Re: svn commit: r346315 - head/lib/libcasper/services/cap_fileargs
It looks this change is causing 'make xdev TARGET=mips TARGET_ARCH=mips' to fail as the following with HEAD checked out under "/usr/obj/freebsd": cc -O2 -pipe -DWITH_CASPER -I/usr/obj/freebsd/contrib/elftoolchain/libelftc - I/usr/obj/freebsd/contrib/elftoolchain/common -g -MD -MF.depend.strings.o -MTstrings.o -std=gnu99 -Qunused-arguments -c /usr/obj/freebsd/contrib/elftoolchain/strings/strings.c -o strings.o /usr/obj/freebsd/contrib/elftoolchain/strings/strings.c:198:55: error: use of undeclared identifier 'FA_OPEN' fa = fileargs_init(argc, argv, O_RDONLY, 0, &rights, FA_OPEN); ^ 1 error generated. *** [strings.o] Error code 1 It looks #include cannot pick up header files from "lib/libcasper/services/cap_fileargs" Hiro Ed Maste wrote: > Author: emaste > Date: Wed Apr 17 16:02:57 2019 > New Revision: 346315 > URL: https://svnweb.freebsd.org/changeset/base/346315 > > Log: > cap_fileargs: add fileargs_lstat service > > Add fileargs_lstat function to cap_fileargs casper service to be able to > lstat files while in capability mode. It can only lstat files given in > fileargs_init. > > Submitted by: Bora 〓zarslan > Reviewed by:oshogbo, cem (partial) > MFC after: 3 weeks > Relnotes: Yes > Sponsored by: The FreeBSD Foundation > Differential Revision: https://reviews.freebsd.org/D19548 > > Modified: > head/lib/libcasper/services/cap_fileargs/cap_fileargs.3 > head/lib/libcasper/services/cap_fileargs/cap_fileargs.c > head/lib/libcasper/services/cap_fileargs/cap_fileargs.h > > Modified: head/lib/libcasper/services/cap_fileargs/cap_fileargs.3 > == > --- head/lib/libcasper/services/cap_fileargs/cap_fileargs.3 Wed Apr 17 > 16:00:33 > 2019 (r346314) +++ head/lib/libcasper/services/cap_fileargs/cap_fileargs.3 > Wed Apr > 17 16:02:57 2019 (r346315) @@ -24,7 +24,7 @@ > .\" > .\" $FreeBSD$ > .\" > -.Dd November 12, 2018 > +.Dd April 17, 2019 > .Dt CAP_FILEARGS 3 > .Os > .Sh NAME > @@ -33,6 +33,7 @@ > .Nm fileargs_init , > .Nm fileargs_initnv , > .Nm fileargs_free , > +.Nm fileargs_lstat , > .Nm fileargs_open , > .Nm fileargs_fopen > .Nd "library for handling files in capability mode" > @@ -43,9 +44,9 @@ > .In libcasper.h > .In casper/cap_fileargs.h > .Ft "fileargs_t *" > -.Fn fileargs_init "int argc" "char *argv[]" "int flags" "mode_t mode" > "cap_rights_t *rightsp" > +.Fn fileargs_init "int argc" "char *argv[]" "int flags" "mode_t mode" > "cap_rights_t *rightsp" > "int operations" .Ft "fileargs_t *" > -.Fn fileargs_cinit "cap_channel_t *cas" "int argc" "char *argv[]" "int > flags" "mode_t mode" > "cap_rights_t *rightsp" +.Fn fileargs_cinit "cap_channel_t *cas" "int argc" > "char *argv[]" "int > flags" "mode_t mode" "cap_rights_t *rightsp" "int operations" .Ft "fileargs_t > *" > .Fn fileargs_cinitnv "cap_channel_t *cas" "nvlist_t *limits" > .Ft "fileargs_t *" > @@ -53,6 +54,8 @@ > .Ft "void" > .Fn fileargs_free "fileargs_t *fa" > .Ft "int" > +.Fn fileargs_lstat "fileargs_t *fa" "const char *path" "struct stat *sb" > +.Ft "int" > .Fn fileargs_open "fileargs_t *fa" "const char *name" > .Ft "FILE *" > .Fn fileargs_fopen "fileargs_t *fa" "const char *name" "const char *mode" > @@ -97,6 +100,22 @@ The > argument contains a list of the capability rights which file should be > limited to. > For more details of the capability rights see > .Xr cap_rights_init 3 . > +The > +.Fa operations > +argument limits the operations that are available using > +.Nm system.fileargs . > +.Fa operations > +is a combination of: > +.Bl -ohang -offset indent > +.It FA_OPEN > +Allow > +.Fn fileargs_open > +and > +.Fn fileargs_fopen . > +.It FA_LSTAT > +Allow > +.Fn fileargs_lstat . > +.El > .Pp > The function > .Fn fileargs_cinit > @@ -126,6 +145,11 @@ The function handle > .Dv NULL > argument. > .Pp > +The function > +.Fn fileargs_lstat > +is equivalent to > +.Xr lstat 2 . > +.Pp > The functions > .Fn fileargs_open > and > @@ -165,6 +189,15 @@ must contain the > The > .Va mode > argument tells which what mode file should be created. > +.It operations (NV_TYPE_NUMBER) > +The > +.Va operations > +limits the usable operations for > +.Fa system.fileargs . > +The possible values are explained as > +.Va operations > +argument with > +.Fn fileargs_init . > .El > .Pp > The > @@ -201,7 +234,7 @@ argv += optind; > > /* Create capability to the system.fileargs service. */ > fa = fileargs_init(argc, argv, O_RDONLY, 0, > -cap_rights_init(&rights, CAP_READ)); > +cap_rights_init(&rights, CAP_READ), FA_OPEN); > if (fa == NULL) > err(1, "unable to open system.fileargs service"); > > @@ -222,6 +255,7 @@ fileargs_free(fa); > .Ed > .Sh SEE ALSO > .Xr cap_enter 2 , > +.Xr lstat 2 , > .Xr open 2 , > .Xr cap_rights_init 3 , > .Xr err 3 , > > Modified: head/lib/libca
Re: svn commit: r346315 - head/lib/libcasper/services/cap_fileargs
Hi Ed and thank you for taking a look. my svn info says 346593 which is after few other fixes were commited. I'm on i386 arch. I haven't done installworld yet after picking up libcasper changes. 'make buildworld' works fine. 'make xdev-build' fails and I tried with both "arm" and "mips" for TARGET and TARGET_ARCH. Both fail same way. Please check your /usr/include/casper/ca_fileargs.h and I suspect that's where you pick up FA_OPEN. I attached a log file this time. Regards, Hiro On Tue, 23 Apr 2019 09:49:00 -0400 Ed Maste wrote: > On Tue, 23 Apr 2019 at 00:07, Yoshihiro Ota wrote: > > > > It looks this change is causing 'make xdev TARGET=mips TARGET_ARCH=mips' to > > fail as the > > following with HEAD checked out under "/usr/obj/freebsd": > > Hello Hiro-san, sorry about that. > > I tried `make xdev TARGET=mips TARGET_ARCH=mips` on HEAD just now (but > it failed on the install as DESTDIR wasn't set and I ran as non-root). > Just `make xdev-build` was successful though. > > What version were you trying to build? There were (several) followup > commits to address issues with the initial commit of cap_fileargs > lstat support. % svn info Path: . Working Copy Root Path: /usr/src URL: https://svn0.us-east.freebsd.org/base/head Relative URL: ^/head Repository Root: https://svn0.us-east.freebsd.org/base Repository UUID: ccf9f872-aa2e-dd11-9fc8-001c23d0bc1f Revision: 346593 Node Kind: directory Schedule: normal Last Changed Author: wma Last Changed Rev: 346593 Last Changed Date: 2019-04-23 02:36:32 -0400 (Tue, 23 Apr 2019) % make xdev-build TARGET=mips TARGET_ARCH=mips mkdir -p /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr mtree -deUW -f /usr/src/etc/mtree/BSD.usr.dist -p /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr >/dev/null ===> lib/clang/libllvmminimal (obj,all,install) ===> usr.bin/clang/llvm-tblgen (obj,all,install) sh /usr/src/tools/install.sh -s -o root -g wheel -m 555 llvm-tblgen /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/llvm-tblgen sh /usr/src/tools/install.sh -o root -g wheel -m 444 llvm-tblgen.debug /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/debug/usr/bin/llvm-tblgen.debug ===> usr.bin/clang/clang-tblgen (obj,all,install) sh /usr/src/tools/install.sh -s -o root -g wheel -m 555 clang-tblgen /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/clang-tblgen sh /usr/src/tools/install.sh -o root -g wheel -m 444 clang-tblgen.debug /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/debug/usr/bin/clang-tblgen.debug ===> gnu/usr.bin/gperf (obj,all,install) sh /usr/src/tools/install.sh -s -o root -g wheel -m 555 gperf /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/gperf sh /usr/src/tools/install.sh -o root -g wheel -m 444 gperf.debug /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/debug/usr/bin/gperf.debug ===> lib/liby (obj,all,install) sh /usr/src/tools/install.sh -C -o root -g wheel -m 444 liby.a /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/ ===> usr.bin/yacc (obj,all,install) sh /usr/src/tools/install.sh -s -o root -g wheel -m 555 yacc /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/yacc sh /usr/src/tools/install.sh -o root -g wheel -m 444 yacc.debug /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/debug/usr/bin/yacc.debug sh /usr/src/tools/install.sh -l h -o root -g wheel -m 555 /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/yacc /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/byacc ===> bin/csh (obj,build-tools) ===> bin/sh (obj,build-tools) ===> lib/ncurses/ncurses (obj,build-tools) ===> lib/ncurses/ncursesw (obj,build-tools) ===> share/syscons/scrnmaps (obj,build-tools) ===> usr.bin/awk (obj,build-tools) ===> lib/libmagic (obj,build-tools) ===> usr.bin/mkesdb_static (obj,build-tools) ===> usr.bin/mkcsmapper_static (obj,build-tools) ===> usr.bin/vi/catalog (obj,build-tools) ===> gnu/usr.bin/cc/cc_tools (obj,build-tools) ===> xdev gnu/usr.bin/binutils (obj,all) ===> gnu/usr.bin/binutils/libiberty (all) ===> gnu/usr.bin/binutils/libbfd (all) ===> gnu/usr.bin/binutils/libopcodes (all) ===> gnu/usr.bin/binutils/doc (all) ===> gnu/usr.bin/binutils/libbinutils (all) ===> gnu/usr.bin/binutils/as (all) ===> gnu/usr.bin/binutils/objdump (all) ===> gnu/usr.bin/binutils/ld (all) ===> xdev lib/libelftc (obj,all) ===> xdev lib/libpe (obj,all) ===> xdev usr.bin/objcopy (obj,all) ===> xdev usr.bin/nm (obj,all) ===> xdev usr.bin/size (obj,all) ===> xdev usr.bin/strings (obj,all) cc -O2 -pipe -DWITH_CASPER -I/usr/src/contrib/elftoolchain/libelftc -I/usr/src/contrib/elftoolchain/common -g -MD -MF.depend.strings.o -MTstrings.o -std=gnu99 -Qunused-arguments -c /usr/src/contrib/elftoolchain/strings/strings.c -o strings.o /usr/src/contrib/elftoolchain/strings/strings.c:198:55: erro
Re: svn commit: r345804 - head/usr.bin/systat
Hi, On Tue, 2 Apr 2019 07:08:17 -0700 (PDT) "Rodney W. Grimes" wrote: > > Author: mr > > Date: Tue Apr 2 14:01:03 2019 > > New Revision: 345804 > > URL: https://svnweb.freebsd.org/changeset/base/345804 > > > > Log: > > systat -zarc to display disk activities like -vm > > > > PR: 213310 > > Submitted by: ota > > MFH: 4 weeks > > ? MFC: > > > Differential Revision:https://reviews.freebsd.org/D18726 > > > > Modified: > > head/usr.bin/systat/devs.c > > head/usr.bin/systat/devs.h > > head/usr.bin/systat/iostat.c > > head/usr.bin/systat/swap.c > > head/usr.bin/systat/systat.h > > head/usr.bin/systat/vmstat.c > > head/usr.bin/systat/zarc.c > > > > Modified: head/usr.bin/systat/devs.c > > == > > --- head/usr.bin/systat/devs.c Tue Apr 2 13:59:04 2019 > > (r345803) > > +++ head/usr.bin/systat/devs.c Tue Apr 2 14:01:03 2019 > > (r345804) > > @@ -2,6 +2,7 @@ > > * SPDX-License-Identifier: BSD-3-Clause > > * > > * Copyright (c) 1998 Kenneth D. Merry. > > + * 2015 Yoshihiro Ota > > * All rights reserved. > > > Can we get in contact with Yoshihiro Ota about his > copyright statements, and correcting this to make > it clear that it is Kenneth D. Merry that is asserting > "All rights reserved" and if Ota does nor does not wish to assert > "All rights reserved". > > As committed this makes a grey area on Kenneth's assertion, > also leaving out the word copyright on Yoshihiro's line is a bit iffy. > > I am only commenting once, this issue appears several times. > We can go back out to D18726 to discuss it if need be. I've fully written zarc.c (and copied copy-right section from another file) but all other changes in other files are refactoring and adjustments. I don't know the policy/procedure of how to update copy right section of FreeBSD. Please adjust them according the project. Regards, Hiro ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r348261 - in stable/11: contrib/zlib lib/libz stand/libsa usr.bin/minigzip
Hi Xin,
I didn't think we would have time for this to hit 11.3 release.
Next important one is the crc32 one.
Unless we do this split, we will need a special aid to work-around zlib/zlib.h
compile error and also zib/crc32.c won't compile at all.
Regards,
Hiro
On Sat, 25 May 2019 01:58:00 + (UTC)
Xin LI wrote:
> Author: delphij
> Date: Sat May 25 01:58:00 2019
> New Revision: 348261
> URL: https://svnweb.freebsd.org/changeset/base/348261
>
> Log:
> MFC r347244:
>
> Move contrib/zlib to sys/contrib/zlib so that we can use it in kernel.
> This is a prerequisite of unifying kernel zlib instances.
>
> Submitted by: Yoshihiro Ota
> Approved by:re (kib)
>
> Added:
> - copied from r347244, head/sys/contrib/zlib/
> Directory Properties:
> stable/11/sys/contrib/zlib/ (props changed)
> Deleted:
> stable/11/contrib/zlib/
> Modified:
> stable/11/lib/libz/FREEBSD-upgrade
> stable/11/lib/libz/Makefile
> stable/11/stand/libsa/Makefile
> stable/11/usr.bin/minigzip/Makefile
> Directory Properties:
> stable/11/ (props changed)
>
> Modified: stable/11/lib/libz/FREEBSD-upgrade
> ==
> --- stable/11/lib/libz/FREEBSD-upgradeSat May 25 00:22:07 2019
> (r348260)
> +++ stable/11/lib/libz/FREEBSD-upgradeSat May 25 01:58:00 2019
> (r348261)
> @@ -8,7 +8,7 @@ benefit other consumers.
> To Update:
>1) Unpack vendor sources into a clean directory.
>2) Import onto the vendor area.
> - 3) Merge the vendor tree to contrib/zlib, which contains a stripped down
> + 3) Merge the vendor tree to sys/contrib/zlib, which contains a stripped
> down
> version of upstream source, resolve any conflicts.
>4) Double check zconf.h, zlib.pc, and Symbol.map to make sure that we
> have the required changes. Test universe and commit them.
>
> Modified: stable/11/lib/libz/Makefile
> ==
> --- stable/11/lib/libz/Makefile Sat May 25 00:22:07 2019
> (r348260)
> +++ stable/11/lib/libz/Makefile Sat May 25 01:58:00 2019
> (r348261)
> @@ -8,7 +8,7 @@ SHLIBDIR?=/lib
> SHLIB_MAJOR= 6
> MAN= zlib.3 zopen.3
>
> -ZLIBSRC= ${SRCTOP}/contrib/zlib
> +ZLIBSRC= ${SRCTOP}/sys/contrib/zlib
>
> .PATH: ${ZLIBSRC}
>
>
> Modified: stable/11/stand/libsa/Makefile
> ==
> --- stable/11/stand/libsa/MakefileSat May 25 00:22:07 2019
> (r348260)
> +++ stable/11/stand/libsa/MakefileSat May 25 01:58:00 2019
> (r348261)
> @@ -99,8 +99,8 @@ CFLAGS+= -DBZ_NO_STDIO -DBZ_NO_COMPRESS
> SRCS+=bzlib.c crctable.c decompress.c huffman.c randtable.c
>
> # decompression functionality from zlib
> -.PATH: ${SRCTOP}/contrib/zlib
> -CFLAGS+=-DHAVE_MEMCPY -I${SRCTOP}/contrib/zlib
> +.PATH: ${SRCTOP}/sys/contrib/zlib
> +CFLAGS+=-DHAVE_MEMCPY -I${SRCTOP}/sys/contrib/zlib
> SRCS+= adler32.c crc32.c
> SRCS+= infback.c inffast.c inflate.c inftrees.c zutil.c
>
>
> Modified: stable/11/usr.bin/minigzip/Makefile
> ==
> --- stable/11/usr.bin/minigzip/Makefile Sat May 25 00:22:07 2019
> (r348260)
> +++ stable/11/usr.bin/minigzip/Makefile Sat May 25 01:58:00 2019
> (r348261)
> @@ -1,6 +1,6 @@
> # $FreeBSD$
>
> -SRCDIR= ${SRCTOP}/contrib/zlib/test
> +SRCDIR= ${SRCTOP}/sys/contrib/zlib/test
> .PATH: ${SRCDIR}
>
> PROG=minigzip
> ___
> svn-src-all@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/svn-src-all
> To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r347968 - head/sys/kern
Hi Konstantin,
I coundn't remount ro->rw a few days ago.
This change isn't related, then.
I will check/tesst it again later, then.
Thanks,
Hiro
On Sun, 19 May 2019 20:51:04 +0300
Konstantin Belousov wrote:
> On Sun, May 19, 2019 at 11:52:28AM -0400, Yoshihiro Ota wrote:
> > Hi,
> >
> > Does this change fix 'mount -u -rw /path' to remount read-only device with
> > write mode?
> > I moticed an issue a couple of days ago but didn't have time to conform nor
> > report.
>
> I do not understand your description. If you mean does this commit has
> any relevance to the issue of remounting ro->rw, it should not. It only
> affects rw->ro remounts, when you have a binary executed from the mount
> point.
>
> >
> > Thanks,
> > Hiro
> >
> > On Sun, 19 May 2019 09:18:10 + (UTC)
> > Konstantin Belousov wrote:
> >
> > > Author: kib
> > > Date: Sun May 19 09:18:09 2019
> > > New Revision: 347968
> > > URL: https://svnweb.freebsd.org/changeset/base/347968
> > >
> > > Log:
> > > Fix rw->ro remount when there is a text vnode mapping.
> > >
> > > Reported and tested by: hrs
> > > Sponsored by: The FreeBSD Foundation
> > > MFC after: 16 days
> > >
> > > Modified:
> > > head/sys/kern/vfs_subr.c
> > >
> > > Modified: head/sys/kern/vfs_subr.c
> > > ==
> > > --- head/sys/kern/vfs_subr.c Sun May 19 06:01:11 2019
> > > (r347967)
> > > +++ head/sys/kern/vfs_subr.c Sun May 19 09:18:09 2019
> > > (r347968)
> > > @@ -3146,7 +3146,7 @@ loop:
> > >
> > > if ((vp->v_type == VNON ||
> > > (error == 0 && vattr.va_nlink > 0)) &&
> > > - (vp->v_writecount == 0 || vp->v_type != VREG)) {
> > > + (vp->v_writecount <= 0 || vp->v_type != VREG)) {
> > > VOP_UNLOCK(vp, 0);
> > > vdropl(vp);
> > > continue;
> > > ___
> > > svn-src-all@freebsd.org mailing list
> > > https://lists.freebsd.org/mailman/listinfo/svn-src-all
> > > To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r347968 - head/sys/kern
Hi,
Does this change fix 'mount -u -rw /path' to remount read-only device with
write mode?
I moticed an issue a couple of days ago but didn't have time to conform nor
report.
Thanks,
Hiro
On Sun, 19 May 2019 09:18:10 + (UTC)
Konstantin Belousov wrote:
> Author: kib
> Date: Sun May 19 09:18:09 2019
> New Revision: 347968
> URL: https://svnweb.freebsd.org/changeset/base/347968
>
> Log:
> Fix rw->ro remount when there is a text vnode mapping.
>
> Reported and tested by: hrs
> Sponsored by: The FreeBSD Foundation
> MFC after: 16 days
>
> Modified:
> head/sys/kern/vfs_subr.c
>
> Modified: head/sys/kern/vfs_subr.c
> ==
> --- head/sys/kern/vfs_subr.c Sun May 19 06:01:11 2019(r347967)
> +++ head/sys/kern/vfs_subr.c Sun May 19 09:18:09 2019(r347968)
> @@ -3146,7 +3146,7 @@ loop:
>
> if ((vp->v_type == VNON ||
> (error == 0 && vattr.va_nlink > 0)) &&
> - (vp->v_writecount == 0 || vp->v_type != VREG)) {
> + (vp->v_writecount <= 0 || vp->v_type != VREG)) {
> VOP_UNLOCK(vp, 0);
> vdropl(vp);
> continue;
> ___
> svn-src-all@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/svn-src-all
> To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r347951 - stable/12/lib/libc/stdlib
I wonder if we can use a tool to confirm coding style like
clang-format or something else.
I tried to setup clang-format and 2 others in the past like
a year ago but wasn't able to do...
Hiro
On Sat, 18 May 2019 06:06:51 -0700 (PDT)
"Rodney W. Grimes" wrote:
> > Hello Konstantin and Bruce,
> >
> > thank you for your comments. Clearly, this should have been addressed in
> > the review and checked by a second pair of eyes. I mostly did the man
> > page changes and can't comment much on the actual coding example. Two
> > people (dab and jilles) approved the revision in the Phabricator review.
> > There was plenty of time to review it between April 16 and May 15 when
> > then the original submitter asked if there was anything else that needs
> > to be done.
>
> The quality of a code review is highly dependent on who is invited
> to do the reviewing, and how much participation and care those invited
> have with respect to the process. Man pages that contain code samples
> need to have more reviewers, you need man page experts, you need subject
> mater experts, and you need coding experts. This review could of used
> more reviewers (hind sight is 20/20) but if they had not been the right
> reviewers the low quality commit would of probably still occurred.
>
> The amount of time in review is not a good measure of review completness,
> that falls more onto the quality of the workmanship of those actually
> commented on a review. I myself consider a LGTM, or a accept by a reviewer
> who makes 0 comments on my code in phab as if that reviewer probably
> just rubber stamped my change, UNLESS I have extremly high trust in
> that person. If bde, imp, or kib stamp an approve on something of mine my
> confidence level is very high and I do consider that code well reviewed,
> if I get a rubber stamp from some others I am not so confident.
>
> Now we have the problem in that note every code review can be done by
> such high quality engineers due to time and work load constraints, so
> we need to raise the level of others closer to these shinning stars
> and offset this lack of avaliable resources by other means (more
> reviewers, long review cycles, automations, etc.)
> >
> > Now, having said that and the the change is now both in HEAD and
> > stable/12, how do we proceed? Revert them both or patch those issues as
> > follow-ups? Note that a src committer should do that and not me trying
> > to add more EXAMPLE sections to man pages with my doc commit bit.
>
> There is another option, a revert is actually a local operation
> to your checked out code, ou can check out head, revert this
> commit, but do not commit, make the corrections, and then commit.
>
> When MFC's this single commit does the right thing.
>
> You could also revert in head, commit and later fix these
> issues, commit, and then MFC both of those commits.
>
> All of your proposed solutions, and my additions are workable,
> and I really have no preference on a single file change like
> this.
>
> Good Day,
> Rod
>
> > Regards,
> > Benedict
> >
> >
> >
> > Am 18.05.19 um 07:45 schrieb Bruce Evans:
> > > On Sat, 18 May 2019, Konstantin Belousov wrote:
> > >
> > >> On Sat, May 18, 2019 at 03:15:08AM +, Benedict Reuschling wrote:
> > >>> Author: bcr (doc committer)
> > >>> Date: Sat May 18 03:15:07 2019
> > >>> New Revision: 347951
> > >>> URL: https://svnweb.freebsd.org/changeset/base/347951
> > >>>
> > >>> Log:
> > >>> ? MFC r347617:
> > >>> ? Add small EXAMPLE section to bsearch.3.
> > >>>
> > >>> ? Submitted by:??? fernape (via Phabricator)
> > >>> ? Reviewed by:??? bcr, jilles, dab
> > >>> ? Approved by:??? bcr (man pages), jilles (src)
> > >>> ? Differential Revision:??? https://reviews.freebsd.org/D19902
> > >>>
> > >>> Modified:
> > >>> ? stable/12/lib/libc/stdlib/bsearch.3
> > >>> Directory Properties:
> > >>> ? stable/12/?? (props changed)
> > >>>
> > >>> Modified: stable/12/lib/libc/stdlib/bsearch.3
> > >>> ==
> > >>>
> > >>> --- stable/12/lib/libc/stdlib/bsearch.3??? Sat May 18 02:02:14
> > >>> 2019??? (r347950)
> > >>> +++ stable/12/lib/libc/stdlib/bsearch.3??? Sat May 18 03:15:07
> > >>> 2019??? (r347951)
> > >>> @@ -32,7 +32,7 @@
> > >>> ?.\" @(#)bsearch.3??? 8.3 (Berkeley) 4/19/94
> > >>> ?.\" $FreeBSD$
> > >>> ?.\"
> > >>> -.Dd February 22, 2013
> > >>> +.Dd May 15, 2019
> > >>> ?.Dt BSEARCH 3
> > >>> ?.Os
> > >>> ?.Sh NAME
> > >>> @@ -83,6 +83,61 @@ The
> > >>> ?function returns a pointer to a matching member of the array, or a null
> > >>> ?pointer if no match is found.
> > >>> ?If two members compare as equal, which member is matched is
> > >>> unspecified.
> > >>> +.Sh EXAMPLES
> > >>> +A sample program that searches people by age in a sorted array:
> > >>> +.Bd -literal
> > >>> +#include
> > >>> +#include
> > >>> +#include
> > >>> +#include
> > >>> +#include
> > >>> +
> > >>> +struct person {
> > >>> +?
Re: svn commit: r346315 - head/lib/libcasper/services/cap_fileargs
I looked into this more. I started wondering if "xdev" has been obsolete after /usr/obj/. dir chaneg. Both of these commands worked okay: % make kernel-toolchain kernel TARGET_ARCH=mips KERNCONF=PB92 -C /usr/src % make buildworld TARGET_ARCH=arm -C /usr/src While I kept getting compile errors by: % make xdev-build TARGET_ARCH=mips TARGET=mips -C /usr/src In other words, if I don't run explict xdev command, buildworld and buildkernel compile okay. Hiro On Wed, 24 Apr 2019 00:24:34 -0400 Yoshihiro Ota wrote: > Hi Ed and thank you for taking a look. > > my svn info says 346593 which is after few other fixes were commited. > > I'm on i386 arch. > I haven't done installworld yet after picking up libcasper changes. > 'make buildworld' works fine. > 'make xdev-build' fails and I tried with both "arm" and "mips" for TARGET and > TARGET_ARCH. > Both fail same way. > Please check your /usr/include/casper/ca_fileargs.h and I suspect that's > where you pick up > FA_OPEN. > > I attached a log file this time. > > Regards, > Hiro > > On Tue, 23 Apr 2019 09:49:00 -0400 > Ed Maste wrote: > > > On Tue, 23 Apr 2019 at 00:07, Yoshihiro Ota wrote: > > > > > > It looks this change is causing 'make xdev TARGET=mips TARGET_ARCH=mips' > > > to fail as the > > > following with HEAD checked out under "/usr/obj/freebsd": > > > > Hello Hiro-san, sorry about that. > > > > I tried `make xdev TARGET=mips TARGET_ARCH=mips` on HEAD just now (but > > it failed on the install as DESTDIR wasn't set and I ran as non-root). > > Just `make xdev-build` was successful though. > > > > What version were you trying to build? There were (several) followup > > commits to address issues with the initial commit of cap_fileargs > > lstat support. ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r346315 - head/lib/libcasper/services/cap_fileargs
Hi Ed and thank you for taking a look. my svn info says 346593 which is after few other fixes were commited. I'm on i386 arch. I haven't done installworld yet after picking up libcasper changes. 'make buildworld' works fine. 'make xdev-build' fails and I tried with both "arm" and "mips" for TARGET and TARGET_ARCH. Both fail same way. Please check your /usr/include/casper/ca_fileargs.h and I suspect that's where you pick up FA_OPEN. I attached a log file this time. Regards, Hiro On Tue, 23 Apr 2019 09:49:00 -0400 Ed Maste wrote: > On Tue, 23 Apr 2019 at 00:07, Yoshihiro Ota wrote: > > > > It looks this change is causing 'make xdev TARGET=mips TARGET_ARCH=mips' to > > fail as the > > following with HEAD checked out under "/usr/obj/freebsd": > > Hello Hiro-san, sorry about that. > > I tried `make xdev TARGET=mips TARGET_ARCH=mips` on HEAD just now (but > it failed on the install as DESTDIR wasn't set and I ran as non-root). > Just `make xdev-build` was successful though. > > What version were you trying to build? There were (several) followup > commits to address issues with the initial commit of cap_fileargs > lstat support. % svn info Path: . Working Copy Root Path: /usr/src URL: https://svn0.us-east.freebsd.org/base/head Relative URL: ^/head Repository Root: https://svn0.us-east.freebsd.org/base Repository UUID: ccf9f872-aa2e-dd11-9fc8-001c23d0bc1f Revision: 346593 Node Kind: directory Schedule: normal Last Changed Author: wma Last Changed Rev: 346593 Last Changed Date: 2019-04-23 02:36:32 -0400 (Tue, 23 Apr 2019) % make xdev-build TARGET=mips TARGET_ARCH=mips mkdir -p /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr mtree -deUW -f /usr/src/etc/mtree/BSD.usr.dist -p /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr >/dev/null ===> lib/clang/libllvmminimal (obj,all,install) ===> usr.bin/clang/llvm-tblgen (obj,all,install) sh /usr/src/tools/install.sh -s -o root -g wheel -m 555 llvm-tblgen /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/llvm-tblgen sh /usr/src/tools/install.sh -o root -g wheel -m 444 llvm-tblgen.debug /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/debug/usr/bin/llvm-tblgen.debug ===> usr.bin/clang/clang-tblgen (obj,all,install) sh /usr/src/tools/install.sh -s -o root -g wheel -m 555 clang-tblgen /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/clang-tblgen sh /usr/src/tools/install.sh -o root -g wheel -m 444 clang-tblgen.debug /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/debug/usr/bin/clang-tblgen.debug ===> gnu/usr.bin/gperf (obj,all,install) sh /usr/src/tools/install.sh -s -o root -g wheel -m 555 gperf /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/gperf sh /usr/src/tools/install.sh -o root -g wheel -m 444 gperf.debug /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/debug/usr/bin/gperf.debug ===> lib/liby (obj,all,install) sh /usr/src/tools/install.sh -C -o root -g wheel -m 444 liby.a /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/ ===> usr.bin/yacc (obj,all,install) sh /usr/src/tools/install.sh -s -o root -g wheel -m 555 yacc /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/yacc sh /usr/src/tools/install.sh -o root -g wheel -m 444 yacc.debug /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/lib/debug/usr/bin/yacc.debug sh /usr/src/tools/install.sh -l h -o root -g wheel -m 555 /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/yacc /usr/obj/usr/src/mips.mips/mips-freebsd/tmp/usr/bin/byacc ===> bin/csh (obj,build-tools) ===> bin/sh (obj,build-tools) ===> lib/ncurses/ncurses (obj,build-tools) ===> lib/ncurses/ncursesw (obj,build-tools) ===> share/syscons/scrnmaps (obj,build-tools) ===> usr.bin/awk (obj,build-tools) ===> lib/libmagic (obj,build-tools) ===> usr.bin/mkesdb_static (obj,build-tools) ===> usr.bin/mkcsmapper_static (obj,build-tools) ===> usr.bin/vi/catalog (obj,build-tools) ===> gnu/usr.bin/cc/cc_tools (obj,build-tools) ===> xdev gnu/usr.bin/binutils (obj,all) ===> gnu/usr.bin/binutils/libiberty (all) ===> gnu/usr.bin/binutils/libbfd (all) ===> gnu/usr.bin/binutils/libopcodes (all) ===> gnu/usr.bin/binutils/doc (all) ===> gnu/usr.bin/binutils/libbinutils (all) ===> gnu/usr.bin/binutils/as (all) ===> gnu/usr.bin/binutils/objdump (all) ===> gnu/usr.bin/binutils/ld (all) ===> xdev lib/libelftc (obj,all) ===> xdev lib/libpe (obj,all) ===> xdev usr.bin/objcopy (obj,all) ===> xdev usr.bin/nm (obj,all) ===> xdev usr.bin/size (obj,all) ===> xdev usr.bin/strings (obj,all) cc -O2 -pipe -DWITH_CASPER -I/usr/src/contrib/elftoolchain/libelftc -I/usr/src/contrib/elftoolchain/common -g -MD -MF.depend.strings.o -MTstrings.o -std=gnu99 -Qunused-arguments -c /usr/src/contrib/elftoolchain/strings/strings.c -o strings.o /usr/src/contrib/elftoolchain/strings/strings.c:198:55: erro
Re: svn commit: r346315 - head/lib/libcasper/services/cap_fileargs
It looks this change is causing 'make xdev TARGET=mips TARGET_ARCH=mips' to fail as the following with HEAD checked out under "/usr/obj/freebsd": cc -O2 -pipe -DWITH_CASPER -I/usr/obj/freebsd/contrib/elftoolchain/libelftc - I/usr/obj/freebsd/contrib/elftoolchain/common -g -MD -MF.depend.strings.o -MTstrings.o -std=gnu99 -Qunused-arguments -c /usr/obj/freebsd/contrib/elftoolchain/strings/strings.c -o strings.o /usr/obj/freebsd/contrib/elftoolchain/strings/strings.c:198:55: error: use of undeclared identifier 'FA_OPEN' fa = fileargs_init(argc, argv, O_RDONLY, 0, &rights, FA_OPEN); ^ 1 error generated. *** [strings.o] Error code 1 It looks #include cannot pick up header files from "lib/libcasper/services/cap_fileargs" Hiro Ed Maste wrote: > Author: emaste > Date: Wed Apr 17 16:02:57 2019 > New Revision: 346315 > URL: https://svnweb.freebsd.org/changeset/base/346315 > > Log: > cap_fileargs: add fileargs_lstat service > > Add fileargs_lstat function to cap_fileargs casper service to be able to > lstat files while in capability mode. It can only lstat files given in > fileargs_init. > > Submitted by: Bora 〓zarslan > Reviewed by:oshogbo, cem (partial) > MFC after: 3 weeks > Relnotes: Yes > Sponsored by: The FreeBSD Foundation > Differential Revision: https://reviews.freebsd.org/D19548 > > Modified: > head/lib/libcasper/services/cap_fileargs/cap_fileargs.3 > head/lib/libcasper/services/cap_fileargs/cap_fileargs.c > head/lib/libcasper/services/cap_fileargs/cap_fileargs.h > > Modified: head/lib/libcasper/services/cap_fileargs/cap_fileargs.3 > == > --- head/lib/libcasper/services/cap_fileargs/cap_fileargs.3 Wed Apr 17 > 16:00:33 > 2019 (r346314) +++ head/lib/libcasper/services/cap_fileargs/cap_fileargs.3 > Wed Apr > 17 16:02:57 2019 (r346315) @@ -24,7 +24,7 @@ > .\" > .\" $FreeBSD$ > .\" > -.Dd November 12, 2018 > +.Dd April 17, 2019 > .Dt CAP_FILEARGS 3 > .Os > .Sh NAME > @@ -33,6 +33,7 @@ > .Nm fileargs_init , > .Nm fileargs_initnv , > .Nm fileargs_free , > +.Nm fileargs_lstat , > .Nm fileargs_open , > .Nm fileargs_fopen > .Nd "library for handling files in capability mode" > @@ -43,9 +44,9 @@ > .In libcasper.h > .In casper/cap_fileargs.h > .Ft "fileargs_t *" > -.Fn fileargs_init "int argc" "char *argv[]" "int flags" "mode_t mode" > "cap_rights_t *rightsp" > +.Fn fileargs_init "int argc" "char *argv[]" "int flags" "mode_t mode" > "cap_rights_t *rightsp" > "int operations" .Ft "fileargs_t *" > -.Fn fileargs_cinit "cap_channel_t *cas" "int argc" "char *argv[]" "int > flags" "mode_t mode" > "cap_rights_t *rightsp" +.Fn fileargs_cinit "cap_channel_t *cas" "int argc" > "char *argv[]" "int > flags" "mode_t mode" "cap_rights_t *rightsp" "int operations" .Ft "fileargs_t > *" > .Fn fileargs_cinitnv "cap_channel_t *cas" "nvlist_t *limits" > .Ft "fileargs_t *" > @@ -53,6 +54,8 @@ > .Ft "void" > .Fn fileargs_free "fileargs_t *fa" > .Ft "int" > +.Fn fileargs_lstat "fileargs_t *fa" "const char *path" "struct stat *sb" > +.Ft "int" > .Fn fileargs_open "fileargs_t *fa" "const char *name" > .Ft "FILE *" > .Fn fileargs_fopen "fileargs_t *fa" "const char *name" "const char *mode" > @@ -97,6 +100,22 @@ The > argument contains a list of the capability rights which file should be > limited to. > For more details of the capability rights see > .Xr cap_rights_init 3 . > +The > +.Fa operations > +argument limits the operations that are available using > +.Nm system.fileargs . > +.Fa operations > +is a combination of: > +.Bl -ohang -offset indent > +.It FA_OPEN > +Allow > +.Fn fileargs_open > +and > +.Fn fileargs_fopen . > +.It FA_LSTAT > +Allow > +.Fn fileargs_lstat . > +.El > .Pp > The function > .Fn fileargs_cinit > @@ -126,6 +145,11 @@ The function handle > .Dv NULL > argument. > .Pp > +The function > +.Fn fileargs_lstat > +is equivalent to > +.Xr lstat 2 . > +.Pp > The functions > .Fn fileargs_open > and > @@ -165,6 +189,15 @@ must contain the > The > .Va mode > argument tells which what mode file should be created. > +.It operations (NV_TYPE_NUMBER) > +The > +.Va operations > +limits the usable operations for > +.Fa system.fileargs . > +The possible values are explained as > +.Va operations > +argument with > +.Fn fileargs_init . > .El > .Pp > The > @@ -201,7 +234,7 @@ argv += optind; > > /* Create capability to the system.fileargs service. */ > fa = fileargs_init(argc, argv, O_RDONLY, 0, > -cap_rights_init(&rights, CAP_READ)); > +cap_rights_init(&rights, CAP_READ), FA_OPEN); > if (fa == NULL) > err(1, "unable to open system.fileargs service"); > > @@ -222,6 +255,7 @@ fileargs_free(fa); > .Ed > .Sh SEE ALSO > .Xr cap_enter 2 , > +.Xr lstat 2 , > .Xr open 2 , > .Xr cap_rights_init 3 , > .Xr err 3 , > > Modified: head/lib/libca
Re: svn commit: r345804 - head/usr.bin/systat
Hi, On Tue, 2 Apr 2019 07:08:17 -0700 (PDT) "Rodney W. Grimes" wrote: > > Author: mr > > Date: Tue Apr 2 14:01:03 2019 > > New Revision: 345804 > > URL: https://svnweb.freebsd.org/changeset/base/345804 > > > > Log: > > systat -zarc to display disk activities like -vm > > > > PR: 213310 > > Submitted by: ota > > MFH: 4 weeks > > ? MFC: > > > Differential Revision:https://reviews.freebsd.org/D18726 > > > > Modified: > > head/usr.bin/systat/devs.c > > head/usr.bin/systat/devs.h > > head/usr.bin/systat/iostat.c > > head/usr.bin/systat/swap.c > > head/usr.bin/systat/systat.h > > head/usr.bin/systat/vmstat.c > > head/usr.bin/systat/zarc.c > > > > Modified: head/usr.bin/systat/devs.c > > == > > --- head/usr.bin/systat/devs.c Tue Apr 2 13:59:04 2019 > > (r345803) > > +++ head/usr.bin/systat/devs.c Tue Apr 2 14:01:03 2019 > > (r345804) > > @@ -2,6 +2,7 @@ > > * SPDX-License-Identifier: BSD-3-Clause > > * > > * Copyright (c) 1998 Kenneth D. Merry. > > + * 2015 Yoshihiro Ota > > * All rights reserved. > > > Can we get in contact with Yoshihiro Ota about his > copyright statements, and correcting this to make > it clear that it is Kenneth D. Merry that is asserting > "All rights reserved" and if Ota does nor does not wish to assert > "All rights reserved". > > As committed this makes a grey area on Kenneth's assertion, > also leaving out the word copyright on Yoshihiro's line is a bit iffy. > > I am only commenting once, this issue appears several times. > We can go back out to D18726 to discuss it if need be. I've fully written zarc.c (and copied copy-right section from another file) but all other changes in other files are refactoring and adjustments. I don't know the policy/procedure of how to update copy right section of FreeBSD. Please adjust them according the project. Regards, Hiro ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r342375 - head/sys/dev/md
Thanks.
I noticed when this was broken upon a new release long time ago but couldn't
get any clue nor when this was broken (since I hadn't started following
trunk/stable back then.)
Was that 7.0-RELEASE or 8.0-RELEASE?
Anyway, good news and I will try to bring this to local 12.0-RELEASE.
Regards,
Hiro
On Sat, 22 Dec 2018 22:59:11 + (UTC)
Bruce Evans wrote:
> Author: bde
> Date: Sat Dec 22 22:59:11 2018
> New Revision: 342375
> URL: https://svnweb.freebsd.org/changeset/base/342375
>
> Log:
> Fix devstat on md devices, second attempt. r341765 depends on
> g_io_deliver() finishing initialization of the bio, but g_io_deliver()
> actually destroys the bio. INVARIANTS makes the bug obvious by
> overwriting the bio with garbage.
>
> Restore the old order for calling devstat (except don't restore not calling
> it for the error case), and translate to the devstat KPI so that this order
> works.
>
> Reviewed by:kib
>
> Modified:
> head/sys/dev/md/md.c
>
> Modified: head/sys/dev/md/md.c
> ==
> --- head/sys/dev/md/md.c Sat Dec 22 21:49:25 2018(r342374)
> +++ head/sys/dev/md/md.c Sat Dec 22 22:59:11 2018(r342375)
> @@ -1241,12 +1241,22 @@ md_kthread(void *arg)
> error = sc->start(sc, bp);
> }
>
> + if (bp->bio_cmd == BIO_READ || bp->bio_cmd == BIO_WRITE) {
> + /*
> + * Devstat uses (bio_bcount, bio_resid) for
> + * determining the length of the completed part of
> + * the i/o. g_io_deliver() will translate from
> + * bio_completed to that, but it also destroys the
> + * bio so we must do our own translation.
> + */
> + bp->bio_bcount = bp->bio_length;
> + bp->bio_resid = (error == -1 ? bp->bio_bcount : 0);
> + devstat_end_transaction_bio(sc->devstat, bp);
> + }
> if (error != -1) {
> bp->bio_completed = bp->bio_length;
> g_io_deliver(bp, error);
> }
> - if (bp->bio_cmd == BIO_READ || bp->bio_cmd == BIO_WRITE)
> - devstat_end_transaction_bio(sc->devstat, bp);
> }
> }
>
> ___
> svn-src-all@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/svn-src-all
> To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r202708 - head/sys/fs/tmpfs
On Wed, 20 Jan 2010 18:50:58 +0100 "Ronald Klop" wrote: > On Wed, 20 Jan 2010 18:34:37 +0100, Ivan Voras wrote: > > > 2010/1/20 Jaakko Heinonen : > >> Author: jh > >> Date: Wed Jan 20 16:56:20 2010 > >> New Revision: 202708 > >> URL: http://svn.freebsd.org/changeset/base/202708 > > > >> Modified: > >> head/sys/fs/tmpfs/tmpfs_subr.c > >> head/sys/fs/tmpfs/tmpfs_vfsops.c > > > > Any chance of downgrading the "considered highly experimental" status > > message of tmpfs to something less scary? Like "will probably not set > > your hamster on fire"? :) > > > > I admit I haven't used it for any really demanding work but it looks > > ok enough for /tmp on some busy systems. > > And since this list is shrinking. > http://www.freebsd.org/cgi/query-pr-summary.cgi?category=&severity=&priority=&class=&state=&sort=none&text=tmpfs&responsible=&multitext=&originator=&release= > > Ronald. I've been using it for all ports build since tmpfs appeared on FreeBSD. Most of times, it is okay although I had to fix a couple of bumps. There seems to be some corner cases as I see PRs come in. Speaking of bugs, tmpfs may have problems if you do "mv a a" where both source and target are the same; file may disappear due to some bug with rename.2. I tested with Linux and Solaris and "mv" prevented problems with some of these archs. But I forgot most of results and didn't have time since then. Hiro ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r197850 - head/sys/fs/tmpfs
Could someone explain what was the cause and why this fixes the issue?
When I read it, it looked like that a vm object could be in 2 locations
and one of them looked the cause of the problem.
But I couldn't figure out thereafter.
Thanks,
Hiro
On Wed, 7 Oct 2009 23:17:15 + (UTC)
Xin LI wrote:
> Author: delphij
> Date: Wed Oct 7 23:17:15 2009
> New Revision: 197850
> URL: http://svn.freebsd.org/changeset/base/197850
>
> Log:
> Add a special workaround to handle UIO_NOCOPY case. This fixes data
> corruption observed when sendfile() is being used.
>
> PR: kern/127213
> Submitted by: gk
> MFC after: 2 weeks
>
> Modified:
> head/sys/fs/tmpfs/tmpfs_vnops.c
>
> Modified: head/sys/fs/tmpfs/tmpfs_vnops.c
> ==
> --- head/sys/fs/tmpfs/tmpfs_vnops.c Wed Oct 7 23:01:31 2009
> (r197849)
> +++ head/sys/fs/tmpfs/tmpfs_vnops.c Wed Oct 7 23:17:15 2009
> (r197850)
> @@ -43,6 +43,8 @@ __FBSDID("$FreeBSD$");
> #include
> #include
> #include
> +#include
> +#include
> #include
> #include
> #include
> @@ -428,15 +430,72 @@ tmpfs_setattr(struct vop_setattr_args *v
> }
>
> /* - */
> +static int
> +tmpfs_nocacheread(vm_object_t tobj, vm_pindex_t idx,
> +vm_offset_t offset, size_t tlen, struct uio *uio)
> +{
> + vm_page_t m;
> + int error;
> +
> + VM_OBJECT_LOCK(tobj);
> + vm_object_pip_add(tobj, 1);
> + m = vm_page_grab(tobj, idx, VM_ALLOC_WIRED |
> + VM_ALLOC_ZERO | VM_ALLOC_NORMAL | VM_ALLOC_RETRY);
> + if (m->valid != VM_PAGE_BITS_ALL) {
> + if (vm_pager_has_page(tobj, idx, NULL, NULL)) {
> + error = vm_pager_get_pages(tobj, &m, 1, 0);
> + if (error != 0) {
> + printf("tmpfs get pages from pager error
> [read]\n");
> + goto out;
> + }
> + } else
> + vm_page_zero_invalid(m, TRUE);
> + }
> + VM_OBJECT_UNLOCK(tobj);
> + error = uiomove_fromphys(&m, offset, tlen, uio);
> + VM_OBJECT_LOCK(tobj);
> +out:
> + vm_page_lock_queues();
> + vm_page_unwire(m, TRUE);
> + vm_page_unlock_queues();
> + vm_page_wakeup(m);
> + vm_object_pip_subtract(tobj, 1);
> + VM_OBJECT_UNLOCK(tobj);
> +
> + return (error);
> +}
> +
> +static __inline int
> +tmpfs_nocacheread_buf(vm_object_t tobj, vm_pindex_t idx,
> +vm_offset_t offset, size_t tlen, void *buf)
> +{
> + struct uio uio;
> + struct iovec iov;
> +
> + uio.uio_iovcnt = 1;
> + uio.uio_iov = &iov;
> + iov.iov_base = buf;
> + iov.iov_len = tlen;
> +
> + uio.uio_offset = 0;
> + uio.uio_resid = tlen;
> + uio.uio_rw = UIO_READ;
> + uio.uio_segflg = UIO_SYSSPACE;
> + uio.uio_td = curthread;
> +
> + return (tmpfs_nocacheread(tobj, idx, offset, tlen, &uio));
> +}
>
> static int
> tmpfs_mappedread(vm_object_t vobj, vm_object_t tobj, size_t len, struct uio
> *uio)
> {
> + struct sf_buf *sf;
> vm_pindex_t idx;
> vm_page_t m;
> vm_offset_t offset;
> off_t addr;
> size_t tlen;
> + char*ma;
> int error;
>
> addr = uio->uio_offset;
> @@ -461,33 +520,30 @@ lookupvpg:
> vm_page_wakeup(m);
> VM_OBJECT_UNLOCK(vobj);
> return (error);
> + } else if (m != NULL && uio->uio_segflg == UIO_NOCOPY) {
> + if (vm_page_sleep_if_busy(m, FALSE, "tmfsmr"))
> + goto lookupvpg;
> + vm_page_busy(m);
> + VM_OBJECT_UNLOCK(vobj);
> + sched_pin();
> + sf = sf_buf_alloc(m, SFB_CPUPRIVATE);
> + ma = (char *)sf_buf_kva(sf);
> + error = tmpfs_nocacheread_buf(tobj, idx, offset, tlen,
> + ma + offset);
> + if (error == 0) {
> + uio->uio_offset += tlen;
> + uio->uio_resid -= tlen;
> + }
> + sf_buf_free(sf);
> + sched_unpin();
> + VM_OBJECT_LOCK(vobj);
> + vm_page_wakeup(m);
> + VM_OBJECT_UNLOCK(vobj);
> + return (error);
> }
> VM_OBJECT_UNLOCK(vobj);
> nocache:
> - VM_OBJECT_LOCK(tobj);
> - vm_object_pip_add(tobj, 1);
> - m = vm_page_grab(tobj, idx, VM_ALLOC_WIRED |
> - VM_ALLOC_ZERO | VM_ALLOC_NORMAL | VM_ALLOC_RETRY);
> - if (m->valid != VM_PAGE_BITS_ALL) {
> - if (vm_pager_has_page(tobj, idx, NULL, NULL)) {
> - error = vm_pager_get_pages(tobj, &m, 1, 0);
> - if (error != 0) {
> - printf("tmpfs get pages from pager error
> [read]\n");
> - goto out;
> -
Re: svn commit: r189625 - head/sys/geom/eli
On Tue, 10 Mar 2009 19:01:57 +0100 Guido van Rooij wrote: > On Tue, Mar 10, 2009 at 05:19:48PM +0100, Fabian Keil wrote: > > > Log: > > > When attaching a geli on boot make sure that it is detached > > > upon last close. (needed for a gmirror to properly shutdown > > > upon reboot when a geli is on top the gmirror) > > > > > > > Detach-on-last-close is known to cause panics when > > scrubbing at least some ZFS pools: > > http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/117158 > > > > Quoting the PR: > > |Quoting Pawel Jakub Dawidek's response to my initial report: > > ||GELI's detach-on-last-close mechanism is a general purpose mechanism, it > > ||may not work correctly with ZFS, because ZFS sometimes closes and reopen > > ||providers, which will make GELI to detach. In other words you shouldn't > > ||configure detach-on-last-close for ZFS components. > > > > Grr. So we should make it tuneable. How about being able to set > this flag with the geli command, like the G_ELI_FLAG_BOOT flag. > > -Guido Can we move its implementation into the GEOM framwork such that all of GEOM classes can have detach-on-last-close? It will need some kind of custom hook such that each class can do something extra, i.e. gjournal sync. I often need to stop providers to securely detach USB disks. Hiro ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
Re: svn commit: r187130 - head/sbin/mount
While you are working on mount, can you take a look at this patch? http://lists.freebsd.org/pipermail/freebsd-current/2007-February/068980.html The problem is that all mount_XXX programs that are not in the base system fail; i.e. everything from the ports or all mount_XXX programs anyone creates out side of the base system. I am not sure what is happening with nmount(2) and mount(1) programs or if these file-system list are temporary or permanent. It will be nice if someone can give me a pointer or short summary on the plan for nmount(). Thanks, Hiro On Tue, 13 Jan 2009 06:08:37 + (UTC) "David E. O'Brien" wrote: > Author: obrien > Date: Tue Jan 13 06:08:37 2009 > New Revision: 187130 > URL: http://svn.freebsd.org/changeset/base/187130 > > Log: > r187093 failed to keep the lifetime of the pointer suitable for reentrancy. > Fix that. Also move the current buffer size into the 'cpa' structure. > > Modified: > head/sbin/mount/mount.c ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
