devel

devel@lists.centos.org

5 participants
4641 discussions

Sharing Software Collections over NFS?
by Honza Horak 09 Feb '18

09 Feb '18

It's been a while since I wrote an article about a cool feature that we introduced together with Software Collections packaging format, which is *the ability to share the environment on multiple machines by NFS sharing *[1]. The idea is very simple -- install Software Collections of your favorite application stack (say Python 3.6 + NodeJS 8) and mount the /opt/rh on workstations. That way you have the same environment on all machines. Now, I'd like to request some feedback -- is there anybody on the list (or outside, but you at least know about it) who uses this feature for the Software Collections packages? Please, let us know! If you can't share it publicly, just drop me a direct mail. Of course, more feedback is welcome, but even simple YES would be also cool. [1] https://developers.redhat.com/blog/2015/06/29/maintain-software-collections… Thanks! Honza

3 2

Re: [CentOS-devel] [scl.org] Sharing Software Collections over NFS?
by Honza Horak 09 Feb '18

09 Feb '18

On 02/09/2018 04:23 PM, Marek Salwerowicz wrote: > Hi Honza, > > On 09.02.2018 16:03, Honza Horak wrote: > >> It's been a while since I wrote an article about a cool feature that >> we introduced together with Software Collections packaging format, >> which is *the ability to share the environment on multiple machines by >> NFS sharing *[1]. >> >> The idea is very simple -- install Software Collections of your >> favorite application stack (say Python 3.6 + NodeJS 8) and mount the >> /opt/rh on workstations. That way you have the same environment on all >> machines. > > Sounds nice, but how about config files that are installed in /etc/opt ? That's tricky, yes, and it's also a reason why we don't do it for stuff like databases (data files are even more tricky). This feature is used only for dynamic languages like python, ruby, nodejs, ... Honza > Cheers, > Marek > >> >> Now, I'd like to request some feedback -- is there anybody on the list >> (or outside, but you at least know about it) who uses this feature for >> the Software Collections packages? >> >> Please, let us know! If you can't share it publicly, just drop me a >> direct mail. Of course, more feedback is welcome, but even simple YES >> would be also cool. >> >> [1] >> https://developers.redhat.com/blog/2015/06/29/maintain-software-collections… >> >> >> Thanks! >> Honza >> >> _______________________________________________ >> SCLorg mailing list >> SCLorg(a)redhat.com >> https://www.redhat.com/mailman/listinfo/sclorg >

3 2

sclo-php-pecl-xdebug 2.6.0 and php-pecl-mongodb 1.4.0 packages available for testing
by Remi Collet 09 Feb '18

09 Feb '18

Updated packages available in centos-sclo-sclo-testing For rh-php56 * sclo-php56-php-pecl-mongodb-1.4.0-1.el6 For rh-php70 * sclo-php70-php-pecl-mongodb-1.4.0-1.el6 * sclo-php70-php-pecl-xdebug-2.6.0-1.el6 For rh-php71 * sclo-php70-php-pecl-mongodb-1.4.0-1.el7 * sclo-php70-php-pecl-xdebug-2.6.0-1.el7 mongodb 1.4.0 with bug fixes and new features See upstream changelog https://pecl.php.net/package-changelog.php?package=mongodb WARNING: The mongodb extension add support for MongoDB 3.6 but drop support for MongoDB 2.4 XDebug 2.6.0 with bug fixes and new features See upstream changelog https://pecl.php.net/package-changelog.php?package=xdebug And author blog post https://derickrethans.nl/xdebug-26.html Feedback welcome. Remi. P.S. https://wiki.centos.org/SpecialInterestGroup/SCLo

1 0

Re: [CentOS-devel] Adjust type of rw in submit_bio from int to unsigned long
by wuzhouhui 09 Feb '18

09 Feb '18

Forward mail to centos-devel mailing list for getting more attentions! Sorry for disturbing. > -----Original Messages----- > From: wuzhouhui <wuzhouhui14(a)mails.ucas.ac.cn> > Sent Time: 2018-02-07 13:44:07 (Wednesday) > To: centos(a)centos.org > Cc: wuzhouhui14 <wuzhouhui14(a)mails.ucas.ac.cn> > Subject: Adjust type of rw in submit_bio from int to unsigned long > > Hi, I am a developer who worked on CentOS. Recently I am working on porting one of my > block drivers from CentOS 6.x to CentOS 7.x. In newest kernel (kernel-3.10.0-693.17.1.el7) > I found there is an issue in submit_bio()'s first argument: > void submit_bio(int rw, struct bio *bio) > The type of bi_rw of struct bio is unsigned long, and the number of enum rq_flags_bits > also exceeds 32, so we should better using unsigned long for rw in submit_bio(), like: > void submit_bio(unsigned long rw, struct bio *bio) > >

1 0

intro of me
by Daan Hoogland 07 Feb '18

07 Feb '18

H, I am Daan Hoogland. A fellow Apache Cloudstack PMC meber of Rohit, who you already met. I hope to get the cloudstack dirstibution into centos as a default rpm. Itried to subscribe to the cloud-sig at centos.org but couldn't find it yet. less intentional me: I develop java, python and hobby with remacs (rust/elisp) regards, -- Daan

2 1

[CentOS-dev][NFV SIG] Informal discussion at Centos Dojo
by Thomas F Herbert 02 Feb '18

02 Feb '18

I am at Centos Dojo. For those at Centos Dojo that are interested in a side discussion about Centos NFV SIG, let me know. -- *Thomas F Herbert* NFV and Fast Data Planes Networking Group Office of the CTO *Red Hat*

1 0

[RDO] Queens Milestone 3 test day, Feb 8, 9
by Rich Bowen 30 Jan '18

30 Jan '18

We will be testing the RDO OpenStack Queens Milestone 3 packages on February 8th and 9th. Details are at http://rdoproject.org/testday/queens/milestone3/ We will be congregating on the #RDO channel on Freenode IRC for questions, commentary, and so on. We will also be deploying a test cloud, like last time, so that you can test without having to deploy it yourself. Details, and signups, are at https://etherpad.openstack.org/p/rdo-queens-m3-cloud If you have any questions, please bring them to the users(a)lists.rdoproject.org mailing list. You can subscribe to this list at http://lists.rdoproject.org/ -- Rich Bowen: Community Architect rbowen(a)redhat.com @rbowen // @RDOCommunity // @CentOSProject 1 859 351 9166

1 0

Re: [CentOS-devel] [PATCH] Fix memory leak in i18n support for sort utility
by Yousong Zhou 30 Jan '18

30 Jan '18

On 30 January 2018 at 15:58, Kamil Dudka <kdudka(a)redhat.com> wrote: > On Tuesday, January 30, 2018 3:01:54 AM CET Yousong Zhou wrote: >> On 30 January 2018 at 02:58, Kamil Dudka <kdudka(a)redhat.com> wrote: >> > On Wednesday, January 24, 2018 1:18:35 PM CET Yousong Zhou wrote: >> >> This can be triggered by using month sort >> >> >> >> The issue with the original code is that after the call to mbsrtowcs() >> >> and wcsrtombs() returns, the source pointer will be set to NULL causing >> >> the later free() call has no effect at all >> >> --- >> >> The issue also exists in c6 branch >> > >> > Thank you for sharing the patch! This was tracked as Fedora bug #1259942: >> > >> > https://bugzilla.redhat.com/1259942 >> > >> > The bug was fixed in coreutils-8.23-11.fc22 and it is going to be fixed >> > in the next major version of RHEL (and CentOS). >> > >> > Kamil >> >> Hi Kamil. I thought there must exist some kind of mechanism for >> backporting bugfixes of this kind into currently maintained versions. >> It should! It's an old bug already known and fixed in 2015 and still >> people are wasting time and getting frustrated by it in 2018! >> >> yousong > > Sorry to hear that. We do not backport non-critical fixes from Fedora to RHEL > proactively. Updates in RHEL are primarily driven by requests of Red Hat's > customers. > > I have cloned the bug for RHEL-7 to get this fixed eventually: > > https://bugzilla.redhat.com/1540059 > > Unfortunately, I cannot guarantee any schedule for merging the fix. If you > are a customer of Red Hat, please contact the product support and tell them > about this bug. > > Kamil Thanks for doing this. I encountered this bug when I was using sort to --merge multiple log files with month fields of the form Jan, Feb etc. It caused the kernel to oom which was unexpected since the very merge sort should run in linear time complexity with fixed amount of memory. yousong

1 0

Re: [CentOS-devel] [PATCH] Fix memory leak in i18n support for sort utility
by Yousong Zhou 30 Jan '18

30 Jan '18

On 30 January 2018 at 02:58, Kamil Dudka <kdudka(a)redhat.com> wrote: > On Wednesday, January 24, 2018 1:18:35 PM CET Yousong Zhou wrote: >> This can be triggered by using month sort >> >> The issue with the original code is that after the call to mbsrtowcs() >> and wcsrtombs() returns, the source pointer will be set to NULL causing >> the later free() call has no effect at all >> --- >> The issue also exists in c6 branch > > Thank you for sharing the patch! This was tracked as Fedora bug #1259942: > > https://bugzilla.redhat.com/1259942 > > The bug was fixed in coreutils-8.23-11.fc22 and it is going to be fixed > in the next major version of RHEL (and CentOS). > > Kamil Hi Kamil. I thought there must exist some kind of mechanism for backporting bugfixes of this kind into currently maintained versions. It should! It's an old bug already known and fixed in 2015 and still people are wasting time and getting frustrated by it in 2018! yousong > >> SOURCES/coreutils-i18n.patch | 45 >> ++++++++++++++++++++++---------------------- 1 file changed, 22 >> insertions(+), 23 deletions(-) >> >> diff --git a/SOURCES/coreutils-i18n.patch b/SOURCES/coreutils-i18n.patch >> index a3c7b33..cdcaffd 100644 >> --- a/SOURCES/coreutils-i18n.patch >> +++ b/SOURCES/coreutils-i18n.patch >> @@ -2963,7 +2963,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c line->keybeg = line_start; >> } >> } >> -@@ -1945,7 +2291,7 @@ human_numcompare (char const *a, char co >> +@@ -1943,7 +2289,7 @@ human_numcompare (char *a, char *b) >> hideously fast. */ >> >> static int >> @@ -2972,7 +2972,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c { >> while (blanks[to_uchar (*a)]) >> a++; >> -@@ -1955,6 +2301,25 @@ numcompare (char const *a, char const *b >> +@@ -1953,6 +2299,25 @@ numcompare (char const *a, char const *b >> return strnumcmp (a, b, decimal_point, thousands_sep); >> } >> >> @@ -2998,7 +2998,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c /* Work around a problem whereby the long double >> value returned by glibc's strtold ("NaN", ...) contains uninitialized bits: >> clear all bytes of A and B before calling strtold. FIXME: remove this >> function once -@@ -2005,7 +2370,7 @@ general_numcompare (char const *sa, >> char >> +@@ -2003,7 +2368,7 @@ general_numcompare (char const *sa, char >> Return 0 if the name in S is not recognized. */ >> >> static int >> @@ -3007,7 +3007,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c { >> size_t lo = 0; >> size_t hi = MONTHS_PER_YEAR; >> -@@ -2280,15 +2645,14 @@ debug_key (struct line const *line, stru >> +@@ -2278,15 +2643,14 @@ debug_key (struct line const *line, stru >> char saved = *lim; >> *lim = '\0'; >> >> @@ -3025,7 +3025,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c else if (key->general_numeric) >> ignore_value (strtold (beg, &tighter_lim)); >> else if (key->numeric || key->human_numeric) >> -@@ -2432,7 +2796,7 @@ key_warnings (struct keyfield const *gke >> +@@ -2430,7 +2794,7 @@ key_warnings (struct keyfield const *gke >> bool maybe_space_aligned = !hard_LC_COLLATE && default_key_compare >> (key) && !(key->schar || key->echar); bool line_offset = key->eword == 0 && >> key->echar != 0; /* -k1.x,1.y */ @@ -3034,7 +3034,7 @@ diff -urNp >> coreutils-8.22-orig/src/sort.c coreutils-8.22/src/sort.c && >> ((!key->skipsblanks && !(implicit_skip || maybe_space_aligned)) >> || (!key->skipsblanks && key->schar) >> || (!key->skipeblanks && key->echar))) >> >> -@@ -2490,11 +2854,87 @@ key_warnings (struct keyfield const *gke >> +@@ -2488,11 +2852,86 @@ key_warnings (struct keyfield const *gke >> error (0, 0, _("option '-r' only applies to last-resort comparison")); >> } >> >> @@ -3047,8 +3047,8 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c + register int lo = 0, hi = MONTHS_PER_YEAR, >> result; >> + char *tmp; >> + size_t wclength, mblength; >> -+ const char **pp; >> -+ const wchar_t **wpp; >> ++ const char *pp; >> ++ const wchar_t *wpp; >> + wchar_t *month_wcs; >> + mbstate_t state; >> + >> @@ -3066,12 +3066,12 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c + tmp = (char *) xmalloc (len + 1); >> + memcpy (tmp, s, len); >> + tmp[len] = '\0'; >> -+ pp = (const char **)&tmp; >> + month_wcs = (wchar_t *) xmalloc ((len + 1) * sizeof (wchar_t)); >> + memset (&state, '\0', sizeof(mbstate_t)); >> + >> -+ wclength = mbsrtowcs (month_wcs, pp, len + 1, &state); >> -+ if (wclength == (size_t)-1 || *pp != NULL) >> ++ pp = tmp; >> ++ wclength = mbsrtowcs (month_wcs, &pp, len + 1, &state); >> ++ if (wclength == (size_t)-1 || pp != NULL) >> + error (SORT_FAILURE, 0, _("Invalid multibyte input %s."), quote(s)); >> + >> + for (i = 0; i < wclength; i++) >> @@ -3084,10 +3084,9 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c + } >> + } >> + >> -+ wpp = (const wchar_t **)&month_wcs; >> -+ >> -+ mblength = wcsrtombs (month, wpp, len + 1, &state); >> -+ assert (mblength != (-1) && *wpp == NULL); >> ++ wpp = month_wcs; >> ++ mblength = wcsrtombs (month, &wpp, len + 1, &state); >> ++ assert (mblength != (-1) && wpp == NULL); >> + >> + do >> + { >> @@ -3123,7 +3122,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c { >> struct keyfield *key = keylist; >> >> -@@ -2579,7 +3019,7 @@ keycompare (struct line const *a, struct >> +@@ -2577,7 +3016,7 @@ keycompare (struct line const *a, struct >> else if (key->human_numeric) >> diff = human_numcompare (ta, tb); >> else if (key->month) >> @@ -3132,7 +3131,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c else if (key->random) >> diff = compare_random (ta, tlena, tb, tlenb); >> else if (key->version) >> -@@ -2695,6 +3135,209 @@ keycompare (struct line const *a, struct >> +@@ -2693,6 +3132,209 @@ keycompare (struct line const *a, struct >> return key->reverse ? -diff : diff; >> } >> >> @@ -3342,7 +3341,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c /* Compare two lines A and B, returning negative, >> zero, or positive depending on whether A compares less than, equal to, or >> greater than B. */ >> >> -@@ -2722,7 +3365,7 @@ compare (struct line const *a, struct li >> +@@ -2720,7 +3362,7 @@ compare (struct line const *a, struct li >> diff = - NONZERO (blen); >> else if (blen == 0) >> diff = 1; >> @@ -3351,7 +3350,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c { >> /* Note xmemcoll0 is a performance enhancement as >> it will not unconditionally write '\0' after the >> -@@ -4113,6 +4756,7 @@ set_ordering (char const *s, struct keyf >> +@@ -4111,6 +4753,7 @@ set_ordering (char const *s, struct keyf >> break; >> case 'f': >> key->translate = fold_toupper; >> @@ -3359,7 +3358,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c break; >> case 'g': >> key->general_numeric = true; >> -@@ -4190,7 +4834,7 @@ main (int argc, char **argv) >> +@@ -4188,7 +4831,7 @@ main (int argc, char **argv) >> initialize_exit_failure (SORT_FAILURE); >> >> hard_LC_COLLATE = hard_locale (LC_COLLATE); >> @@ -3368,7 +3367,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c hard_LC_TIME = hard_locale (LC_TIME); >> #endif >> >> -@@ -4211,6 +4855,29 @@ main (int argc, char **argv) >> +@@ -4209,6 +4852,29 @@ main (int argc, char **argv) >> thousands_sep = -1; >> } >> >> @@ -3398,7 +3397,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c have_read_stdin = false; >> inittables (); >> >> -@@ -4485,13 +5152,34 @@ main (int argc, char **argv) >> +@@ -4483,13 +5149,34 @@ main (int argc, char **argv) >> >> case 't': >> { >> @@ -3437,7 +3436,7 @@ diff -urNp coreutils-8.22-orig/src/sort.c >> coreutils-8.22/src/sort.c else >> { >> /* Provoke with 'sort -txx'. Complain about >> -@@ -4502,9 +5190,12 @@ main (int argc, char **argv) >> +@@ -4500,9 +5187,12 @@ main (int argc, char **argv) >> quote (optarg)); >> } >> }

1 0

IBM ppc64{le} builders offline for CBS/koji
by Fabian Arrotin 29 Jan '18

29 Jan '18

Hi, Our monitoring platform informed us that the ppc64{le} builders used in our Koji environment are actually unreachable (same for the underlying IBM P8 hypervisor node) After investigation it seems that it's hardware issue I'll work with the DC people next monday to see how we can try to fix the issue and I'll also investigate some options to have at least one builder per arch asap (so ppc64 and ppc64le) That means that people building for multiple arches will be impacted as long as those builders will not be back online (or some equivalent node that can take those $arch jobs) I'll keep the list informed about a status update as soon as possible. -- Fabian Arrotin The CentOS Project | https://www.centos.org gpg key: 56BEC54E | twitter: @arrfab

1 2

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

devel