Re: update misc/memcached to 1.5.14
On 2019/05/13 21:13, Giovanni Bechis wrote: > On Mon, May 13, 2019 at 12:24:29PM +0100, Stuart Henderson wrote: > > On 2019/05/13 12:23, Stuart Henderson wrote: > > > On 2019/05/13 09:02, Giovanni Bechis wrote: > > > > On 5/13/19 8:42 AM, David CARLIER wrote: > > > > > If you mention testapp executable it does not coredump. No worries if > > > > > you wanna keep the no_drop_privileges patch tho. > > > > > Regards. > > > > > > > > > > On Mon, 13 May 2019 at 07:32, Giovanni Bechis > > > > > wrote: > > > > >> > > > > >> On Sun, May 12, 2019 at 01:08:51PM +0100, David CARLIER wrote: > > > > >>> Hi, > > > > >>> > > > > >>> Here a proposal straight to 1.5.14, removing a patch due to > > > > >>> drop_privilege support feature. > > > > >>> > > > > >> why you removed that patch ? I think in OpenBSD dropping privileges > > > > >> should be > > > > >> mandatory. > > > > >> Does it pass regression tests (at least the same tests that passes > > > > >> with 1.5.6) ? > > > > >> I have a 1.5.12 port on my tree that core dumps during regression > > > > >> tests, I haven't > > > > >> checked 1.5.14 yet. > > > > >> Giovanni > > > > > > > > > New diff attached, I will commit it in a few days. > > > > > > Maybe worth adding --enable-tls (and syncing WANTLIB)? > > > > > > OK either way. > > > > > > > Oh actually... hold for a little while, I want to make sure it still builds > > on arm64. > > > I do not know if it's useful or not, there is a "--enable-arm-crc32 (ARMv8 > CRC32 instructions)" > option as well in configure. > Giovanni > Yep that is why I wanted to check, in case it autodetected it and triggered breakage there. Confirmed it builds, so I am OK to go ahead (either with or without --enable-tls).
Re: update misc/memcached to 1.5.14
On Mon, May 13, 2019 at 12:24:29PM +0100, Stuart Henderson wrote: > On 2019/05/13 12:23, Stuart Henderson wrote: > > On 2019/05/13 09:02, Giovanni Bechis wrote: > > > On 5/13/19 8:42 AM, David CARLIER wrote: > > > > If you mention testapp executable it does not coredump. No worries if > > > > you wanna keep the no_drop_privileges patch tho. > > > > Regards. > > > > > > > > On Mon, 13 May 2019 at 07:32, Giovanni Bechis > > > > wrote: > > > >> > > > >> On Sun, May 12, 2019 at 01:08:51PM +0100, David CARLIER wrote: > > > >>> Hi, > > > >>> > > > >>> Here a proposal straight to 1.5.14, removing a patch due to > > > >>> drop_privilege support feature. > > > >>> > > > >> why you removed that patch ? I think in OpenBSD dropping privileges > > > >> should be > > > >> mandatory. > > > >> Does it pass regression tests (at least the same tests that passes > > > >> with 1.5.6) ? > > > >> I have a 1.5.12 port on my tree that core dumps during regression > > > >> tests, I haven't > > > >> checked 1.5.14 yet. > > > >> Giovanni > > > > > > > New diff attached, I will commit it in a few days. > > > > Maybe worth adding --enable-tls (and syncing WANTLIB)? > > > > OK either way. > > > > Oh actually... hold for a little while, I want to make sure it still builds > on arm64. > I do not know if it's useful or not, there is a "--enable-arm-crc32 (ARMv8 CRC32 instructions)" option as well in configure. Giovanni
Re: update misc/memcached to 1.5.14
On 2019/05/13 12:23, Stuart Henderson wrote: > On 2019/05/13 09:02, Giovanni Bechis wrote: > > On 5/13/19 8:42 AM, David CARLIER wrote: > > > If you mention testapp executable it does not coredump. No worries if > > > you wanna keep the no_drop_privileges patch tho. > > > Regards. > > > > > > On Mon, 13 May 2019 at 07:32, Giovanni Bechis wrote: > > >> > > >> On Sun, May 12, 2019 at 01:08:51PM +0100, David CARLIER wrote: > > >>> Hi, > > >>> > > >>> Here a proposal straight to 1.5.14, removing a patch due to > > >>> drop_privilege support feature. > > >>> > > >> why you removed that patch ? I think in OpenBSD dropping privileges > > >> should be > > >> mandatory. > > >> Does it pass regression tests (at least the same tests that passes with > > >> 1.5.6) ? > > >> I have a 1.5.12 port on my tree that core dumps during regression tests, > > >> I haven't > > >> checked 1.5.14 yet. > > >> Giovanni > > > > > New diff attached, I will commit it in a few days. > > Maybe worth adding --enable-tls (and syncing WANTLIB)? > > OK either way. > Oh actually... hold for a little while, I want to make sure it still builds on arm64.
Re: update misc/memcached to 1.5.14
On 2019/05/13 09:02, Giovanni Bechis wrote: > On 5/13/19 8:42 AM, David CARLIER wrote: > > If you mention testapp executable it does not coredump. No worries if > > you wanna keep the no_drop_privileges patch tho. > > Regards. > > > > On Mon, 13 May 2019 at 07:32, Giovanni Bechis wrote: > >> > >> On Sun, May 12, 2019 at 01:08:51PM +0100, David CARLIER wrote: > >>> Hi, > >>> > >>> Here a proposal straight to 1.5.14, removing a patch due to > >>> drop_privilege support feature. > >>> > >> why you removed that patch ? I think in OpenBSD dropping privileges should > >> be > >> mandatory. > >> Does it pass regression tests (at least the same tests that passes with > >> 1.5.6) ? > >> I have a 1.5.12 port on my tree that core dumps during regression tests, I > >> haven't > >> checked 1.5.14 yet. > >> Giovanni > > > New diff attached, I will commit it in a few days. Maybe worth adding --enable-tls (and syncing WANTLIB)? OK either way.
Re: update misc/memcached to 1.5.14
On Sun, May 12, 2019 at 01:08:51PM +0100, David CARLIER wrote: > Hi, > > Here a proposal straight to 1.5.14, removing a patch due to > drop_privilege support feature. > why you removed that patch ? I think in OpenBSD dropping privileges should be mandatory. Does it pass regression tests (at least the same tests that passes with 1.5.6) ? I have a 1.5.12 port on my tree that core dumps during regression tests, I haven't checked 1.5.14 yet. Giovanni
Re: update misc/memcached to 1.5.14
On 5/13/19 8:42 AM, David CARLIER wrote: > If you mention testapp executable it does not coredump. No worries if > you wanna keep the no_drop_privileges patch tho. > Regards. > > On Mon, 13 May 2019 at 07:32, Giovanni Bechis wrote: >> >> On Sun, May 12, 2019 at 01:08:51PM +0100, David CARLIER wrote: >>> Hi, >>> >>> Here a proposal straight to 1.5.14, removing a patch due to >>> drop_privilege support feature. >>> >> why you removed that patch ? I think in OpenBSD dropping privileges should be >> mandatory. >> Does it pass regression tests (at least the same tests that passes with >> 1.5.6) ? >> I have a 1.5.12 port on my tree that core dumps during regression tests, I >> haven't >> checked 1.5.14 yet. >> Giovanni > New diff attached, I will commit it in a few days. Cheers Giovanni Index: Makefile === RCS file: /var/cvs/ports/misc/memcached/Makefile,v retrieving revision 1.45 diff -u -p -r1.45 Makefile --- Makefile 4 Sep 2018 12:46:16 - 1.45 +++ Makefile 13 May 2019 06:26:23 - @@ -2,9 +2,8 @@ COMMENT= distributed memory object caching system -DISTNAME= memcached-1.5.6 +DISTNAME= memcached-1.5.14 CATEGORIES= misc -REVISION= 0 HOMEPAGE= https://www.memcached.org/ Index: distinfo === RCS file: /var/cvs/ports/misc/memcached/distinfo,v retrieving revision 1.27 diff -u -p -r1.27 distinfo --- distinfo 3 Mar 2018 13:39:18 - 1.27 +++ distinfo 13 May 2019 06:26:23 - @@ -1,2 +1,2 @@ -SHA256 (memcached-1.5.6.tar.gz) = lnXuhZ19gfepUPGQpoEnILJvCCKNNWBE7FF9TVryXwM= -SIZE (memcached-1.5.6.tar.gz) = 452038 +SHA256 (memcached-1.5.14.tar.gz) = nFvfKaeA+2xvfJ6q7toFg+/fZjGTdYw+MWyWmlEK8qk= +SIZE (memcached-1.5.14.tar.gz) = 485076 Index: patches/patch-configure === RCS file: /var/cvs/ports/misc/memcached/patches/patch-configure,v retrieving revision 1.12 diff -u -p -r1.12 patch-configure --- patches/patch-configure 3 Mar 2018 13:39:18 - 1.12 +++ patches/patch-configure 13 May 2019 06:57:11 - @@ -2,7 +2,7 @@ $OpenBSD: patch-configure,v 1.12 2018/03 Index: configure --- configure.orig +++ configure -@@ -6575,7 +6575,7 @@ $as_echo "#define _GNU_SOURCE 1" >>confdefs.h +@@ -6818,7 +6818,7 @@ $as_echo "#define _GNU_SOURCE 1" >>confdefs.h elif test "$GCC" = "yes" then GCC_VERSION=`$CC -dumpversion` Index: patches/patch-memcached_c === RCS file: /var/cvs/ports/misc/memcached/patches/patch-memcached_c,v retrieving revision 1.16 diff -u -p -r1.16 patch-memcached_c --- patches/patch-memcached_c 3 Mar 2018 13:39:18 - 1.16 +++ patches/patch-memcached_c 13 May 2019 06:57:58 - @@ -1,15 +1,16 @@ -$OpenBSD: patch-memcached_c,v 1.16 2018/03/03 13:39:18 sthen Exp $ +$OpenBSD$ Index: memcached.c --- memcached.c.orig +++ memcached.c -@@ -7746,6 +7746,9 @@ int main (int argc, char **argv) { - /* Drop privileges no longer needed */ - if (settings.drop_privileges) { - drop_privileges(); -+} else { -+fprintf(stderr, "no_drop_privileges option unsupported on this operating system.\n"); -+exit(EXIT_FAILURE); - } - - /* Initialize the uriencode lookup table. */ +@@ -7719,8 +7719,8 @@ int main (int argc, char **argv) { + start_lru_maintainer = false; + break; + case NO_DROP_PRIVILEGES: +-settings.drop_privileges = false; +-break; ++ fprintf(stderr, "no_drop_privileges option unsupported on this operating system.\n"); ++ return 1; + case DROP_PRIVILEGES: + settings.drop_privileges = true; + break; Index: pkg/PLIST === RCS file: /var/cvs/ports/misc/memcached/pkg/PLIST,v retrieving revision 1.12 diff -u -p -r1.12 PLIST --- pkg/PLIST 4 Sep 2018 12:46:16 - 1.12 +++ pkg/PLIST 13 May 2019 06:26:23 - @@ -1,6 +1,7 @@ @comment $OpenBSD: PLIST,v 1.12 2018/09/04 12:46:16 espie Exp $ @newgroup _memcached:637 @newuser _memcached:637:_memcached:daemon:memcache daemon:/var/empty:/sbin/nologin +@rcscript ${RCDIR}/memcached @bin bin/memcached bin/memcached-tool include/memcached/ @@ -13,5 +14,5 @@ share/doc/memcached/protocol.txt share/doc/memcached/readme.txt share/doc/memcached/storage.txt share/doc/memcached/threads.txt +share/doc/memcached/tls.txt share/doc/pkg-readmes/${PKGSTEM} -@rcscript ${RCDIR}/memcached
Re: update misc/memcached to 1.5.14
If you mention testapp executable it does not coredump. No worries if you wanna keep the no_drop_privileges patch tho. Regards. On Mon, 13 May 2019 at 07:32, Giovanni Bechis wrote: > > On Sun, May 12, 2019 at 01:08:51PM +0100, David CARLIER wrote: > > Hi, > > > > Here a proposal straight to 1.5.14, removing a patch due to > > drop_privilege support feature. > > > why you removed that patch ? I think in OpenBSD dropping privileges should be > mandatory. > Does it pass regression tests (at least the same tests that passes with > 1.5.6) ? > I have a 1.5.12 port on my tree that core dumps during regression tests, I > haven't > checked 1.5.14 yet. > Giovanni
Re: UPDATE: misc/memcached
On Sun, Dec 17 2017, Giovanni Bechiswrote: [...] > correct, what about that then ? LGTM -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
Re: UPDATE: misc/memcached
On Sun, Dec 17, 2017 at 01:24:30PM +0100, Jeremie Courreges-Anglas wrote: > On Sun, Dec 17 2017, Giovanni Bechiswrote: > > Il 17 dicembre 2017 11:54:22 CET, Jeremie Courreges-Anglas > > ha scritto: > >>On Sat, Dec 16 2017, Giovanni Bechis wrote: > >>> On 12/15/17 17:51, Jeremie Courreges-Anglas wrote: > On Wed, Nov 22 2017, Giovanni Bechis wrote: > > Hi, > > update to latest release, some bug fixes and pledge(2) support > > committed upstream. > > I don't use this but the update looks fine ports-wise. ok jca@ fwiw > > > Pledge(2) support is enabled only if the daemon is not run with "-o > >>no_drop_privileges" parameter; > > do we want to go the way upstream goes or should we disable the > >>possibility to disable pledge(2) ? > > I would not bother disabling this, but if you do make it obvious > >>that > -o no_drop_privileges won't work. If people use the option as > a workaround and slack off instead of reporting bugs, they're the > >>ones > not benefiting from pledge(2), which is not a smart thing to do. > > >>> what about this one ? > >> > >>I have a knee-jerk reaction whenever I see #ifdef SOME_OS, I would have > >>implemented this as a configure-time option which would be usable on > >>other systems that provide sandboxing. Words are cheap and I don't > >>care > >>enough to write a diff, so please go ahead with whatever suits you. ;) > > > > As done in other diffs, I would like to go with a diff as simple as > > possible for our tree, then push a proper autoconf diff to upstream for > > next release. > > Cool! But then you could make your patch even shorter: no need to add > the #ifdef __OpenBSD__ goo. :) > correct, what about that then ? Cheers Giovanni Index: Makefile === RCS file: /var/cvs/ports/misc/memcached/Makefile,v retrieving revision 1.39 diff -u -p -r1.39 Makefile --- Makefile31 Oct 2017 07:39:25 - 1.39 +++ Makefile15 Dec 2017 23:28:28 - @@ -2,7 +2,7 @@ COMMENT= distributed memory object caching system -DISTNAME= memcached-1.5.2 +DISTNAME= memcached-1.5.3 CATEGORIES=misc HOMEPAGE= https://www.memcached.org/ Index: distinfo === RCS file: /var/cvs/ports/misc/memcached/distinfo,v retrieving revision 1.25 diff -u -p -r1.25 distinfo --- distinfo31 Oct 2017 07:39:25 - 1.25 +++ distinfo15 Dec 2017 23:28:30 - @@ -1,2 +1,2 @@ -SHA256 (memcached-1.5.2.tar.gz) = mskxE7210DfnnGEnc4ZWSsLl4x1J5ZTxHlVOTBSbckU= -SIZE (memcached-1.5.2.tar.gz) = 407715 +SHA256 (memcached-1.5.3.tar.gz) = JYzD3bdhNoVGWs/QIV+CciCju90Wf9LAgGMhBbLS884= +SIZE (memcached-1.5.3.tar.gz) = 409131 Index: patches/patch-configure === RCS file: /var/cvs/ports/misc/memcached/patches/patch-configure,v retrieving revision 1.10 diff -u -p -r1.10 patch-configure --- patches/patch-configure 31 Oct 2017 07:39:25 - 1.10 +++ patches/patch-configure 15 Dec 2017 23:30:49 - @@ -2,7 +2,7 @@ $OpenBSD: patch-configure,v 1.10 2017/10 Index: configure --- configure.orig +++ configure -@@ -6517,7 +6517,7 @@ $as_echo "#define _GNU_SOURCE 1" >>confdefs.h +@@ -6551,7 +6551,7 @@ $as_echo "#define _GNU_SOURCE 1" >>confdefs.h elif test "$GCC" = "yes" then GCC_VERSION=`$CC -dumpversion` Index: patches/patch-memcached_c === RCS file: /var/cvs/ports/misc/memcached/patches/patch-memcached_c,v retrieving revision 1.14 diff -u -p -r1.14 patch-memcached_c --- patches/patch-memcached_c 31 Oct 2017 07:39:25 - 1.14 +++ patches/patch-memcached_c 17 Dec 2017 19:31:41 - @@ -1,31 +1,15 @@ -$OpenBSD: patch-memcached_c,v 1.14 2017/10/31 07:39:25 giovanni Exp $ +$OpenBSD$ + Index: memcached.c --- memcached.c.orig +++ memcached.c -@@ -23,6 +23,7 @@ - #include - #include - #include -+#include - - /* some POSIX systems need the following definition - * to get mlockall flags out of sys/mman.h. */ -@@ -6861,6 +6862,18 @@ int main (int argc, char **argv) { - - if (pid_file != NULL) { - save_pid(pid_file); -+} -+ -+if (settings.socketpath != NULL) { -+ if (pledge("stdio unix", NULL) == -1) { -+ fprintf(stderr, "%s: pledge: %s\n", argv[0], strerror(errno)); -+ exit(1); -+ } +@@ -6927,6 +6927,9 @@ int main (int argc, char **argv) { + /* Drop privileges no longer needed */ + if (settings.drop_privileges) { + drop_privileges(); +} else { -+ if (pledge("stdio inet", NULL) == -1) { -+ fprintf(stderr, "%s: pledge: %s\n", argv[0], strerror(errno)); -+ exit(1); -+ } ++fprintf(stderr, "no_drop_privileges
Re: UPDATE: misc/memcached
On Sun, Dec 17 2017, Giovanni Bechiswrote: > Il 17 dicembre 2017 11:54:22 CET, Jeremie Courreges-Anglas > ha scritto: >>On Sat, Dec 16 2017, Giovanni Bechis wrote: >>> On 12/15/17 17:51, Jeremie Courreges-Anglas wrote: On Wed, Nov 22 2017, Giovanni Bechis wrote: > Hi, > update to latest release, some bug fixes and pledge(2) support > committed upstream. I don't use this but the update looks fine ports-wise. ok jca@ fwiw > Pledge(2) support is enabled only if the daemon is not run with "-o >>no_drop_privileges" parameter; > do we want to go the way upstream goes or should we disable the >>possibility to disable pledge(2) ? I would not bother disabling this, but if you do make it obvious >>that -o no_drop_privileges won't work. If people use the option as a workaround and slack off instead of reporting bugs, they're the >>ones not benefiting from pledge(2), which is not a smart thing to do. >>> what about this one ? >> >>I have a knee-jerk reaction whenever I see #ifdef SOME_OS, I would have >>implemented this as a configure-time option which would be usable on >>other systems that provide sandboxing. Words are cheap and I don't >>care >>enough to write a diff, so please go ahead with whatever suits you. ;) > > As done in other diffs, I would like to go with a diff as simple as possible > for our tree, then push a proper autoconf diff to upstream for next release. Cool! But then you could make your patch even shorter: no need to add the #ifdef __OpenBSD__ goo. :) -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE signature.asc Description: PGP signature
Re: UPDATE: misc/memcached
Il 17 dicembre 2017 11:54:22 CET, Jeremie Courreges-Anglasha scritto: >On Sat, Dec 16 2017, Giovanni Bechis wrote: >> On 12/15/17 17:51, Jeremie Courreges-Anglas wrote: >>> On Wed, Nov 22 2017, Giovanni Bechis wrote: Hi, update to latest release, some bug fixes and pledge(2) support committed upstream. >>> >>> I don't use this but the update looks fine ports-wise. ok jca@ fwiw >>> Pledge(2) support is enabled only if the daemon is not run with "-o >no_drop_privileges" parameter; do we want to go the way upstream goes or should we disable the >possibility to disable pledge(2) ? >>> >>> I would not bother disabling this, but if you do make it obvious >that >>> -o no_drop_privileges won't work. If people use the option as >>> a workaround and slack off instead of reporting bugs, they're the >ones >>> not benefiting from pledge(2), which is not a smart thing to do. >>> >> what about this one ? > >I have a knee-jerk reaction whenever I see #ifdef SOME_OS, I would have >implemented this as a configure-time option which would be usable on >other systems that provide sandboxing. Words are cheap and I don't >care >enough to write a diff, so please go ahead with whatever suits you. ;) As done in other diffs, I would like to go with a diff as simple as possible for our tree, then push a proper autoconf diff to upstream for next release. Giovanni
Re: UPDATE: misc/memcached
On Sat, Dec 16 2017, Giovanni Bechiswrote: > On 12/15/17 17:51, Jeremie Courreges-Anglas wrote: >> On Wed, Nov 22 2017, Giovanni Bechis wrote: >>> Hi, >>> update to latest release, some bug fixes and pledge(2) support >>> committed upstream. >> >> I don't use this but the update looks fine ports-wise. ok jca@ fwiw >> >>> Pledge(2) support is enabled only if the daemon is not run with "-o >>> no_drop_privileges" parameter; >>> do we want to go the way upstream goes or should we disable the possibility >>> to disable pledge(2) ? >> >> I would not bother disabling this, but if you do make it obvious that >> -o no_drop_privileges won't work. If people use the option as >> a workaround and slack off instead of reporting bugs, they're the ones >> not benefiting from pledge(2), which is not a smart thing to do. >> > what about this one ? I have a knee-jerk reaction whenever I see #ifdef SOME_OS, I would have implemented this as a configure-time option which would be usable on other systems that provide sandboxing. Words are cheap and I don't care enough to write a diff, so please go ahead with whatever suits you. ;) -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
Re: UPDATE: misc/memcached
On 12/15/17 17:51, Jeremie Courreges-Anglas wrote: > On Wed, Nov 22 2017, Giovanni Bechiswrote: >> Hi, >> update to latest release, some bug fixes and pledge(2) support >> committed upstream. > > I don't use this but the update looks fine ports-wise. ok jca@ fwiw > >> Pledge(2) support is enabled only if the daemon is not run with "-o >> no_drop_privileges" parameter; >> do we want to go the way upstream goes or should we disable the possibility >> to disable pledge(2) ? > > I would not bother disabling this, but if you do make it obvious that > -o no_drop_privileges won't work. If people use the option as > a workaround and slack off instead of reporting bugs, they're the ones > not benefiting from pledge(2), which is not a smart thing to do. > what about this one ? Cheers & Thanks Giovanni > My two cents, > Index: Makefile === RCS file: /var/cvs/ports/misc/memcached/Makefile,v retrieving revision 1.39 diff -u -p -r1.39 Makefile --- Makefile 31 Oct 2017 07:39:25 - 1.39 +++ Makefile 15 Dec 2017 23:28:28 - @@ -2,7 +2,7 @@ COMMENT= distributed memory object caching system -DISTNAME= memcached-1.5.2 +DISTNAME= memcached-1.5.3 CATEGORIES= misc HOMEPAGE= https://www.memcached.org/ Index: distinfo === RCS file: /var/cvs/ports/misc/memcached/distinfo,v retrieving revision 1.25 diff -u -p -r1.25 distinfo --- distinfo 31 Oct 2017 07:39:25 - 1.25 +++ distinfo 15 Dec 2017 23:28:30 - @@ -1,2 +1,2 @@ -SHA256 (memcached-1.5.2.tar.gz) = mskxE7210DfnnGEnc4ZWSsLl4x1J5ZTxHlVOTBSbckU= -SIZE (memcached-1.5.2.tar.gz) = 407715 +SHA256 (memcached-1.5.3.tar.gz) = JYzD3bdhNoVGWs/QIV+CciCju90Wf9LAgGMhBbLS884= +SIZE (memcached-1.5.3.tar.gz) = 409131 Index: patches/patch-configure === RCS file: /var/cvs/ports/misc/memcached/patches/patch-configure,v retrieving revision 1.10 diff -u -p -r1.10 patch-configure --- patches/patch-configure 31 Oct 2017 07:39:25 - 1.10 +++ patches/patch-configure 15 Dec 2017 23:30:49 - @@ -2,7 +2,7 @@ $OpenBSD: patch-configure,v 1.10 2017/10 Index: configure --- configure.orig +++ configure -@@ -6517,7 +6517,7 @@ $as_echo "#define _GNU_SOURCE 1" >>confdefs.h +@@ -6551,7 +6551,7 @@ $as_echo "#define _GNU_SOURCE 1" >>confdefs.h elif test "$GCC" = "yes" then GCC_VERSION=`$CC -dumpversion` Index: patches/patch-memcached_c === RCS file: /var/cvs/ports/misc/memcached/patches/patch-memcached_c,v retrieving revision 1.14 diff -u -p -r1.14 patch-memcached_c --- patches/patch-memcached_c 31 Oct 2017 07:39:25 - 1.14 +++ patches/patch-memcached_c 15 Dec 2017 23:30:43 - @@ -1,31 +1,17 @@ -$OpenBSD: patch-memcached_c,v 1.14 2017/10/31 07:39:25 giovanni Exp $ +$OpenBSD$ + Index: memcached.c --- memcached.c.orig +++ memcached.c -@@ -23,6 +23,7 @@ - #include - #include - #include -+#include - - /* some POSIX systems need the following definition - * to get mlockall flags out of sys/mman.h. */ -@@ -6861,6 +6862,18 @@ int main (int argc, char **argv) { - - if (pid_file != NULL) { - save_pid(pid_file); -+} -+ -+if (settings.socketpath != NULL) { -+ if (pledge("stdio unix", NULL) == -1) { -+ fprintf(stderr, "%s: pledge: %s\n", argv[0], strerror(errno)); -+ exit(1); -+ } +@@ -6927,6 +6927,11 @@ int main (int argc, char **argv) { + /* Drop privileges no longer needed */ + if (settings.drop_privileges) { + drop_privileges(); +} else { -+ if (pledge("stdio inet", NULL) == -1) { -+ fprintf(stderr, "%s: pledge: %s\n", argv[0], strerror(errno)); -+ exit(1); -+ } ++#ifdef __OpenBSD__ ++fprintf(stderr, "no_drop_privileges option unsupported on this operating system.\n"); ++exit(EXIT_FAILURE); ++#endif } - /* Drop privileges no longer needed */ + /* Initialize the uriencode lookup table. */
Re: UPDATE: misc/memcached
On Wed, Nov 22 2017, Giovanni Bechiswrote: > Hi, > update to latest release, some bug fixes and pledge(2) support > committed upstream. I don't use this but the update looks fine ports-wise. ok jca@ fwiw > Pledge(2) support is enabled only if the daemon is not run with "-o > no_drop_privileges" parameter; > do we want to go the way upstream goes or should we disable the possibility > to disable pledge(2) ? I would not bother disabling this, but if you do make it obvious that -o no_drop_privileges won't work. If people use the option as a workaround and slack off instead of reporting bugs, they're the ones not benefiting from pledge(2), which is not a smart thing to do. My two cents, -- jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
Re: UPDATE misc/memcached
On Thu, Dec 22, 2011 at 07:06:14PM +0100, Sebastian Reitenbach wrote: Maintainer timed out. I'd like to get this in. This time its inlined, the only change is that I moved the -d parameter to daemon, instead of the daemon_flags, similar to other changes done to other ports so that it always daemonizes. is it OK? OK for me. Slightly tested with devel/subversion on amd64. It seems to run fine: stats STAT pid 12951 STAT uptime 1314 STAT time 1324808489 STAT version 1.4.10 STAT libevent 1.4.14b-stable STAT pointer_size 64 STAT rusage_user 0.11 STAT rusage_system 0.36 STAT curr_connections 10 STAT total_connections 750 STAT connection_structures 13 STAT reserved_fds 20 STAT cmd_get 8112 STAT cmd_set 5508 STAT cmd_flush 0 STAT cmd_touch 0 STAT get_hits 2604 STAT get_misses 5508 STAT delete_misses 0 STAT delete_hits 0 STAT incr_misses 0 STAT incr_hits 0 STAT decr_misses 0 STAT decr_hits 0 STAT cas_misses 0 STAT cas_hits 0 STAT cas_badval 0 STAT touch_hits 0 STAT touch_misses 0 STAT auth_cmds 0 STAT auth_errors 0 STAT bytes_read 1485917 STAT bytes_written 497951 STAT limit_maxbytes 67108864 STAT accepting_conns 1 STAT listen_disabled_num 0 STAT threads 4 STAT conn_yields 0 STAT hash_power_level 16 STAT hash_bytes 524288 STAT hash_is_expanding 0 STAT expired_unfetched 0 STAT evicted_unfetched 0 STAT bytes 861008 STAT curr_items 4409 STAT total_items 5508 STAT evictions 0 STAT reclaimed 0 ? memcached-1.4.10.diff Index: Makefile === RCS file: /cvs/ports/misc/memcached/Makefile,v retrieving revision 1.15 diff -u -r1.15 Makefile --- Makefile 11 Jul 2011 12:13:03 - 1.15 +++ Makefile 14 Nov 2011 09:33:08 - @@ -2,8 +2,7 @@ COMMENT= distributed memory object caching system -DISTNAME=memcached-1.4.5 -REVISION=0 +DISTNAME = memcached-1.4.10 CATEGORIES= misc HOMEPAGE=http://www.danga.com/memcached/ Index: distinfo === RCS file: /cvs/ports/misc/memcached/distinfo,v retrieving revision 1.10 diff -u -r1.10 distinfo --- distinfo 14 Oct 2010 12:33:03 - 1.10 +++ distinfo 14 Nov 2011 09:33:08 - @@ -1,5 +1,5 @@ -MD5 (memcached-1.4.5.tar.gz) = WDRBol+Tc2BiQCTyiB5eqA== -RMD160 (memcached-1.4.5.tar.gz) = wfU0+8WFoeoDO70S1vyUx3oIUDI= -SHA1 (memcached-1.4.5.tar.gz) = x9ZRd2S4LSOuLedrVsJJQ0PFPwI= -SHA256 (memcached-1.4.5.tar.gz) = lXG0uFSE5Gs7EPB8y6d6H6l9YGYLMoWfmQ7/77MAX5E= -SIZE (memcached-1.4.5.tar.gz) = 302516 +MD5 (memcached-1.4.10.tar.gz) = jhgFTsXt/Zb33ofwJiIFKg== +RMD160 (memcached-1.4.10.tar.gz) = Z/Rne6ZEsbFRUaaGWoN3lm/hlwY= +SHA1 (memcached-1.4.10.tar.gz) = CuMA+Fi3Z6v4EgCdU+1YZHzrSYo= +SHA256 (memcached-1.4.10.tar.gz) = LMPeNBtWFGTaKfxdXipPrMD2qvwVHw4dlCKHBeB1j98= +SIZE (memcached-1.4.10.tar.gz) = 311926 Index: pkg/MESSAGE === RCS file: /cvs/ports/misc/memcached/pkg/MESSAGE,v retrieving revision 1.1 diff -u -r1.1 MESSAGE --- pkg/MESSAGE 7 Aug 2009 10:46:41 - 1.1 +++ pkg/MESSAGE 14 Nov 2011 09:33:08 - @@ -1,13 +1,3 @@ -To start memcached at boot time, add the following to /etc/rc.local: - -if [ -x ${PREFIX}/bin/memcached ]; then -echo -n ' memcached' -${PREFIX}/bin/memcached -d \ - -u _memcached -P /var/run/memcached.pid -fi - -To see further configuration options, run memcached -h. - Since memcached has no authentication mechanisms, it is designed for internal use only, and should be firewalled accordingly. You may wish to add a rule such as the following to /etc/pf.conf: Index: pkg/PLIST === RCS file: /cvs/ports/misc/memcached/pkg/PLIST,v retrieving revision 1.5 diff -u -r1.5 PLIST --- pkg/PLIST 21 Apr 2010 12:23:37 - 1.5 +++ pkg/PLIST 14 Nov 2011 09:33:08 - @@ -8,4 +8,4 @@ share/doc/memcached/ share/doc/memcached/protocol.txt share/doc/memcached/readme.txt -share/doc/memcached/threads.txt +@rcscript ${RCDIR}/memcached Index: pkg/memcached.rc === RCS file: pkg/memcached.rc diff -N pkg/memcached.rc --- /dev/null 1 Jan 1970 00:00:00 - +++ pkg/memcached.rc 14 Nov 2011 09:33:08 - @@ -0,0 +1,12 @@ +#!/bin/sh +# +# $OpenBSD$ + +daemon=${TRUEPREFIX}/bin/memcached -d +daemon_flags=-u _memcached -P /var/run/memcached.pid + +. /etc/rc.d/rc.subr + +rc_reload=NO + +rc_cmd $1
Re: UPDATE misc/memcached
On Thursday, December 15, 2011 11:54 CET, Sebastian Reitenbach sebas...@l00-bugdead-prods.de wrote: On Tuesday, November 22, 2011 00:09 CET, Sebastian Reitenbach sebas...@l00-bugdead-prods.de wrote: On Monday, November 14, 2011 11:14 CET, Sebastian Reitenbach sebas...@l00-bugdead-prods.de wrote: Hi, updated memcached to 1.4.10, and added a rc script, which was long overdue ;) Tested with www/sogo, and updated libmemcached which will come in next mail. It seems to still work well with the new version. is it OK? Sebastian Maintainer timed out. I'd like to get this in. This time its inlined, the only change is that I moved the -d parameter to daemon, instead of the daemon_flags, similar to other changes done to other ports so that it always daemonizes. is it OK? ? memcached-1.4.10.diff Index: Makefile === RCS file: /cvs/ports/misc/memcached/Makefile,v retrieving revision 1.15 diff -u -r1.15 Makefile --- Makefile11 Jul 2011 12:13:03 - 1.15 +++ Makefile14 Nov 2011 09:33:08 - @@ -2,8 +2,7 @@ COMMENT= distributed memory object caching system -DISTNAME= memcached-1.4.5 -REVISION= 0 +DISTNAME = memcached-1.4.10 CATEGORIES=misc HOMEPAGE= http://www.danga.com/memcached/ Index: distinfo === RCS file: /cvs/ports/misc/memcached/distinfo,v retrieving revision 1.10 diff -u -r1.10 distinfo --- distinfo14 Oct 2010 12:33:03 - 1.10 +++ distinfo14 Nov 2011 09:33:08 - @@ -1,5 +1,5 @@ -MD5 (memcached-1.4.5.tar.gz) = WDRBol+Tc2BiQCTyiB5eqA== -RMD160 (memcached-1.4.5.tar.gz) = wfU0+8WFoeoDO70S1vyUx3oIUDI= -SHA1 (memcached-1.4.5.tar.gz) = x9ZRd2S4LSOuLedrVsJJQ0PFPwI= -SHA256 (memcached-1.4.5.tar.gz) = lXG0uFSE5Gs7EPB8y6d6H6l9YGYLMoWfmQ7/77MAX5E= -SIZE (memcached-1.4.5.tar.gz) = 302516 +MD5 (memcached-1.4.10.tar.gz) = jhgFTsXt/Zb33ofwJiIFKg== +RMD160 (memcached-1.4.10.tar.gz) = Z/Rne6ZEsbFRUaaGWoN3lm/hlwY= +SHA1 (memcached-1.4.10.tar.gz) = CuMA+Fi3Z6v4EgCdU+1YZHzrSYo= +SHA256 (memcached-1.4.10.tar.gz) = LMPeNBtWFGTaKfxdXipPrMD2qvwVHw4dlCKHBeB1j98= +SIZE (memcached-1.4.10.tar.gz) = 311926 Index: pkg/MESSAGE === RCS file: /cvs/ports/misc/memcached/pkg/MESSAGE,v retrieving revision 1.1 diff -u -r1.1 MESSAGE --- pkg/MESSAGE 7 Aug 2009 10:46:41 - 1.1 +++ pkg/MESSAGE 14 Nov 2011 09:33:08 - @@ -1,13 +1,3 @@ -To start memcached at boot time, add the following to /etc/rc.local: - -if [ -x ${PREFIX}/bin/memcached ]; then -echo -n ' memcached' -${PREFIX}/bin/memcached -d \ - -u _memcached -P /var/run/memcached.pid -fi - -To see further configuration options, run memcached -h. - Since memcached has no authentication mechanisms, it is designed for internal use only, and should be firewalled accordingly. You may wish to add a rule such as the following to /etc/pf.conf: Index: pkg/PLIST === RCS file: /cvs/ports/misc/memcached/pkg/PLIST,v retrieving revision 1.5 diff -u -r1.5 PLIST --- pkg/PLIST 21 Apr 2010 12:23:37 - 1.5 +++ pkg/PLIST 14 Nov 2011 09:33:08 - @@ -8,4 +8,4 @@ share/doc/memcached/ share/doc/memcached/protocol.txt share/doc/memcached/readme.txt -share/doc/memcached/threads.txt +@rcscript ${RCDIR}/memcached Index: pkg/memcached.rc === RCS file: pkg/memcached.rc diff -N pkg/memcached.rc --- /dev/null 1 Jan 1970 00:00:00 - +++ pkg/memcached.rc14 Nov 2011 09:33:08 - @@ -0,0 +1,12 @@ +#!/bin/sh +# +# $OpenBSD$ + +daemon=${TRUEPREFIX}/bin/memcached -d +daemon_flags=-u _memcached -P /var/run/memcached.pid + +. /etc/rc.d/rc.subr + +rc_reload=NO + +rc_cmd $1
Re: UPDATE misc/memcached
On Tuesday, November 22, 2011 00:09 CET, Sebastian Reitenbach sebas...@l00-bugdead-prods.de wrote: On Monday, November 14, 2011 11:14 CET, Sebastian Reitenbach sebas...@l00-bugdead-prods.de wrote: Hi, updated memcached to 1.4.10, and added a rc script, which was long overdue ;) Tested with www/sogo, and updated libmemcached which will come in next mail. It seems to still work well with the new version. is it OK? Sebastian ping! pong. reattached to ease review. Sebastian ? memcached-1.4.10.diff Index: Makefile === RCS file: /cvs/ports/misc/memcached/Makefile,v retrieving revision 1.15 diff -u -r1.15 Makefile --- Makefile 11 Jul 2011 12:13:03 - 1.15 +++ Makefile 14 Nov 2011 09:33:08 - @@ -2,8 +2,7 @@ COMMENT= distributed memory object caching system -DISTNAME= memcached-1.4.5 -REVISION= 0 +DISTNAME = memcached-1.4.10 CATEGORIES= misc HOMEPAGE= http://www.danga.com/memcached/ Index: distinfo === RCS file: /cvs/ports/misc/memcached/distinfo,v retrieving revision 1.10 diff -u -r1.10 distinfo --- distinfo 14 Oct 2010 12:33:03 - 1.10 +++ distinfo 14 Nov 2011 09:33:08 - @@ -1,5 +1,5 @@ -MD5 (memcached-1.4.5.tar.gz) = WDRBol+Tc2BiQCTyiB5eqA== -RMD160 (memcached-1.4.5.tar.gz) = wfU0+8WFoeoDO70S1vyUx3oIUDI= -SHA1 (memcached-1.4.5.tar.gz) = x9ZRd2S4LSOuLedrVsJJQ0PFPwI= -SHA256 (memcached-1.4.5.tar.gz) = lXG0uFSE5Gs7EPB8y6d6H6l9YGYLMoWfmQ7/77MAX5E= -SIZE (memcached-1.4.5.tar.gz) = 302516 +MD5 (memcached-1.4.10.tar.gz) = jhgFTsXt/Zb33ofwJiIFKg== +RMD160 (memcached-1.4.10.tar.gz) = Z/Rne6ZEsbFRUaaGWoN3lm/hlwY= +SHA1 (memcached-1.4.10.tar.gz) = CuMA+Fi3Z6v4EgCdU+1YZHzrSYo= +SHA256 (memcached-1.4.10.tar.gz) = LMPeNBtWFGTaKfxdXipPrMD2qvwVHw4dlCKHBeB1j98= +SIZE (memcached-1.4.10.tar.gz) = 311926 Index: pkg/MESSAGE === RCS file: /cvs/ports/misc/memcached/pkg/MESSAGE,v retrieving revision 1.1 diff -u -r1.1 MESSAGE --- pkg/MESSAGE 7 Aug 2009 10:46:41 - 1.1 +++ pkg/MESSAGE 14 Nov 2011 09:33:08 - @@ -1,13 +1,3 @@ -To start memcached at boot time, add the following to /etc/rc.local: - -if [ -x ${PREFIX}/bin/memcached ]; then -echo -n ' memcached' -${PREFIX}/bin/memcached -d \ - -u _memcached -P /var/run/memcached.pid -fi - -To see further configuration options, run memcached -h. - Since memcached has no authentication mechanisms, it is designed for internal use only, and should be firewalled accordingly. You may wish to add a rule such as the following to /etc/pf.conf: Index: pkg/PLIST === RCS file: /cvs/ports/misc/memcached/pkg/PLIST,v retrieving revision 1.5 diff -u -r1.5 PLIST --- pkg/PLIST 21 Apr 2010 12:23:37 - 1.5 +++ pkg/PLIST 14 Nov 2011 09:33:08 - @@ -8,4 +8,4 @@ share/doc/memcached/ share/doc/memcached/protocol.txt share/doc/memcached/readme.txt -share/doc/memcached/threads.txt +@rcscript ${RCDIR}/memcached Index: pkg/memcached.rc === RCS file: pkg/memcached.rc diff -N pkg/memcached.rc --- /dev/null 1 Jan 1970 00:00:00 - +++ pkg/memcached.rc 14 Nov 2011 09:33:08 - @@ -0,0 +1,12 @@ +#!/bin/sh +# +# $OpenBSD$ + +daemon=${TRUEPREFIX}/bin/memcached +daemon_flags=-d -u _memcached -P /var/run/memcached.pid + +. /etc/rc.d/rc.subr + +rc_reload=NO + +rc_cmd $1
Re: UPDATE misc/memcached
On Monday, November 14, 2011 11:14 CET, Sebastian Reitenbach sebas...@l00-bugdead-prods.de wrote: Hi, updated memcached to 1.4.10, and added a rc script, which was long overdue ;) Tested with www/sogo, and updated libmemcached which will come in next mail. It seems to still work well with the new version. is it OK? Sebastian ping! reattached to ease review. Sebastian memcached-1.4.10.diff Description: Binary data
Re: UPDATE: misc/memcached
Scott Vokes wrote: This updates memcached from 1.2.6 to 1.2.8 (they strongly encourage updating). Please note that the patch-memcached_c file is no longer necessary, and should be deleted. I don't use memcahced on anything -current, so I cannot test it, but FYI, something along $ rm patches/patch-memcached_c $ cvs delete patches/patch-memcached_c $ cvs -q diff -uNp should make the deletion included in your diff. /Alexander
Re: [+] UPDATE: misc/memcached
Le Thu, Nov 06, 2008 at 11:03:12AM +0300, [EMAIL PROTECTED] ecrivait : Trivial diff to update memcached to it's latest stable version. Works very well. It looks the way it manages its stored objects has changed a bit, and with the same high water limit, it takes longer before it starts pruning entries. Thanks for the update. -- Frank Denis - j [at] pureftpd.org - http://00f.net - http://www.cotery.com
Re: UPDATE: misc/memcached
On 2008/11/06 11:03, [EMAIL PROTECTED] wrote: Trivial diff to update memcached to it's latest stable version. /usr/ports/packages/amd64/all/memcached-1.2.6.tgz: Extra: pthread.11
Re: UPDATE: misc/memcached
On Thu, 6 Nov 2008 12:27:02 + Stuart Henderson [EMAIL PROTECTED] wrote: /usr/ports/packages/amd64/all/memcached-1.2.6.tgz: Extra: pthread.11 Thanks. Here is updated diff. memcached-1.2.6.diff Description: Binary data