Re: svn commit: r301777 - head/lib/libc/stdio
Please back it out. __sccl() use plain chars, you can't just call wide
chars __wcollate_range_cmp() from it. Never intermix plain chars and
wide chars without conversion. The problem is somewhere else. The code
used for [a-z] ranges in the format. Moreover, even if conversion added
this code can't work with wchars by definition since uses the same loop
as in regcomp()
for (i = 0; i < 256; i ++)
Counting completely broken regcomp, our locale quickly becomes
intermixing junk of chars and wchars without any conversion. Please
anybody stop hacking here, especially blindly adding wchars everywhere
without analyzing rest of code.
If you see 256 or UCHAR_MAX somewhere, this function can't be converted
to wchars.
On 10.06.2016 8:21, Pedro F. Giffuni wrote:
> Author: pfg
> Date: Fri Jun 10 05:21:52 2016
> New Revision: 301777
> URL: https://svnweb.freebsd.org/changeset/base/301777
>
> Log:
> Fix regression from r301461.
>
> The fix to the __collate_range_cmp() ABI breakage missed some replacements
> in libc's vfscanf(). Replace them with __wcollate_range_cmp() which
> does what is expected.
>
> This was breaking applications like xterm and pidgin when using wide
> characters.
>
> Reported by:Vitalij Satanivskij
> Approved by:re
>
> Modified:
> head/lib/libc/stdio/vfscanf.c
>
> Modified: head/lib/libc/stdio/vfscanf.c
> ==
> --- head/lib/libc/stdio/vfscanf.c Fri Jun 10 04:04:55 2016
> (r301776)
> +++ head/lib/libc/stdio/vfscanf.c Fri Jun 10 05:21:52 2016
> (r301777)
> @@ -873,7 +873,7 @@ doswitch:
> n = *fmt;
> if (n == ']'
> || (table->__collate_load_error ? n < c :
> - __collate_range_cmp (table, n, c) < 0
> + __wcollate_range_cmp(table, n, c) < 0
> )
> ) {
> c = '-';
> @@ -887,8 +887,8 @@ doswitch:
> } while (c < n);
> } else {
> for (i = 0; i < 256; i ++)
> - if ( __collate_range_cmp (table, c,
> i) < 0
> - && __collate_range_cmp (table, i,
> n) <= 0
> + if (__wcollate_range_cmp(table, c, i) <
> 0 &&
> + __wcollate_range_cmp(table, i, n)
> <= 0
> )
> tab[i] = v;
> }
>
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r301779 - head/share/misc
Author: tz (ports committer) Date: Fri Jun 10 08:33:15 2016 New Revision: 301779 URL: https://svnweb.freebsd.org/changeset/base/301779 Log: - Add myself (tz / Torsten Zuehlsdorff) to ports commiters - Add mentors junovitch, pi, swills Approved by: re (gjb), junovitch (mentor) Modified: head/share/misc/committers-ports.dot Modified: head/share/misc/committers-ports.dot == --- head/share/misc/committers-ports.dotFri Jun 10 06:04:53 2016 (r301778) +++ head/share/misc/committers-ports.dotFri Jun 10 08:33:15 2016 (r301779) @@ -226,6 +226,7 @@ tota [label="TAKATSU Tomonari\ntota@Free trasz [label="Edward Tomasz Napierala\ntr...@freebsd.org\n2007/04/12"] trhodes [label="Tom Rhodes\ntrho...@freebsd.org\n2004/07/06"] trociny [label="Mikolaj Golub\ntroc...@freebsd.org\n2013/10/17"] +tz [label="Torsten Zuehlsdorff\n...@freebsd.org\n2016/06/04"] uqs [label="Ulrich Spoerlein\n...@freebsd.org\n2012/01/19"] vd [label="Vasil Dimov\n...@freebsd.org\n2006/01/19"] vg [label="Veniamin Gvozdikov\n...@freebsd.org\n2013/06/11"] @@ -413,6 +414,8 @@ jadawin -> wen joerg -> netchild +junovitch -> tz + knu -> daichi knu -> maho knu -> nobutaka @@ -530,6 +533,8 @@ pgollucci -> swills philip -> koitsu +pi -> tz + rafan -> chinsan rakuco -> alonso @@ -573,6 +578,7 @@ swills -> milki swills -> pclin swills -> robak swills -> rpaulo +swills -> tz swills -> xmj tabthorpe -> ashish ___ svn-src-head@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r301777 - head/lib/libc/stdio
(Sorry I have been offline/sleep for a while)
Hello;
I cannot back this out because of this report:
http://docs.freebsd.org/cgi/mid.cgi?20160608132224.GA15454
This change only re-states what has been working before the ABI breakage
fix in r301461.
The fix to regex is completely independent.
Pedro.
On 06/10/16 03:20, Andrey Chernov wrote:
Please back it out. __sccl() use plain chars, you can't just call wide
chars __wcollate_range_cmp() from it. Never intermix plain chars and
wide chars without conversion. The problem is somewhere else. The code
used for [a-z] ranges in the format. Moreover, even if conversion added
this code can't work with wchars by definition since uses the same loop
as in regcomp()
for (i = 0; i < 256; i ++)
Counting completely broken regcomp, our locale quickly becomes
intermixing junk of chars and wchars without any conversion. Please
anybody stop hacking here, especially blindly adding wchars everywhere
without analyzing rest of code.
If you see 256 or UCHAR_MAX somewhere, this function can't be converted
to wchars.
On 10.06.2016 8:21, Pedro F. Giffuni wrote:
Author: pfg
Date: Fri Jun 10 05:21:52 2016
New Revision: 301777
URL: https://svnweb.freebsd.org/changeset/base/301777
Log:
Fix regression from r301461.
The fix to the __collate_range_cmp() ABI breakage missed some replacements
in libc's vfscanf(). Replace them with __wcollate_range_cmp() which
does what is expected.
This was breaking applications like xterm and pidgin when using wide
characters.
Reported by: Vitalij Satanivskij
Approved by: re
Modified:
head/lib/libc/stdio/vfscanf.c
Modified: head/lib/libc/stdio/vfscanf.c
==
--- head/lib/libc/stdio/vfscanf.c Fri Jun 10 04:04:55 2016
(r301776)
+++ head/lib/libc/stdio/vfscanf.c Fri Jun 10 05:21:52 2016
(r301777)
@@ -873,7 +873,7 @@ doswitch:
n = *fmt;
if (n == ']'
|| (table->__collate_load_error ? n < c :
- __collate_range_cmp (table, n, c) < 0
+ __wcollate_range_cmp(table, n, c) < 0
)
) {
c = '-';
@@ -887,8 +887,8 @@ doswitch:
} while (c < n);
} else {
for (i = 0; i < 256; i ++)
- if ( __collate_range_cmp (table, c, i)
< 0
- && __collate_range_cmp (table, i, n)
<= 0
+ if (__wcollate_range_cmp(table, c, i) < 0
&&
+ __wcollate_range_cmp(table, i, n)
<= 0
)
tab[i] = v;
}
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r301777 - head/lib/libc/stdio
On 10.06.2016 15:03, Pedro Giffuni wrote: > (Sorry I have been offline/sleep for a while) > > Hello; > > I cannot back this out because of this report: > > http://docs.freebsd.org/cgi/mid.cgi?20160608132224.GA15454 > > This change only re-states what has been working before the ABI breakage > fix in r301461. Because now strcoll_l() is used as before and it is broken in wide char converting process. It is broken for _all_, not just for vfscanf() ranges. It should never fail with any args. Real fix should be in strcoll_l(). > The fix to regex is completely independent. It is dependent. vfscanf() and regcomp() use the same code for range collation. If we remove range collation from everywhere, we save vfscanf(), but strcoll_l() remains broken. ___ svn-src-head@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r301777 - head/lib/libc/stdio
On 06/10/16 07:19, Andrey Chernov wrote: On 10.06.2016 15:03, Pedro Giffuni wrote: (Sorry I have been offline/sleep for a while) Hello; I cannot back this out because of this report: http://docs.freebsd.org/cgi/mid.cgi?20160608132224.GA15454 This change only re-states what has been working before the ABI breakage fix in r301461. Because now strcoll_l() is used as before and it is broken in wide char converting process. It is broken for _all_, not just for vfscanf() ranges. It should never fail with any args. Real fix should be in strcoll_l(). Yes, it is broken as before however the objective of r301461 was only to address the ABI breakage. If I revert r301777 then I have to backout r301461 too, and then we have again the ABI breakage. I understand you want to fix all but we are under code freeze and I have to go back to a known working (although still broken) state. The fix to regex is completely independent. It is dependent. vfscanf() and regcomp() use the same code for range collation. If we remove range collation from everywhere, we save vfscanf(), but strcoll_l() remains broken. We have had a broken regex for a very long time, and we new that before collation was introduced. I didn't take that decision but moving to libtre was a lot more work and we really had to have collation now. Pedro. ___ svn-src-head@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r301787 - head/usr.sbin/bsdinstall/scripts
Author: allanjude Date: Fri Jun 10 14:31:59 2016 New Revision: 301787 URL: https://svnweb.freebsd.org/changeset/base/301787 Log: Fix bsdinstall for root-on-zfs with MBR partitioning Fix an error where vfs.root.mountfrom was not always set as required when creating a bootpool. After the recent geliboot changes, it was only set if the main pool was encrypted. Also resolve an error where the bootpool was unmounted twice causing bsdinstall to stop with an error message about the failed command. Approved by: re (gjb) Sponsored by: BSDCan Hacker Lounge Modified: head/usr.sbin/bsdinstall/scripts/zfsboot Modified: head/usr.sbin/bsdinstall/scripts/zfsboot == --- head/usr.sbin/bsdinstall/scripts/zfsbootFri Jun 10 14:31:03 2016 (r301786) +++ head/usr.sbin/bsdinstall/scripts/zfsbootFri Jun 10 14:31:59 2016 (r301787) @@ -1194,11 +1194,6 @@ zfs_create_boot() f_eval_catch $funcname chmod "$CHMOD_MODE" \ go-wrx "$bootpool/$zroot_key" || return $FAILURE - else - # Clean up - f_eval_catch $funcname zfs "$ZFS_UNMOUNT" \ -"$bootpool_name" || return $FAILURE - f_eval_catch -d $funcname umount "$UMOUNT" /mnt # tmpfs fi fi @@ -1409,6 +1404,11 @@ zfs_create_boot() "cachefile=\"$BSDINSTALL_CHROOT/boot/zfs/zpool.cache\"" \ "$zroot_name" || return $FAILURE + if [ "$ZFSBOOT_BOOT_POOL" ]; then + f_eval_catch $funcname printf "$PRINTF_CONF" \ + vfs.root.mountfrom "\"zfs:$zroot_name/$zroot_bootfs\"" \ + $BSDINSTALL_TMPBOOT/loader.conf.root || return $FAILURE + fi # # Set canmount=noauto so that the default Boot Environment (BE) does not # get mounted if a different BE is selected from the beastie menu @@ -1486,10 +1486,6 @@ zfs_create_boot() $BSDINSTALL_TMPBOOT/loader.conf.zfs || return $FAILURE done - f_eval_catch $funcname printf "$PRINTF_CONF" vfs.root.mountfrom \ - "\"zfs:$zroot_name/$zroot_bootfs\"" \ - $BSDINSTALL_TMPBOOT/loader.conf.root || return $FAILURE - return $SUCCESS } ___ svn-src-head@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r301572 - in head/lib/libcasper: libcasper services/cap_dns services/cap_grp services/cap_pwd services/cap_random services/cap_sysctl
On Wed, Jun 08, 2016 at 02:03:53AM +, Mariusz Zaborski wrote:
> Author: oshogbo
> Date: Wed Jun 8 02:03:53 2016
> New Revision: 301572
> URL: https://svnweb.freebsd.org/changeset/base/301572
> Log:
> Add flags to the Casper services.
> CASPER_SERVICE_STDIO - Casper will not close the first three descriptors
> (stdin,
> stdout and stderr) this can be helpful for debugging.
> CASPER_SERVICE_FD - Capser will not close all other descriptors, this can
> be useful for a filesystem service.
This reminds me that there are some common cases where it is wrong to
close descriptors you don't know about. The non-POSIX command
diff <(cmd1) <(cmd2)
that compares the outputs of the two commands, when executed with bash
that was compiled with the full /dev/fd visible (as in poudriere), will
actually run something like
diff /dev/fd/63 /dev/fd/62
passing two file descriptors to pipes.
When created by a shell, these pathnames will start with /dev/fd/, but
people could create symlinks to these special files.
> [snip]
> +static void
> +stdnull(void)
> +{
> + int fd;
> +
> + fd = open(_PATH_DEVNULL, O_RDWR);
> + if (fd == -1)
> + errx(1, "Unable to open %s", _PATH_DEVNULL);
> +
> + if (setsid() == -1)
> + errx(1, "Unable to detach from session");
There is an implicit assumption here that stdnull() is only called from
a process that was forked off from here, since setsid() will not and
cannot work when called from a process that is already a session leader.
If the application is running from a shell, this setsid() will exclude
the process from most signals, including terminal ^C/^\/^Z, kill % and
hangups. More generally, this might make it more likely for the process
to hang around indefinitely after the parent is gone.
> +
> + if (dup2(fd, STDIN_FILENO) == -1)
> + errx(1, "Unable to cover stdin");
> + if (dup2(fd, STDOUT_FILENO) == -1)
> + errx(1, "Unable to cover stdout");
> + if (dup2(fd, STDERR_FILENO) == -1)
> + errx(1, "Unable to cover stderr");
> +
> + close(fd);
This was not broken by this commit, but fd should not be closed if it is
equal to STDIN_FILENO, STDOUT_FILENO or STDERR_FILENO.
> [snip]
--
Jilles Tjoelker
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r301794 - head/usr.sbin/services_mkdb
Author: allanjude Date: Fri Jun 10 14:59:11 2016 New Revision: 301794 URL: https://svnweb.freebsd.org/changeset/base/301794 Log: Fix a miss merge in the services_mkdb(8) man page Restore the cross reference to getservent(3) to the correct line Approved by: re (gjb) Sponsored by: BSDCan Hacker Lounge Modified: head/usr.sbin/services_mkdb/services_mkdb.8 Modified: head/usr.sbin/services_mkdb/services_mkdb.8 == --- head/usr.sbin/services_mkdb/services_mkdb.8 Fri Jun 10 14:51:11 2016 (r301793) +++ head/usr.sbin/services_mkdb/services_mkdb.8 Fri Jun 10 14:59:11 2016 (r301794) @@ -75,6 +75,7 @@ Print the services file to stdout, omitt .El .Pp The databases are used by the C library services routines (see +.Xr getservent 3 ) . .Pp The .Fl b @@ -82,7 +83,6 @@ and .Fl l flags are mutually exclusive. The default byte ordering is the current host order. -.Xr getservent 3 ) . .Sh FILES .Bl -tag -width ".Pa /var/db/services.db.tmp" -compact .It Pa /var/db/services.db ___ svn-src-head@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r301777 - head/lib/libc/stdio
On 10.06.2016 17:14, Pedro Giffuni wrote:
>> Because now strcoll_l() is used as before and it is broken in wide char
>> converting process. It is broken for _all_, not just for vfscanf()
>> ranges. It should never fail with any args.
>> Real fix should be in strcoll_l().
>>
>
> Yes, it is broken as before however the objective of r301461 was only
> to address the ABI breakage. If I revert r301777 then I have to backout
> r301461 too, and then we have again the ABI breakage.
>
> I understand you want to fix all but we are under code freeze and
> I have to go back to a known working (although still broken) state.
I can't make strcoll or vfscanf to drop core in my simple tests.
Too little info in the bug report.
In any case here is vfscanf.c fix attached (by removing collation range).
>> It is dependent. vfscanf() and regcomp() use the same code for range
>> collation. If we remove range collation from everywhere, we save
>> vfscanf(), but strcoll_l() remains broken.
>>
>>
>
> We have had a broken regex for a very long time, and we new that
> before collation was introduced. I didn't take that decision but
> moving to libtre was a lot more work and we really had to have
> collation now.
No, we have perfectly working single byte regex with collation ranges
for many years until those wchars was invaded the code very recently.
Index: vfscanf.c
===
--- vfscanf.c (revision 301779)
+++ vfscanf.c (working copy)
@@ -816,9 +816,7 @@
static const u_char *
__sccl(char *tab, const u_char *fmt)
{
- int c, n, v, i;
- struct xlocale_collate *table =
- (struct
xlocale_collate*)__get_locale()->components[XLC_COLLATE];
+ int c, n, v;
/* first `clear' the whole table */
c = *fmt++; /* first char hat => negated scanset */
@@ -871,29 +869,15 @@
* we just stored in the table (c).
*/
n = *fmt;
- if (n == ']'
- || (table->__collate_load_error ? n < c :
- __wcollate_range_cmp(table, n, c) < 0
- )
- ) {
+ if (n == ']' || n < c) {
c = '-';
break; /* resume the for(;;) */
}
fmt++;
- /* fill in the range */
- if (table->__collate_load_error) {
- do {
- tab[++c] = v;
- } while (c < n);
- } else {
- for (i = 0; i < 256; i ++)
- if (__wcollate_range_cmp(table, c, i) <
0 &&
- __wcollate_range_cmp(table, i, n)
<= 0
- )
- tab[i] = v;
- }
+ do {/* fill in the range */
+ tab[++c] = v;
+ } while (c < n);
#if 1 /* XXX another disgusting compatibility hack */
- c = n;
/*
* Alas, the V7 Unix scanf also treats formats
* such as [a-c-e] as `the letters a through e'.
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r301797 - head/release/doc/en_US.ISO8859-1/relnotes
Author: lidl Date: Fri Jun 10 15:05:35 2016 New Revision: 301797 URL: https://svnweb.freebsd.org/changeset/base/301797 Log: Relnotes entries for blacklist project Approved by: re (gjb) Sponsored by: The FreeBSD Foundation Modified: head/release/doc/en_US.ISO8859-1/relnotes/article.xml Modified: head/release/doc/en_US.ISO8859-1/relnotes/article.xml == --- head/release/doc/en_US.ISO8859-1/relnotes/article.xml Fri Jun 10 15:03:45 2016(r301796) +++ head/release/doc/en_US.ISO8859-1/relnotes/article.xml Fri Jun 10 15:05:35 2016(r301797) @@ -608,6 +608,29 @@ OpenBSM has been updated to version 1.2 alpha 4. + The NetBSD + Project's &man.libblacklist.3; library and applications + have been ported and integrated into the system. Packet + filtering support for the &man.pf.4; packet filtering systems + has been implemented. The blacklist + system provides the blacklistd + daemon, the helper script + blacklistd-helper to make changes + to the running packet filter system and the + blacklistctl control program. + A selection of system daemons, including: + fingerd, + ftpd, + rlogind, + rshd, and + sshd have been modified to support + sending notifications to the blacklistd + daemon. + + Support for + the &man.ipfw.4; packet filter has been added to the + blacklistd-helper script. + ___ svn-src-head@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r301800 - head/sys/rpc
Author: ngie
Date: Fri Jun 10 17:53:28 2016
New Revision: 301800
URL: https://svnweb.freebsd.org/changeset/base/301800
Log:
Deobfuscate cleanup path in clnt_bck_create(..)
Similar to r300836, cl and ct will always be non-NULL as they're allocated
using the mem_alloc routines, which always use `malloc(..., M_WAITOK)`.
Deobfuscating the cleanup path fixes a leak where if cl was NULL and
ct was not, ct would not be free'd, and also removes a duplicate test for
cl not being NULL.
Approved by: re (gjb)
Differential Revision: https://reviews.freebsd.org/D6801
MFC after: 1 week
Reported by: Coverity
CID: 122
Reviewed by: cem
Sponsored by: EMC / Isilon Storage Division
Modified:
head/sys/rpc/clnt_bck.c
Modified: head/sys/rpc/clnt_bck.c
==
--- head/sys/rpc/clnt_bck.c Fri Jun 10 15:47:20 2016(r301799)
+++ head/sys/rpc/clnt_bck.c Fri Jun 10 17:53:28 2016(r301800)
@@ -175,14 +175,9 @@ clnt_bck_create(
return (cl);
err:
- if (cl) {
- if (ct) {
- mtx_destroy(&ct->ct_lock);
- mem_free(ct, sizeof (struct ct_data));
- }
- if (cl)
- mem_free(cl, sizeof (CLIENT));
- }
+ mtx_destroy(&ct->ct_lock);
+ mem_free(ct, sizeof (struct ct_data));
+ mem_free(cl, sizeof (CLIENT));
return (NULL);
}
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
svn commit: r301814 - in head/sys: conf modules
Author: jtl
Date: Fri Jun 10 19:06:11 2016
New Revision: 301814
URL: https://svnweb.freebsd.org/changeset/base/301814
Log:
Change the default build behavior so we don't compile extra TCP modules by
default. At least initially, the feature to support multiple TCP stacks is
aimed at supporting advanced use cases and TCP development, but it is not
necessarily aimed at a wide audience. Therefore, there is no need to build
and install the extra TCP stacks by default. Instead, the people who are
using or developing this functionality can add the extra option to build/
install the extra TCP stacks.
However, we do want to build the extra TCP stacks as part of test builds
(e.g. LINT or tinderbox) to ensure that developers who are testing their
changes will know that their changes do not break the additional TCP
stack modules.
After this change, a user will need to add WITH_EXTRA_TCP_STACKS=1 to
make.conf or the kernel config in order to build the extra TCP modules.
Differential Revision:https://reviews.freebsd.org/D6795
Reviewed by: sjg
Approved by: re (kib)
Modified:
head/sys/conf/kern.opts.mk
head/sys/conf/kern.post.mk
head/sys/modules/Makefile
Modified: head/sys/conf/kern.opts.mk
==
--- head/sys/conf/kern.opts.mk Fri Jun 10 18:47:11 2016(r301813)
+++ head/sys/conf/kern.opts.mk Fri Jun 10 19:06:11 2016(r301814)
@@ -45,6 +45,7 @@ __DEFAULT_YES_OPTIONS = \
__DEFAULT_NO_OPTIONS = \
EISA \
+EXTRA_TCP_STACKS \
NAND \
OFED
Modified: head/sys/conf/kern.post.mk
==
--- head/sys/conf/kern.post.mk Fri Jun 10 18:47:11 2016(r301813)
+++ head/sys/conf/kern.post.mk Fri Jun 10 19:06:11 2016(r301814)
@@ -23,6 +23,10 @@ MKMODULESENV+= CONF_CFLAGS="${CONF_CFLAG
MKMODULESENV+= WITH_CTF="${WITH_CTF}"
.endif
+.if defined(WITH_EXTRA_TCP_STACKS)
+MKMODULESENV+= WITH_EXTRA_TCP_STACKS="${WITH_EXTRA_TCP_STACKS}"
+.endif
+
# Allow overriding the kernel debug directory, so kernel and user debug may be
# installed in different directories. Setting it to "" restores the historical
# behavior of installing debug files in the kernel directory.
Modified: head/sys/modules/Makefile
==
--- head/sys/modules/Makefile Fri Jun 10 18:47:11 2016(r301813)
+++ head/sys/modules/Makefile Fri Jun 10 19:06:11 2016(r301814)
@@ -353,7 +353,7 @@ SUBDIR= \
${_syscons} \
sysvipc \
${_ti} \
- tcp/fastpath \
+ ${_tcp_fastpath} \
tests/framework \
tests/callout_test \
tl \
@@ -436,6 +436,10 @@ _random_other= random_other
SUBDIR+= cuse
.endif
+.if ${MK_EXTRA_TCP_STACKS} != "no" || defined(ALL_MODULES)
+_tcp_fastpath= tcp/fastpath
+.endif
+
.if (${MK_INET_SUPPORT} != "no" || ${MK_INET6_SUPPORT} != "no") || \
defined(ALL_MODULES)
_carp= carp
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r301572 - in head/lib/libcasper: libcasper services/cap_dns services/cap_grp services/cap_pwd services/cap_random services/cap_sysctl
On Fri, Jun 10, 2016 at 04:47:28PM +0200, Jilles Tjoelker wrote:
> On Wed, Jun 08, 2016 at 02:03:53AM +, Mariusz Zaborski wrote:
> > Author: oshogbo
> > Date: Wed Jun 8 02:03:53 2016
> > New Revision: 301572
> > URL: https://svnweb.freebsd.org/changeset/base/301572
>
> > Log:
> > Add flags to the Casper services.
>
> > CASPER_SERVICE_STDIO - Casper will not close the first three descriptors
> > (stdin,
> >stdout and stderr) this can be helpful for debugging.
> > CASPER_SERVICE_FD - Capser will not close all other descriptors, this can
> > be useful for a filesystem service.
>
> This reminds me that there are some common cases where it is wrong to
> close descriptors you don't know about. The non-POSIX command
> diff <(cmd1) <(cmd2)
> that compares the outputs of the two commands, when executed with bash
> that was compiled with the full /dev/fd visible (as in poudriere), will
> actually run something like
> diff /dev/fd/63 /dev/fd/62
> passing two file descriptors to pipes.
>
> When created by a shell, these pathnames will start with /dev/fd/, but
> people could create symlinks to these special files.
This is why the CASPER_SERVICE_FD was added. :)
> > [snip]
> > +static void
> > +stdnull(void)
> > +{
> > + int fd;
> > +
> > + fd = open(_PATH_DEVNULL, O_RDWR);
> > + if (fd == -1)
> > + errx(1, "Unable to open %s", _PATH_DEVNULL);
> > +
> > + if (setsid() == -1)
> > + errx(1, "Unable to detach from session");
>
> There is an implicit assumption here that stdnull() is only called from
> a process that was forked off from here, since setsid() will not and
> cannot work when called from a process that is already a session leader.
>
> If the application is running from a shell, this setsid() will exclude
> the process from most signals, including terminal ^C/^\/^Z, kill % and
> hangups. More generally, this might make it more likely for the process
> to hang around indefinitely after the parent is gone.
I'm not sure but if the process descriptor not solve that?
If we close all process descriptor to the process it should die then, so you
need to kill just the process which is using service.
> > +
> > + if (dup2(fd, STDIN_FILENO) == -1)
> > + errx(1, "Unable to cover stdin");
> > + if (dup2(fd, STDOUT_FILENO) == -1)
> > + errx(1, "Unable to cover stdout");
> > + if (dup2(fd, STDERR_FILENO) == -1)
> > + errx(1, "Unable to cover stderr");
> > +
> > + close(fd);
>
> This was not broken by this commit, but fd should not be closed if it is
> equal to STDIN_FILENO, STDOUT_FILENO or STDERR_FILENO.
Yes you are in 100% right. Thanks, I will fix that.
Thanks,
--
Mariusz Zaborski
oshogbo//vx | http://oshogbo.vexillium.org
FreeBSD commiter| https://freebsd.org
Software developer | http://wheelsystems.com
If it's not broken, let's fix it till it is!!1
signature.asc
Description: PGP signature
svn commit: r301815 - head/sys/conf
Author: gjb
Date: Fri Jun 10 19:29:55 2016
New Revision: 301815
URL: https://svnweb.freebsd.org/changeset/base/301815
Log:
Update 11.0 to ALPHA3 in preparation for new snapshot builds.
Approved by: re (implicit)
Sponsored by: The FreeBSD Foundation
Modified:
head/sys/conf/newvers.sh
Modified: head/sys/conf/newvers.sh
==
--- head/sys/conf/newvers.shFri Jun 10 19:06:11 2016(r301814)
+++ head/sys/conf/newvers.shFri Jun 10 19:29:55 2016(r301815)
@@ -32,7 +32,7 @@
TYPE="FreeBSD"
REVISION="11.0"
-BRANCH="ALPHA2"
+BRANCH="ALPHA3"
if [ -n "${BRANCH_OVERRIDE}" ]; then
BRANCH=${BRANCH_OVERRIDE}
fi
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
Re: svn commit: r301572 - in head/lib/libcasper: libcasper services/cap_dns services/cap_grp services/cap_pwd services/cap_random services/cap_sysctl
On Fri, Jun 10, 2016 at 09:17:45PM +0200, Mariusz Zaborski wrote: > On Fri, Jun 10, 2016 at 04:47:28PM +0200, Jilles Tjoelker wrote: > > On Wed, Jun 08, 2016 at 02:03:53AM +, Mariusz Zaborski wrote: > > > + if (setsid() == -1) > > > + errx(1, "Unable to detach from session"); > > There is an implicit assumption here that stdnull() is only called from > > a process that was forked off from here, since setsid() will not and > > cannot work when called from a process that is already a session leader. > > If the application is running from a shell, this setsid() will exclude > > the process from most signals, including terminal ^C/^\/^Z, kill % and > > hangups. More generally, this might make it more likely for the process > > to hang around indefinitely after the parent is gone. > I'm not sure but if the process descriptor not solve that? > If we close all process descriptor to the process it should die then, so you > need to kill just the process which is using service. Oh, right. For termination this is probably even better since terminating properly may require casper daemons, and for stopping it probably doesn't matter much. For opening files in utilities called from an interactive shell it is probably still bad since it breaks /dev/tty and obscures blocking opens such as fifos. -- Jilles Tjoelker ___ svn-src-head@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"
