VECTOR]: Do not go through
GLRO(dl_auxv) list.
Mike Frysinger [Tue, 9 Feb 2016 20:28:52 +0000 (15:28 -0500)]
localedata: CLDRv28: update LC_PAPER values
These locales should be using A4 paper size rather than US-Letter.
Update the copy points to match the others in the file. All other
locales have been verified against the CLDR and hand checking.
Joseph Myers [Fri, 8 Apr 2016 22:52:51 +0000 (22:52 +0000)]
Fix limits.h NL_NMAX namespace (bug 19929).
bits/xopen_lim.h (included by limits.h if __USE_XOPEN) defines
NL_NMAX, but this constant was removed in the 2008 edition of POSIX so
should not be defined in that case. This patch duly disables that
define for __USE_XOPEN2K8. It remains enabled for __USE_GNU to avoid
affecting sysconf (_SC_NL_NMAX), the implementation of which uses
"#ifdef NL_NMAX".
Tested for x86_64 and x86 (testsuite, and that installed stripped
shared libraries are unchanged by the patch).
[BZ #19929]
* include/bits/xopen_lim.h (NL_NMAX): Do not define if
[__USE_XOPEN2K8 && !__USE_GNU].
* conform/Makefile (test-xfail-XOPEN2K8/limits.h/conform): Remove
variable.
Joseph Myers [Fri, 8 Apr 2016 18:16:09 +0000 (18:16 +0000)]
Fix termios.h XCASE namespace (bug 19925).
bits/termios.h (various versions under sysdeps/unix/sysv/linux)
defines XCASE if defined __USE_MISC || defined __USE_XOPEN. This
macro was removed in the 2001 edition of POSIX, and is not otherwise
reserved, so should not be defined for 2001 and later versions of
POSIX. This patch fixes the conditions accordingly (leaving the macro
defined for __USE_MISC, so still in the default namespace).
Tested for x86_64 and x86 (testsuite, and that installed shared
libraries are unchanged by the patch).
Paul E. Murphy [Mon, 14 Mar 2016 21:40:46 +0000 (17:40 -0400)]
powerpc: Add optimized P8 strspn
This utilizes vectors and bitmasks. For small needle, large
haystack, the performance improvement is upto 8x. For short
strings (0-4B), the cost of computing the bitmask dominates,
and is a tad slower.
H.J. Lu [Wed, 6 Apr 2016 17:19:16 +0000 (10:19 -0700)]
X86-64: Prepare memmove-vec-unaligned-erms.S
Prepare memmove-vec-unaligned-erms.S to make the SSE2 version as the
default memcpy, mempcpy and memmove.
* sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S
(MEMCPY_SYMBOL): New.
(MEMPCPY_SYMBOL): Likewise.
(MEMMOVE_CHK_SYMBOL): Likewise.
Replace MEMMOVE_SYMBOL with MEMMOVE_CHK_SYMBOL on __mempcpy_chk
symbols. Replace MEMMOVE_SYMBOL with MEMPCPY_SYMBOL on
__mempcpy symbols. Provide alias for __memcpy_chk in libc.a.
Provide alias for memcpy in libc.a and ld.so.
H.J. Lu [Fri, 1 Apr 2016 22:08:48 +0000 (15:08 -0700)]
Remove Fast_Copy_Backward from Intel Core processors
Intel Core i3, i5 and i7 processors have fast unaligned copy and
copy backward is ignored. Remove Fast_Copy_Backward from Intel Core
processors to avoid confusion.
* sysdeps/x86/cpu-features.c (init_cpu_features): Don't set
bit_arch_Fast_Copy_Backward for Intel Core proessors.
Remove powerpc64 strspn, strcspn, and strpbrk implementation
This patch removes the powerpc64 optimized strspn, strcspn, and
strpbrk assembly implementation now that the default C one
implements the same strategy. On internal glibc benchtests
current implementations shows similar performance with -O2.
With now a faster strcspn implementation, it is faster to just use
it with some return tests than reimplementing strpbrk itself.
As for strcspn optimization, it is generally at least 10 times faster
than the existing implementation on bench-strspn on a few AArch64
implementations.
Also the string/bits/string2.h inlines make no longer sense, as current
implementation will already implement most of the optimizations.
As for strcspn, this patch improves strspn performance using a much
faster algorithm. It first constructs a 256-entry table based on
the accept string and then uses it as a lookup table for the
input string. As for strcspn optimization, it is generally at least
10 times faster than the existing implementation on bench-strspn
on a few AArch64 implementations.
Also the string/bits/string2.h inlines make no longer sense, as current
implementation will already implement most of the optimizations.
Wilco Dijkstra [Fri, 25 Mar 2016 19:44:26 +0000 (16:44 -0300)]
Improve generic strcspn performance
Improve strcspn performance using a much faster algorithm. It is kept simple
so it works well on most targets. It is generally at least 10 times faster
than the existing implementation on bench-strcspn on a few AArch64
implementations, and for some tests 100 times as fast (repeatedly calling
strchr on a small string is extremely slow...).
In fact the string/bits/string2.h inlines make no longer sense, as GCC
already uses strlen if reject is an empty string, strchrnul is 5 times as
fast as __strcspn_c1, while __strcspn_c2 and __strcspn_c3 are slower than
the strcspn main loop for large strings (though reject length 2-4 could be
special cased in the future to gain even more performance).
H.J. Lu [Thu, 31 Mar 2016 17:05:51 +0000 (10:05 -0700)]
Add x86-64 memset with unaligned store and rep stosb
Implement x86-64 memset with unaligned store and rep movsb. Support
16-byte, 32-byte and 64-byte vector register sizes. A single file
provides 2 implementations of memset, one with rep stosb and the other
without rep stosb. They share the same codes when size is between 2
times of vector register size and REP_STOSB_THRESHOLD which defaults
to 2KB.
Key features:
1. Use overlapping store to avoid branch.
2. For size <= 4 times of vector register size, fully unroll the loop.
3. For size > 4 times of vector register size, store 4 times of vector
register size at a time.
H.J. Lu [Thu, 31 Mar 2016 17:04:26 +0000 (10:04 -0700)]
Add x86-64 memmove with unaligned load/store and rep movsb
Implement x86-64 memmove with unaligned load/store and rep movsb.
Support 16-byte, 32-byte and 64-byte vector register sizes. When
size <= 8 times of vector register size, there is no check for
address overlap bewteen source and destination. Since overhead for
overlap check is small when size > 8 times of vector register size,
memcpy is an alias of memmove.
A single file provides 2 implementations of memmove, one with rep movsb
and the other without rep movsb. They share the same codes when size is
between 2 times of vector register size and REP_MOVSB_THRESHOLD which
is 2KB for 16-byte vector register size and scaled up by large vector
register size.
Key features:
1. Use overlapping load and store to avoid branch.
2. For size <= 8 times of vector register size, load all sources into
registers and store them together.
3. If there is no address overlap bewteen source and destination, copy
from both ends with 4 times of vector register size at a time.
4. If address of destination > address of source, backward copy 8 times
of vector register size at a time.
5. Otherwise, forward copy 8 times of vector register size at a time.
6. Use rep movsb only for forward copy. Avoid slow backward rep movsb
by fallbacking to backward copy 8 times of vector register size at a
time.
7. Skip when address of destination == address of source.
Stefan Liebler [Thu, 31 Mar 2016 15:37:16 +0000 (17:37 +0200)]
S390: Extend structs La_s390_regs / La_s390_retval with vector-registers.
Starting with z13, vector registers can also occur as argument registers.
Thus the passed input/output register structs for
la_s390_[32|64]_gnu_plt[enter|exit] functions should reflect those new
registers. This patch extends these structs La_s390_regs and La_s390_retval
and adjusts _dl_runtime_profile() to handle those fields in case of
running on a z13 machine.
Stefan Liebler [Thu, 31 Mar 2016 15:37:16 +0000 (17:37 +0200)]
S390: Save and restore fprs/vrs while resolving symbols.
On s390, no fpr/vrs were saved while resolving a symbol
via _dl_runtime_resolve/_dl_runtime_profile.
According to the abi, the fpr-arguments are defined as call clobbered.
In leaf-functions, gcc 4.9 and newer can use fprs for saving/restoring gprs
instead of saving them to the stack.
If gcc do this in one of the resolver-functions, then the floating point
arguments of a library-function are invalid for the first library-function-call.
Thus, this patch saves/restores the fprs around the resolving code.
The same could occur for vector registers. Furthermore an ifunc-resolver
could also clobber the vector/floating point argument registers.
Thus this patch provides the further variants _dl_runtime_resolve_vx/
_dl_runtime_profile_vx, which are used if the kernel claims, that
we run on a machine with vector registers.
Furthermore, if _dl_runtime_profile calls _dl_call_pltexit,
the pointers to inregs-/outregs-structs were setup invalid.
Now they point to the correct location in the stack-frame.
Before branching back to the caller, the return values are now
restored instead of containing the return values of the
_dl_call_pltexit() call.
On s390-32, an endless loop occurs if _dl_call_pltexit() should be called.
Now, this code-path branches to this function instead of just after the
preceding basr-instruction.
ChangeLog:
* sysdeps/s390/s390-32/dl-trampoline.S: Include dl-trampoline.h twice
to create a non-vector/vector version for _dl_runtime_resolve and
_dl_runtime_profile. Move implementation to ...
* sysdeps/s390/s390-32/dl-trampoline.h: ... here.
(_dl_runtime_resolve) Save and restore fpr/vrs.
(_dl_runtime_profile) Save and restore vrs and fix some issues
if _dl_call_pltexit is called.
* sysdeps/s390/s390-32/dl-machine.h (elf_machine_runtime_setup):
Choose the correct resolver function if running on a machine with vx.
* sysdeps/s390/s390-64/dl-trampoline.S: Include dl-trampoline.h twice
to create a non-vector/vector version for _dl_runtime_resolve and
_dl_runtime_profile. Move implementation to ...
* sysdeps/s390/s390-64/dl-trampoline.h: ... here.
(_dl_runtime_resolve) Save and restore fpr/vrs.
(_dl_runtime_profile) Save and restore vrs and fix some issues
* sysdeps/s390/s390-64/dl-machine.h: (elf_machine_runtime_setup):
Choose the correct resolver function if running on a machine with vx.
Florian Weimer [Thu, 31 Mar 2016 09:26:55 +0000 (11:26 +0200)]
Report dlsym, dlvsym lookup errors using dlerror [BZ #19509]
* elf/dl-lookup.c (_dl_lookup_symbol_x): Report error even if
skip_map != NULL.
* elf/tst-dlsym-error.c: New file.
* elf/Makefile (tests): Add tst-dlsym-error.
(tst-dlsym-error): Link against libdl.
Joseph Myers [Tue, 29 Mar 2016 22:13:36 +0000 (22:13 +0000)]
[microblaze] Remove __ASSUME_FUTIMESAT.
MicroBlaze has a special version of futimesat.c because it gained the
futimesat syscall later than other non-asm-generic architectures. Now
the minimum kernel is recent enough that this syscall can always be
assumed to be present for MicroBlaze, so this patch removes the
special version and the __ASSUME_FUTIMESAT macro, resulting in the
sysdeps/unix/sysv/linux/futimesat.c version being used.
Florian Weimer [Tue, 29 Mar 2016 09:27:32 +0000 (11:27 +0200)]
nss_db: Propagate ERANGE error if parse_line fails [BZ #19837]
Reproducer (needs to run as root):
perl -e \
'print "large:x:999:" . join(",", map {"user$_"} (1 .. 135))."\n"' \
>> /etc/group
cd /var/db
make
getent -s db group
After the fix, the last command should list the "large" group.
The magic number 135 has been chosen so that the line is shorter than
1024 bytes, but the pointers required to encode the member array will
cross the threshold, triggering the bug.
H.J. Lu [Tue, 29 Mar 2016 02:22:59 +0000 (19:22 -0700)]
Initial Enhanced REP MOVSB/STOSB (ERMS) support
The newer Intel processors support Enhanced REP MOVSB/STOSB (ERMS) which
has a feature bit in CPUID. This patch adds the Enhanced REP MOVSB/STOSB
(ERMS) bit to x86 cpu-features.
H.J. Lu [Mon, 28 Mar 2016 20:15:59 +0000 (13:15 -0700)]
Make __memcpy_avx512_no_vzeroupper an alias
Since x86-64 memcpy-avx512-no-vzeroupper.S implements memmove, make
__memcpy_avx512_no_vzeroupper an alias of __memmove_avx512_no_vzeroupper
to reduce code size of libc.so.
* sysdeps/x86_64/multiarch/Makefile (sysdep_routines): Remove
memcpy-avx512-no-vzeroupper.
* sysdeps/x86_64/multiarch/memcpy-avx512-no-vzeroupper.S: Renamed
to ...
* sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S: This.
(MEMCPY): Don't define.
(MEMCPY_CHK): Likewise.
(MEMPCPY): Likewise.
(MEMPCPY_CHK): Likewise.
(MEMPCPY_CHK): Renamed to ...
(__mempcpy_chk_avx512_no_vzeroupper): This.
(MEMPCPY_CHK): Renamed to ...
(__mempcpy_chk_avx512_no_vzeroupper): This.
(MEMCPY_CHK): Renamed to ...
(__memmove_chk_avx512_no_vzeroupper): This.
(MEMCPY): Renamed to ...
(__memmove_avx512_no_vzeroupper): This.
(__memcpy_avx512_no_vzeroupper): New alias.
(__memcpy_chk_avx512_no_vzeroupper): Likewise.
H.J. Lu [Mon, 28 Mar 2016 11:39:48 +0000 (04:39 -0700)]
[x86] Add a feature bit: Fast_Unaligned_Copy
On AMD processors, memcpy optimized with unaligned SSE load is
slower than emcpy optimized with aligned SSSE3 while other string
functions are faster with unaligned SSE load. A feature bit,
Fast_Unaligned_Copy, is added to select memcpy optimized with
unaligned SSE load.
[BZ #19583]
* sysdeps/x86/cpu-features.c (init_cpu_features): Set
Fast_Unaligned_Copy with Fast_Unaligned_Load for Intel
processors. Set Fast_Copy_Backward for AMD Excavator
processors.
* sysdeps/x86/cpu-features.h (bit_arch_Fast_Unaligned_Copy):
New.
(index_arch_Fast_Unaligned_Copy): Likewise.
* sysdeps/x86_64/multiarch/memcpy.S (__new_memcpy): Check
Fast_Unaligned_Copy instead of Fast_Unaligned_Load.
Florian Weimer [Fri, 25 Mar 2016 10:49:51 +0000 (11:49 +0100)]
resolv: Always set *resplen2 out parameter in send_dg [BZ #19791]
Since commit 44d20bca52ace85850012b0ead37b360e3ecd96e (Implement
second fallback mode for DNS requests), there is a code path which
returns early, before *resplen2 is initialized. This happens if the
name server address is immediately recognized as invalid (because of
lack of protocol support, or if it is a broadcast address such
255.255.255.255, or another invalid address).
If this happens and *resplen2 was non-zero (which is the case if a
previous query resulted in a failure), __libc_res_nquery would reuse
an existing second answer buffer. This answer has been previously
identified as unusable (for example, it could be an NXDOMAIN
response). Due to the presence of a second answer, no name server
switching will occur. The result is a name resolution failure,
although a successful resolution would have been possible if name
servers have been switched and queries had proceeded along the search
path.
The above paragraph still simplifies the situation. Before glibc
2.23, if the second answer needed malloc, the stub resolver would
still attempt to reuse the second answer, but this is not possible
because __libc_res_nsearch has freed it, after the unsuccessful call
to __libc_res_nquerydomain, and set the buffer pointer to NULL. This
eventually leads to an assertion failure in __libc_res_nquery:
/* Make sure both hp and hp2 are defined */
assert((hp != NULL) && (hp2 != NULL));
If assertions are disabled, the consequence is a NULL pointer
dereference on the next line.
Starting with glibc 2.23, as a result of commit e9db92d3acfe1822d56d11abcea5bfc4c41cf6ca (CVE-2015-7547: getaddrinfo()
stack-based buffer overflow (Bug 18665)), the second answer is always
allocated with malloc. This means that the assertion failure happens
with small responses as well because there is no buffer to reuse, as
soon as there is a name resolution failure which triggers a search for
an answer along the search path.
This commit addresses the issue by ensuring that *resplen2 is
initialized before the send_dg function returns.
This commit also addresses a bug where an invalid second reply is
incorrectly returned as a valid to the caller.
Joseph Myers [Thu, 24 Mar 2016 01:32:52 +0000 (01:32 +0000)]
Fix x86_64 / x86 powl inaccuracy for integer exponents (bug 19848).
Bug 19848 reports cases where powl on x86 / x86_64 has error
accumulation, for small integer exponents, larger than permitted by
glibc's accuracy goals, at least in some rounding modes. This patch
further restricts the exponent range for which the
small-integer-exponent logic is used to limit the possible error
accumulation.
Tested for x86_64 and x86 and ulps updated accordingly.
[BZ #19848]
* sysdeps/i386/fpu/e_powl.S (p3): Rename to p2 and change value
from 8 to 4.
(__ieee754_powl): Compare integer exponent against 4 not 8.
* sysdeps/x86_64/fpu/e_powl.S (p3): Rename to p2 and change value
from 8 to 4.
(__ieee754_powl): Compare integer exponent against 4 not 8.
* math/auto-libm-test-in: Add more tests of pow.
* math/auto-libm-test-out: Regenerated.
* sysdeps/i386/i686/fpu/multiarch/libm-test-ulps: Update.
* sysdeps/x86_64/fpu/libm-test-ulps: Likewise.
Nick Alcock [Wed, 23 Mar 2016 12:40:14 +0000 (13:40 +0100)]
x86, pthread_cond_*wait: Do not depend on %eax not being clobbered
The x86-specific versions of both pthread_cond_wait and
pthread_cond_timedwait have (in their fall-back-to-futex-wait slow
paths) calls to __pthread_mutex_cond_lock_adjust followed by
__pthread_mutex_unlock_usercnt, which load the parameters before the
first call but then assume that the first parameter, in %eax, will
survive unaffected. This happens to have been true before now, but %eax
is a call-clobbered register, and this assumption is not safe: it could
change at any time, at GCC's whim, and indeed the stack-protector canary
checking code clobbers %eax while checking that the canary is
uncorrupted.
So reload %eax before calling __pthread_mutex_unlock_usercnt. (Do this
unconditionally, even when stack-protection is not in use, because it's
the right thing to do, it's a slow path, and anything else is dicing
with death.)
H.J. Lu [Tue, 22 Mar 2016 14:46:56 +0000 (07:46 -0700)]
Set index_arch_AVX_Fast_Unaligned_Load only for Intel processors
Since only Intel processors with AVX2 have fast unaligned load, we
should set index_arch_AVX_Fast_Unaligned_Load only for Intel processors.
Move AVX, AVX2, AVX512, FMA and FMA4 detection into get_common_indeces
and call get_common_indeces for other processors.
Add CPU_FEATURES_CPU_P and CPU_FEATURES_ARCH_P to aoid loading
GLRO(dl_x86_cpu_features) in cpu-features.c.
[BZ #19583]
* sysdeps/x86/cpu-features.c (get_common_indeces): Remove
inline. Check family before setting family, model and
extended_model. Set AVX, AVX2, AVX512, FMA and FMA4 usable
bits here.
(init_cpu_features): Replace HAS_CPU_FEATURE and
HAS_ARCH_FEATURE with CPU_FEATURES_CPU_P and
CPU_FEATURES_ARCH_P. Set index_arch_AVX_Fast_Unaligned_Load
for Intel processors with usable AVX2. Call get_common_indeces
for other processors with family == NULL.
* sysdeps/x86/cpu-features.h (CPU_FEATURES_CPU_P): New macro.
(CPU_FEATURES_ARCH_P): Likewise.
(HAS_CPU_FEATURE): Use CPU_FEATURES_CPU_P.
(HAS_ARCH_FEATURE): Use CPU_FEATURES_ARCH_P.
Samuel Thibault [Tue, 22 Mar 2016 08:58:48 +0000 (09:58 +0100)]
Fix malloc threaded tests link on non-Linux
* malloc/Makefile ($(objpfx)tst-malloc-backtrace,
$(objpfx)tst-malloc-thread-exit, $(objpfx)tst-malloc-thread-fail): Use
$(shared-thread-library) instead of hardcoding the path to libpthread.
Joseph Myers [Tue, 22 Mar 2016 00:32:20 +0000 (00:32 +0000)]
Remove __ASSUME_GETDENTS64_SYSCALL.
This patch removes the __ASSUME_GETDENTS64_SYSCALL macro, as its
definition is constant given the new kernel version requirements (and
was constant anyway before those requirements except for MIPS n32).
Note that the "#ifdef __NR_getdents64" conditional *is* still needed,
because MIPS n64 only has the getdents syscall (being a 64-bit ABI,
that syscall is 64-bit; the difference between the two on 64-bit
architectures is where d_type goes). If MIPS n64 were to gain the
getdents64 syscall and we wanted to use it conditionally on the kernel
version at runtime we'd have to revert this patch, but I think that's
unlikely (and in any case, we could follow the simpler approach of
undefining __NR_getdents64 if the syscall can't be assumed, just like
we do for accept4 / recvmmsg / sendmmsg syscalls on architectures
where socketcall support came first).
Most of the getdents.c changes are reindentation.
Tested for x86_64 and x86 that installed stripped shared libraries are
unchanged by the patch.
Joseph Myers [Mon, 21 Mar 2016 16:30:05 +0000 (16:30 +0000)]
Remove __ASSUME_SIGNALFD4.
Current Linux kernel version requirements mean the signalfd4 syscall
can always be assumed to be available. This patch removes
__ASSUME_SIGNALFD4 and associated conditionals.
Tested for x86_64 and x86 that installed stripped shared libraries are
unchanged by the patch.
* sysdeps/unix/sysv/linux/kernel-features.h (__ASSUME_SIGNALFD4):
Remove macro.
* sysdeps/unix/sysv/linux/signalfd.c: Do not include
<kernel-features.h>.
(signalfd) [__NR_signalfd4]: Make code unconditional.
(signalfd) [!__ASSUME_SIGNALFD4]: Remove conditional code.
H.J. Lu [Mon, 21 Mar 2016 13:38:17 +0000 (06:38 -0700)]
Use JUMPTARGET in x86-64 pthread
When PLT may be used, JUMPTARGET should be used instead calling the
function directly.
* sysdeps/unix/sysv/linux/x86_64/cancellation.S
(__pthread_enable_asynccancel): Use JUMPTARGET to call
__pthread_unwind.
* sysdeps/unix/sysv/linux/x86_64/pthread_cond_timedwait.S
(__condvar_cleanup2): Use JUMPTARGET to call _Unwind_Resume.
* sysdeps/unix/sysv/linux/x86_64/pthread_cond_wait.S
(__condvar_cleanup1): Likewise.
Dylan Alex Simon [Tue, 15 Mar 2016 17:20:01 +0000 (13:20 -0400)]
math: don't clobber old libm.so on install [BZ #19822]
When installing glibc (w/mathvec enabled) in-place on a system with
a glibc w/out mathvec enabled, the install will clobber the existing
libm.so (e.g., /lib64/libm-2.21.so) with a linker script. This is
because libm.so is a symlink to libm.so.6 which is a symlink to the
final libm-2.21.so file. When the makefile writes the linker script
directly to libm.so, it gets clobbered.
The simple patch below to math/Makefile fixes this. It is based on
the nptl Makefile, which does exactly the same thing in a safer way.
Samuel Thibault [Sun, 20 Mar 2016 16:56:47 +0000 (17:56 +0100)]
hurd: Do not hide rtld symbols which need to be preempted
* sysdeps/generic/dl-fcntl.h: New file, adds attribute_hidden to __open
and __fcntl.
* sysdeps/mach/hurd/dl-fcntl.h: New file, adds attribute_hidden to
__fcntl only.
* include/fcntl.h [IS_IN (rtld)]: Include <dl-fcntl.h> instead of
adding attribute_hidden to __open and __fcntl.
Generating errnos.d does not actually need libc-modules.h.
* sysdeps/mach/hurd/Makefile ($(common-objpfx)errnos.d): Strip
"-include $(common-objpfx)libc-modules.h" from CPPFLAGS, and do not
depend on libc-modules.h,
Joseph Myers [Thu, 17 Mar 2016 19:07:39 +0000 (19:07 +0000)]
Remove __ASSUME_EVENTFD2, move eventfd to syscalls.list.
Given current Linux kernel version requirements, we can assume the
presence of the eventfd2 syscall. This means that __ASSUME_EVENTFD2
can be removed, and a syscalls.list entry suffices for eventfd instead
of needing a .c file. This patch implements those changes.
Tested for x86_64 and x86 (not that that means much, given the lack of
testsuite coverage for eventfd).
Joseph Myers [Thu, 17 Mar 2016 12:15:51 +0000 (12:15 +0000)]
Remove __ASSUME_FALLOCATE.
Given current Linux kernel version requirements, we can always assume
the fallocate syscall to be available. This patch removes
__ASSUME_FALLOCATE and a test for whether __NR_fallocate is defined.
Tested for x86_64 and x86 that installed stripped shared libraries are
unchanged by the patch.
* sysdeps/unix/sysv/linux/kernel-features.h (__ASSUME_FALLOCATE):
Remove macro.
* sysdeps/unix/sysv/linux/wordsize-64/posix_fallocate.c: Do not
include <kernel-features.h>.
[!__ASSUME_FALLOCATE]: Remove conditional code.
(posix_fallocate) [__NR_fallocate]: Make code unconditional.