Re: svn commit: r289085 - head/share/mk
On Fri, Oct 9, 2015 at 9:57 PM, Baptiste Daroussin wrote: > Author: bapt > Date: Fri Oct 9 21:57:42 2015 > New Revision: 289085 > URL: https://svnweb.freebsd.org/changeset/base/289085 > > Log: > Add a new bsd.confs.mk similar to bsd.files.mk or bsd.incs.mk > > It defines a CONFS variable for all files supposed to be installed as a > configuration file and handle as such Hi, This breaks lots of ports: - bsd.confs.mk used but not installed - ETCDIR defined to /etc in bsd.own.mk Cheers, Antoine (with hat: portmgr) ___ 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: r289109 - head/sys/dev/drm2/i915
Author: dumbbell
Date: Sat Oct 10 07:43:02 2015
New Revision: 289109
URL: https://svnweb.freebsd.org/changeset/base/289109
Log:
drm/i915: Remove "Attempting to unbind pinned buffer" message
This error message is removed in later versions of Linux and currently,
it spams users.
PR: 200712
MFC after:1 week
Modified:
head/sys/dev/drm2/i915/i915_gem.c
Modified: head/sys/dev/drm2/i915/i915_gem.c
==
--- head/sys/dev/drm2/i915/i915_gem.c Sat Oct 10 06:43:44 2015
(r289108)
+++ head/sys/dev/drm2/i915/i915_gem.c Sat Oct 10 07:43:02 2015
(r289109)
@@ -2527,10 +2527,8 @@ i915_gem_object_unbind(struct drm_i915_g
if (obj->gtt_space == NULL)
return 0;
- if (obj->pin_count) {
- DRM_ERROR("Attempting to unbind pinned buffer\n");
+ if (obj->pin_count)
return -EINVAL;
- }
ret = i915_gem_object_finish_gpu(obj);
if (ret == -ERESTARTSYS || ret == -EINTR)
___
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: r289110 - in head: sbin/geom/class/nop sys/geom/nop
Author: trasz
Date: Sat Oct 10 09:03:31 2015
New Revision: 289110
URL: https://svnweb.freebsd.org/changeset/base/289110
Log:
Make geom_nop(4) collect statistics on all types of BIOs, not just
reads and writes.
PR: kern/198405
Submitted by: Matthew D. Fuller
MFC after:1 month
Sponsored by: The FreeBSD Foundation
Differential Revision:https://reviews.freebsd.org/D3679
Modified:
head/sbin/geom/class/nop/gnop.8
head/sys/geom/nop/g_nop.c
head/sys/geom/nop/g_nop.h
Modified: head/sbin/geom/class/nop/gnop.8
==
--- head/sbin/geom/class/nop/gnop.8 Sat Oct 10 07:43:02 2015
(r289109)
+++ head/sbin/geom/class/nop/gnop.8 Sat Oct 10 09:03:31 2015
(r289110)
@@ -24,7 +24,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd September 15, 2015
+.Dd October 10, 2015
.Dt GNOP 8
.Os
.Sh NAME
@@ -72,9 +72,10 @@ The
utility is used for setting up transparent providers on existing ones.
Its main purpose is testing other GEOM classes, as it allows forced provider
removal and I/O error simulation with a given probability.
-It also gathers the following statistics: number of read requests, number of
-write requests, number of bytes read and number of bytes written.
-In addition, it can be used as a good starting point for implementing new GEOM
+It also gathers statistics on the number of read, write, delete,
+getattr, flush, and other requests, and the number of bytes read and written.
+.Nm
+can also be used as a good starting point for implementing new GEOM
classes.
.Pp
The first argument to
Modified: head/sys/geom/nop/g_nop.c
==
--- head/sys/geom/nop/g_nop.c Sat Oct 10 07:43:02 2015(r289109)
+++ head/sys/geom/nop/g_nop.c Sat Oct 10 09:03:31 2015(r289110)
@@ -119,6 +119,24 @@ g_nop_start(struct bio *bp)
sc->sc_wrotebytes += bp->bio_length;
failprob = sc->sc_wfailprob;
break;
+ case BIO_DELETE:
+ sc->sc_deletes++;
+ break;
+ case BIO_GETATTR:
+ sc->sc_getattrs++;
+ break;
+ case BIO_FLUSH:
+ sc->sc_flushes++;
+ break;
+ case BIO_CMD0:
+ sc->sc_cmd0s++;
+ break;
+ case BIO_CMD1:
+ sc->sc_cmd1s++;
+ break;
+ case BIO_CMD2:
+ sc->sc_cmd2s++;
+ break;
}
mtx_unlock(&sc->sc_lock);
if (failprob > 0) {
@@ -238,6 +256,12 @@ g_nop_create(struct gctl_req *req, struc
sc->sc_wfailprob = wfailprob;
sc->sc_reads = 0;
sc->sc_writes = 0;
+ sc->sc_deletes = 0;
+ sc->sc_getattrs = 0;
+ sc->sc_flushes = 0;
+ sc->sc_cmd0s = 0;
+ sc->sc_cmd1s = 0;
+ sc->sc_cmd2s = 0;
sc->sc_readbytes = 0;
sc->sc_wrotebytes = 0;
mtx_init(&sc->sc_lock, "gnop lock", NULL, MTX_DEF);
@@ -602,6 +626,12 @@ g_nop_ctl_reset(struct gctl_req *req, st
sc = pp->geom->softc;
sc->sc_reads = 0;
sc->sc_writes = 0;
+ sc->sc_deletes = 0;
+ sc->sc_getattrs = 0;
+ sc->sc_flushes = 0;
+ sc->sc_cmd0s = 0;
+ sc->sc_cmd1s = 0;
+ sc->sc_cmd2s = 0;
sc->sc_readbytes = 0;
sc->sc_wrotebytes = 0;
}
@@ -659,6 +689,12 @@ g_nop_dumpconf(struct sbuf *sb, const ch
sbuf_printf(sb, "%s%d\n", indent, sc->sc_error);
sbuf_printf(sb, "%s%ju\n", indent, sc->sc_reads);
sbuf_printf(sb, "%s%ju\n", indent, sc->sc_writes);
+ sbuf_printf(sb, "%s%ju\n", indent, sc->sc_deletes);
+ sbuf_printf(sb, "%s%ju\n", indent,
sc->sc_getattrs);
+ sbuf_printf(sb, "%s%ju\n", indent, sc->sc_flushes);
+ sbuf_printf(sb, "%s%ju\n", indent, sc->sc_cmd0s);
+ sbuf_printf(sb, "%s%ju\n", indent, sc->sc_cmd1s);
+ sbuf_printf(sb, "%s%ju\n", indent, sc->sc_cmd2s);
sbuf_printf(sb, "%s%ju\n", indent,
sc->sc_readbytes);
sbuf_printf(sb, "%s%ju\n", indent,
Modified: head/sys/geom/nop/g_nop.h
==
--- head/sys/geom/nop/g_nop.h Sat Oct 10 07:43:02 2015(r289109)
+++ head/sys/geom/nop/g_nop.h Sat Oct 10 09:03:31 2015(r289110)
@@ -65,6 +65,12 @@ struct g_nop_softc {
u_int sc_wfailprob;
uintmax_t sc_reads;
uintmax_t sc_writes;
+ uintmax_t sc_deletes;
+ uintmax_t sc_getattrs;
+ uintmax_t sc_flushes;
+ uintmax_t sc_cmd0s;
+ uintmax_t sc_cmd1s;
+ uintmax_t sc_cmd2s;
uintmax_t sc_readbytes;
uintmax_t sc_wrotebytes;
struct mtx sc_lock;
_
svn commit: r289111 - head/sys/kern
Author: trasz
Date: Sat Oct 10 09:21:55 2015
New Revision: 289111
URL: https://svnweb.freebsd.org/changeset/base/289111
Log:
Provide better debug message on kernel module name clash.
Reviewed by: kib@
MFC after:1 month
Sponsored by: The FreeBSD Foundation
Modified:
head/sys/kern/kern_module.c
Modified: head/sys/kern/kern_module.c
==
--- head/sys/kern/kern_module.c Sat Oct 10 09:03:31 2015(r289110)
+++ head/sys/kern/kern_module.c Sat Oct 10 09:21:55 2015(r289111)
@@ -158,16 +158,12 @@ module_register(const moduledata_t *data
newmod = module_lookupbyname(data->name);
if (newmod != NULL) {
MOD_XUNLOCK;
- printf("module_register: module %s already exists!\n",
- data->name);
+ printf("%s: cannot register %s from %s; already loaded from
%s\n",
+ __func__, data->name, container->filename,
newmod->file->filename);
return (EEXIST);
}
namelen = strlen(data->name) + 1;
newmod = malloc(sizeof(struct module) + namelen, M_MODULE, M_WAITOK);
- if (newmod == NULL) {
- MOD_XUNLOCK;
- return (ENOMEM);
- }
newmod->refs = 1;
newmod->id = nextid++;
newmod->name = (char *)(newmod + 1);
___
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: r289112 - in head: lib/libc/sys sys/kern
Author: trasz
Date: Sat Oct 10 09:29:47 2015
New Revision: 289112
URL: https://svnweb.freebsd.org/changeset/base/289112
Log:
Change the default setting of kern.ipc.shm_allow_removed from 0 to 1.
This removes the need for manually changing this flag for Google Chrome
users. It also improves compatibility with Linux applications running under
Linuxulator compatibility layer, and possibly also helps in porting software
from Linux.
Generally speaking, the flag allows applications to create the shared memory
segment, attach it, remove it, and then continue to use it and to reattach it
later. This means that the kernel will automatically "clean up" after the
application exits.
It could be argued that it's against POSIX. However, SUSv3 says this
about IPC_RMID: "Remove the shared memory identifier specified by shmid from
the system and destroy the shared memory segment and shmid_ds data structure
associated with it." From my reading, we break it in any case by deferring
removal of the segment until it's detached; we won't break it any more
by also deferring removal of the identifier.
This is the behaviour exhibited by Linux since... probably always, and
also by OpenBSD since the following commit:
revision 1.54
date: 2011/10/27 07:56:28; author: robert; state: Exp; lines: +3 -8;
Allow segments to be used even after they were marked for deletion with
the IPC_RMID flag.
This is permitted as an extension beyond the standards and this is similar
to what other operating systems like linux do.
MFC after:1 month
Relnotes: yes
Sponsored by: The FreeBSD Foundation
Differential Revision:https://reviews.freebsd.org/D3603
Modified:
head/lib/libc/sys/shmctl.2
head/sys/kern/sysv_shm.c
Modified: head/lib/libc/sys/shmctl.2
==
--- head/lib/libc/sys/shmctl.2 Sat Oct 10 09:21:55 2015(r289111)
+++ head/lib/libc/sys/shmctl.2 Sat Oct 10 09:29:47 2015(r289112)
@@ -25,7 +25,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd July 17, 1995
+.Dd October 10, 2015
.Dt SHMCTL 2
.Os
.Sh NAME
@@ -72,15 +72,18 @@ or it must have superuser privileges.
.It Dv IPC_RMID
Removes the segment from the system.
The removal will not take
-effect until all processes having attached the segment have exited;
-however, once the IPC_RMID operation has taken place, no further
-processes will be allowed to attach the segment.
+effect until all processes having attached the segment have exited.
For the operation
to succeed, the calling process's effective uid must match
.Fa shm_perm.uid
or
.Fa shm_perm.cuid ,
or the process must have superuser privileges.
+If the
+.Va kern.ipc.shm_allow_removed
+.Xr sysctl 3
+variable is set to 0, once the IPC_RMID operation has taken place,
+no further processes will be allowed to attach the segment.
.\" .It Dv SHM_LOCK
.\" Locks the segment in memory. The calling process must have
.\" superuser privileges. Not implemented in FreeBSD.
Modified: head/sys/kern/sysv_shm.c
==
--- head/sys/kern/sysv_shm.cSat Oct 10 09:21:55 2015(r289111)
+++ head/sys/kern/sysv_shm.cSat Oct 10 09:29:47 2015(r289112)
@@ -162,7 +162,7 @@ struct shminfo shminfo = {
};
static int shm_use_phys;
-static int shm_allow_removed;
+static int shm_allow_removed = 1;
SYSCTL_ULONG(_kern_ipc, OID_AUTO, shmmax, CTLFLAG_RWTUN, &shminfo.shmmax, 0,
"Maximum shared memory segment size");
___
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: r289109 - head/sys/dev/drm2/i915
On Sat, Oct 10, 2015 at 07:43:02AM +, Jean-S'ebastien P'edron wrote: > Author: dumbbell > Date: Sat Oct 10 07:43:02 2015 > New Revision: 289109 > URL: https://svnweb.freebsd.org/changeset/base/289109 > > Log: > drm/i915: Remove "Attempting to unbind pinned buffer" message > > This error message is removed in later versions of Linux and currently, > it spams users. > > PR: 200712 > MFC after: 1 week What about don't worked video playback? https://lists.freebsd.org/pipermail/freebsd-current/2015-August/057326.html ___ 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: r289113 - head/sbin/mdconfig
Author: trasz
Date: Sat Oct 10 09:37:31 2015
New Revision: 289113
URL: https://svnweb.freebsd.org/changeset/base/289113
Log:
Add 'p' postfix to mdconfig(8).
MFC after:1 month
Sponsored by: The FreeBSD Foundation
Modified:
head/sbin/mdconfig/mdconfig.8
head/sbin/mdconfig/mdconfig.c
Modified: head/sbin/mdconfig/mdconfig.8
==
--- head/sbin/mdconfig/mdconfig.8 Sat Oct 10 09:29:47 2015
(r289112)
+++ head/sbin/mdconfig/mdconfig.8 Sat Oct 10 09:37:31 2015
(r289113)
@@ -37,7 +37,7 @@
.\"
.\" $FreeBSD$
.\"
-.Dd August 6, 2015
+.Dd October 10, 2015
.Dt MDCONFIG 8
.Os
.Sh NAME
@@ -159,11 +159,11 @@ prefix.
Size of the memory disk.
.Ar Size
is the number of 512 byte sectors unless suffixed with a
-.Cm b , k , m , g ,
+.Cm b , k , m , g , t ,
or
-.Cm t
+.Cm p
which
-denotes byte, kilobyte, megabyte, gigabyte and terabyte respectively.
+denotes byte, kilobyte, megabyte, gigabyte, terabyte and petabyte respectively.
When used without the
.Fl r
option, the
Modified: head/sbin/mdconfig/mdconfig.c
==
--- head/sbin/mdconfig/mdconfig.c Sat Oct 10 09:29:47 2015
(r289112)
+++ head/sbin/mdconfig/mdconfig.c Sat Oct 10 09:37:31 2015
(r289113)
@@ -88,8 +88,8 @@ usage(void)
fprintf(stderr, "\t\ttype = {malloc, vnode, swap}\n");
fprintf(stderr, "\t\toption = {cluster, compress, reserve}\n");
fprintf(stderr, "\t\tsize = %%d (512 byte blocks), %%db (B),\n");
- fprintf(stderr, "\t\t %%dk (kB), %%dm (MB), %%dg (GB) or\n");
- fprintf(stderr, "\t\t %%dt (TB)\n");
+ fprintf(stderr, "\t\t %%dk (kB), %%dm (MB), %%dg (GB), \n");
+ fprintf(stderr, "\t\t %%dt (TB), or %%dp (PB)\n");
exit(1);
}
@@ -217,6 +217,9 @@ main(int argc, char **argv)
else if (*p == 't' || *p == 'T') {
mdio.md_mediasize <<= 30;
mdio.md_mediasize <<= 10;
+ } else if (*p == 'p' || *p == 'P') {
+ mdio.md_mediasize <<= 30;
+ mdio.md_mediasize <<= 20;
} else
errx(1, "unknown suffix on -s argument");
break;
___
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: r289109 - head/sys/dev/drm2/i915
On 10.10.2015 11:35, Slawa Olhovchenkov wrote: > On Sat, Oct 10, 2015 at 07:43:02AM +, Jean-S'ebastien P'edron wrote: > What about don't worked video playback? > https://lists.freebsd.org/pipermail/freebsd-current/2015-August/057326.html Hi! Hopefully it will be fixed by the coming i915 update. Please see the following wiki article for status update and instructions if you want to test it: https://wiki.freebsd.org/Graphics/Update%20i915%20GPU%20driver%20to%20Linux%203.8 -- Jean-Sébastien Pédron signature.asc Description: OpenPGP digital signature
Re: svn commit: r289109 - head/sys/dev/drm2/i915
On Sat, Oct 10, 2015 at 12:14:50PM +0200, Jean-S'ebastien P'edron wrote: > On 10.10.2015 11:35, Slawa Olhovchenkov wrote: > > On Sat, Oct 10, 2015 at 07:43:02AM +, Jean-S'ebastien P'edron wrote: > > What about don't worked video playback? > > https://lists.freebsd.org/pipermail/freebsd-current/2015-August/057326.html > > Hi! > > Hopefully it will be fixed by the coming i915 update. Please see the > following wiki article for status update and instructions if you want to > test it: > https://wiki.freebsd.org/Graphics/Update%20i915%20GPU%20driver%20to%20Linux%203.8 Can I try this on 10.2 system? w/o recompile base and X11? ___ 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: r289114 - in head: libexec/dma/dmagent share/mk usr.sbin/ppp
Author: bapt
Date: Sat Oct 10 10:48:09 2015
New Revision: 289114
URL: https://svnweb.freebsd.org/changeset/base/289114
Log:
Install bsd.confs.mk
Rename ETCDIR into CONFIGDIR to avoid conflicts with the ports tree
Modified:
head/libexec/dma/dmagent/Makefile
head/share/mk/Makefile
head/share/mk/bsd.confs.mk
head/share/mk/bsd.own.mk
head/usr.sbin/ppp/Makefile
Modified: head/libexec/dma/dmagent/Makefile
==
--- head/libexec/dma/dmagent/Makefile Sat Oct 10 09:37:31 2015
(r289113)
+++ head/libexec/dma/dmagent/Makefile Sat Oct 10 10:48:09 2015
(r289114)
@@ -17,7 +17,7 @@ SRCS= aliases_parse.y \
util.c
MAN8= dma.8
CONFS= dma.conf
-CONFSDIR= ${ETCDIR}/dma
+CONFSDIR= ${CONFIGDIR}/dma
YFLAGS+= -i
CLEANFILES=aliases_parse.i
FILES= mailer.conf
Modified: head/share/mk/Makefile
==
--- head/share/mk/Makefile Sat Oct 10 09:37:31 2015(r289113)
+++ head/share/mk/Makefile Sat Oct 10 10:48:09 2015(r289114)
@@ -13,6 +13,7 @@ FILES=\
bsd.README \
bsd.arch.inc.mk \
bsd.compiler.mk \
+ bsd.confs.mk \
bsd.cpu.mk \
bsd.crunchgen.mk \
bsd.dep.mk \
Modified: head/share/mk/bsd.confs.mk
==
--- head/share/mk/bsd.confs.mk Sat Oct 10 09:37:31 2015(r289113)
+++ head/share/mk/bsd.confs.mk Sat Oct 10 10:48:09 2015(r289114)
@@ -22,7 +22,7 @@ all: buildconfig
${group}OWN?= ${SHAREOWN}
${group}GRP?= ${SHAREGRP}
${group}MODE?= ${CONFMODE}
-${group}DIR?= ${ETCDIR}/
+${group}DIR?= ${CONFIGDIR}/
STAGE_SETS+= ${group}
STAGE_DIR.${group}= ${STAGE_OBJTOP}${${group}DIR}
STAGE_SYMLINKS_DIR.${group}= ${STAGE_OBJTOP}
Modified: head/share/mk/bsd.own.mk
==
--- head/share/mk/bsd.own.mkSat Oct 10 09:37:31 2015(r289113)
+++ head/share/mk/bsd.own.mkSat Oct 10 10:48:09 2015(r289114)
@@ -213,7 +213,7 @@ NLSGRP?=${SHAREGRP}
NLSMODE?= ${NOBINMODE}
INCLUDEDIR?= /usr/include
-ETCDIR?= /etc
+CONFIGDIR?=/etc
#
# install(1) parameters.
Modified: head/usr.sbin/ppp/Makefile
==
--- head/usr.sbin/ppp/Makefile Sat Oct 10 09:37:31 2015(r289113)
+++ head/usr.sbin/ppp/Makefile Sat Oct 10 10:48:09 2015(r289114)
@@ -22,7 +22,7 @@ PPP_NO_RADIUS=
PPP_NO_SUID=
.endif
CONFS= ppp.conf
-CONFSDIR= ${ETCDIR}/ppp
+CONFSDIR= ${CONFIGDIR}/ppp
CONFSMODE= 600
.if ${MK_ATM} == "no"
___
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: r289072 - in head/contrib: libc++/src llvm/lib/Transforms/Vectorize
On 10 October 2015 at 06:08, Konstantin Belousov wrote: > On Fri, Oct 09, 2015 at 06:21:45PM +, Dimitry Andric wrote: >> Author: dim >> Date: Fri Oct 9 18:21:45 2015 >> New Revision: 289072 >> URL: https://svnweb.freebsd.org/changeset/base/289072 >> >> Log: >> Temporarily revert upstream llvm trunk r240144 (by Michael Zolotukhin): > > This might be indeed a clang bug, but probably not in the revision which > you reverted. Or it might be a libc++ bug. Yes, the bug is almost certainly not r240144; reverting it is a short-term workaround until the underlying issue is fixed. > The story, from what I understand from the discussion at some other > place, is that the fault occurs on access to the common unaligned > symbol. Apparently, ELF does allow to specify alignment of the common > symbols, and static linker, when finally allocating .bss space for the > object, must obey the requirement. The symbol value for the common > symbol must be interpreted as the desired alignment. > > Could you look up the symbol reference in the .o files (again) and see > which alignment is requested ? I think that it is probably clang which > should set it to be at least 16 bytes to generate ABI-compliant code. The declaration is extern ostream cout; which correctly has 8 byte alignment. In the implementation (contrib/libc++/src/iostream.cpp) it is defined as a char array which is where the 16-byte ABI alignment would arise, but explicitly specifies 8-byte alignment: _ALIGNAS_TYPE (ostream) _LIBCPP_FUNC_VIS char cout[sizeof(ostream)]; Compiling with -emit-llvm --save-temps shows the desired 8-byte alignment for cout in the LLVM IR but 16-byte in the assembly file. ___ 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: r289072 - in head/contrib: libc++/src llvm/lib/Transforms/Vectorize
On Sat, Oct 10, 2015 at 11:37:26AM +, Ed Maste wrote: > On 10 October 2015 at 06:08, Konstantin Belousov wrote: > > On Fri, Oct 09, 2015 at 06:21:45PM +, Dimitry Andric wrote: > >> Author: dim > >> Date: Fri Oct 9 18:21:45 2015 > >> New Revision: 289072 > >> URL: https://svnweb.freebsd.org/changeset/base/289072 > >> > >> Log: > >> Temporarily revert upstream llvm trunk r240144 (by Michael Zolotukhin): > > > > This might be indeed a clang bug, but probably not in the revision which > > you reverted. Or it might be a libc++ bug. > > Yes, the bug is almost certainly not r240144; reverting it is a > short-term workaround until the underlying issue is fixed. > > > The story, from what I understand from the discussion at some other > > place, is that the fault occurs on access to the common unaligned > > symbol. Apparently, ELF does allow to specify alignment of the common > > symbols, and static linker, when finally allocating .bss space for the > > object, must obey the requirement. The symbol value for the common > > symbol must be interpreted as the desired alignment. > > > > Could you look up the symbol reference in the .o files (again) and see > > which alignment is requested ? I think that it is probably clang which > > should set it to be at least 16 bytes to generate ABI-compliant code. > > The declaration is extern ostream cout; which correctly has 8 byte alignment. The alignment of the declaration does not matter, most likely. I am not sure what you mean by saying that 8 byte alignment is desirable, while compiler generates accesses (movdqa) which assume 16-byte alignment. > > In the implementation (contrib/libc++/src/iostream.cpp) it is defined > as a char array which is where the 16-byte ABI alignment would arise, > but explicitly specifies 8-byte alignment: > _ALIGNAS_TYPE (ostream) _LIBCPP_FUNC_VIS char cout[sizeof(ostream)]; > > Compiling with -emit-llvm --save-temps shows the desired 8-byte > alignment for cout in the LLVM IR but 16-byte in the assembly file. Could somebody show the readelf -s iostream.So and readelf -s libc++.so.X output lines for the symbol, please ? ___ 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: r289109 - head/sys/dev/drm2/i915
On 10.10.2015 12:22, Slawa Olhovchenkov wrote: > On Sat, Oct 10, 2015 at 12:14:50PM +0200, Jean-S'ebastien P'edron wrote: >> Hopefully it will be fixed by the coming i915 update. Please see the >> following wiki article for status update and instructions if you want to >> test it: >> https://wiki.freebsd.org/Graphics/Update%20i915%20GPU%20driver%20to%20Linux%203.8 > > Can I try this on 10.2 system? > w/o recompile base and X11? Yes, you can (and should) only build the kernel from that branch, no need to rebuild anything else. That said, I'm not sure a kernel from HEAD will work nicely with userland from 10.2. That's worth a try. -- Jean-Sébastien Pédron signature.asc Description: OpenPGP digital signature
Re: svn commit: r289072 - in head/contrib: libc++/src llvm/lib/Transforms/Vectorize
On 10 October 2015 at 12:03, Konstantin Belousov wrote: >> The declaration is extern ostream cout; which correctly has 8 byte alignment. > The alignment of the declaration does not matter, most likely. > I am not sure what you mean by saying that 8 byte alignment is desirable, > while compiler generates accesses (movdqa) which assume 16-byte alignment. I mean that this object is naturally expected to have 8-byte alignment. In previous versions of libc++.so shipped with FreeBSD it did, and can't be changed now (cout comes from the binary itself, via a copy relocation). >> In the implementation (contrib/libc++/src/iostream.cpp) it is defined >> as a char array which is where the 16-byte ABI alignment would arise, >> but explicitly specifies 8-byte alignment: >> _ALIGNAS_TYPE (ostream) _LIBCPP_FUNC_VIS char cout[sizeof(ostream)]; >> >> Compiling with -emit-llvm --save-temps shows the desired 8-byte >> alignment for cout in the LLVM IR but 16-byte in the assembly file. > Could somebody show the readelf -s iostream.So and readelf -s libc++.so.X > output lines for the symbol, please ? readelf -s iostream.So: 120: 0150 160 OBJECT GLOBAL DEFAULT5 _ZNSt3__14coutE libc++.so.1.full 1280: 002bb4a0 160 OBJECT GLOBAL DEFAULT 25 _ZNSt3__14coutE ___ 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: r289072 - in head/contrib: libc++/src llvm/lib/Transforms/Vectorize
On Sat, Oct 10, 2015 at 12:24:33PM +, Ed Maste wrote: > On 10 October 2015 at 12:03, Konstantin Belousov wrote: > >> The declaration is extern ostream cout; which correctly has 8 byte > >> alignment. > > The alignment of the declaration does not matter, most likely. > > I am not sure what you mean by saying that 8 byte alignment is desirable, > > while compiler generates accesses (movdqa) which assume 16-byte alignment. > > I mean that this object is naturally expected to have 8-byte > alignment. In previous versions of libc++.so shipped with FreeBSD it > did, and can't be changed now (cout comes from the binary itself, via > a copy relocation). No, the natural alignment for char[] is 1, while ABI requires an array longer than 15 bytes to have 16-bytes alignment. > > >> In the implementation (contrib/libc++/src/iostream.cpp) it is defined > >> as a char array which is where the 16-byte ABI alignment would arise, > >> but explicitly specifies 8-byte alignment: > >> _ALIGNAS_TYPE (ostream) _LIBCPP_FUNC_VIS char cout[sizeof(ostream)]; > >> > >> Compiling with -emit-llvm --save-temps shows the desired 8-byte > >> alignment for cout in the LLVM IR but 16-byte in the assembly file. > > Could somebody show the readelf -s iostream.So and readelf -s libc++.so.X > > output lines for the symbol, please ? > > readelf -s iostream.So: >120: 0150 160 OBJECT GLOBAL DEFAULT5 _ZNSt3__14coutE > > libc++.so.1.full > 1280: 002bb4a0 160 OBJECT GLOBAL DEFAULT 25 _ZNSt3__14coutE And there the object has right alignment. ___ 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: r289109 - head/sys/dev/drm2/i915
On Sat, Oct 10, 2015 at 04:37:43PM +0300, Dmitry Morozovsky wrote: > On Sat, 10 Oct 2015, Jean-S?bastien P?dron wrote: > > > That said, I'm not sure a kernel from HEAD will work nicely with > > userland from 10.2. That's worth a try. > > well, it should -- otherwise stable/10 jails would not work under -current. smartctl from 8.x don't work under 10.2. > Not the other way os course. More interesting about X11. ___ 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: r289109 - head/sys/dev/drm2/i915
On Sat, 10 Oct 2015, Jean-S?bastien P?dron wrote: > That said, I'm not sure a kernel from HEAD will work nicely with > userland from 10.2. That's worth a try. well, it should -- otherwise stable/10 jails would not work under -current. Not the other way os course. -- Sincerely, D.Marck [DM5020, MCK-RIPE, DM3-RIPN] [ FreeBSD committer: ma...@freebsd.org ] *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- ma...@rinet.ru *** ___ 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: r289109 - head/sys/dev/drm2/i915
On 10.10.2015 15:37, Dmitry Morozovsky wrote: > On Sat, 10 Oct 2015, Jean-S?bastien P?dron wrote: >> That said, I'm not sure a kernel from HEAD will work nicely with >> userland from 10.2. That's worth a try. > > well, it should -- otherwise stable/10 jails would not work under -current. I was thinking about changes affecting rc.d, such as the disappearance of the wifi device (other than wlan0) in ifconfig (ie. there is only wlan0 in HEAD, not wlan0+iwn0). Perhaps there is no problem with that, I just don't know. -- Jean-Sébastien Pédron signature.asc Description: OpenPGP digital signature
Re: svn commit: r289109 - head/sys/dev/drm2/i915
On Sat, Oct 10, 2015 at 04:29:04PM +0200, Jean-S'ebastien P'edron wrote: > On 10.10.2015 15:37, Dmitry Morozovsky wrote: > > On Sat, 10 Oct 2015, Jean-S?bastien P?dron wrote: > >> That said, I'm not sure a kernel from HEAD will work nicely with > >> userland from 10.2. That's worth a try. > > > > well, it should -- otherwise stable/10 jails would not work under -current. > > I was thinking about changes affecting rc.d, such as the disappearance > of the wifi device (other than wlan0) in ifconfig (ie. there is only > wlan0 in HEAD, not wlan0+iwn0). Perhaps there is no problem with that, I > just don't know. OK, I am need wired connection for notebook. ___ 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: r289115 - head
Author: bapt
Date: Sat Oct 10 14:43:46 2015
New Revision: 289115
URL: https://svnweb.freebsd.org/changeset/base/289115
Log:
Fix installing config files as non root
Reported by: adrian
Modified:
head/Makefile.inc1
Modified: head/Makefile.inc1
==
--- head/Makefile.inc1 Sat Oct 10 10:48:09 2015(r289114)
+++ head/Makefile.inc1 Sat Oct 10 14:43:46 2015(r289115)
@@ -1003,7 +1003,9 @@ distrib-dirs: .MAKE
distribution: .MAKE
cd ${.CURDIR}/etc; ${CROSSENV} PATH=${TMPPATH} ${MAKE} \
${IMAKE_INSTALL} ${IMAKE_MTREE} METALOG=${METALOG} ${.TARGET}
- ${_+_}cd ${.CURDIR}; ${MAKE} -f Makefile.inc1 installconfig
+ ${_+_}cd ${.CURDIR}; ${CROSSENV} PATH=${TMPPATH} \
+ ${MAKE} -f Makefile.inc1 ${IMAKE_INSTALL} \
+ METALOG=${METALOG} installconfig
#
# buildkernel and installkernel
___
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: r289087 - in head: etc etc/dma libexec libexec/dma libexec/dma-mbox-create libexec/dma/dma libexec/dma/dma-mbox-create share/examples share/examples/dma
On Fri, Oct 09, 2015 at 08:15:57PM -0700, Adrian Chadd wrote: > I think this broke being able to do installworld as non-root > > 21:03 <@adrian> ===> libexec/dma/dmagent (installconfig) > 21:03 <@adrian> install: > /home/adrian/work/freebsd/head-embedded/src/../root/mips_ap/etc/dma/dma.conf: > chown/chgrp: Operation not permitted > 21:03 <@adrian> *** Error code 71 > Fixed in r289115 Bapt signature.asc Description: PGP signature
Re: svn commit: r289072 - in head/contrib: libc++/src llvm/lib/Transforms/Vectorize
On Oct 10, 2015 8:53 AM, "Konstantin Belousov" wrote: > > On Sat, Oct 10, 2015 at 12:24:33PM +, Ed Maste wrote: > > On 10 October 2015 at 12:03, Konstantin Belousov wrote: > > >> The declaration is extern ostream cout; which correctly has 8 byte alignment. > > > The alignment of the declaration does not matter, most likely. > > > I am not sure what you mean by saying that 8 byte alignment is desirable, > > > while compiler generates accesses (movdqa) which assume 16-byte alignment. > > > > I mean that this object is naturally expected to have 8-byte > > alignment. In previous versions of libc++.so shipped with FreeBSD it > > did, and can't be changed now (cout comes from the binary itself, via > > a copy relocation). > No, the natural alignment for char[] is 1, while ABI requires an array > longer than 15 bytes to have 16-bytes alignment. Right, but the proper alignment for ostream is 8; it's only because of the slightly bizarre definition as a char array that we end up with this 16 byte ABI alignment. ___ 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: r289116 - head/share/man/man4
Author: phk Date: Sat Oct 10 17:32:06 2015 New Revision: 289116 URL: https://svnweb.freebsd.org/changeset/base/289116 Log: Remove a double "as well as" Modified: head/share/man/man4/vtnet.4 Modified: head/share/man/man4/vtnet.4 == --- head/share/man/man4/vtnet.4 Sat Oct 10 14:43:46 2015(r289115) +++ head/share/man/man4/vtnet.4 Sat Oct 10 17:32:06 2015(r289116) @@ -52,9 +52,9 @@ device driver provides support for VirtI If the hypervisor advertises the appreciate features, the .Nm driver supports TCP/UDP checksum offload for both transmit and receive, -TCP segmentation offload (TSO), TCP large receive offload (LRO), and -hardware VLAN tag stripping/insertion features, as well as a multicast -hash filter, as well as Jumbo Frames (up to 9216 bytes), which can be +TCP segmentation offload (TSO), TCP large receive offload (LRO), +hardware VLAN tag stripping/insertion features, a multicast hash filter, +as well as Jumbo Frames (up to 9216 bytes), which can be configured via the interface MTU setting. Selecting an MTU larger than 1500 bytes with the .Xr ifconfig 8 ___ 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: r289117 - head/sys/mips/conf
Author: sbruno Date: Sat Oct 10 19:08:34 2015 New Revision: 289117 URL: https://svnweb.freebsd.org/changeset/base/289117 Log: There's no way a fbsd install + kernel will fit into 4MB of flash. Assume and enforce the fact that this will always boot a rootfs from usb. Modified: head/sys/mips/conf/TP-MR3020 Modified: head/sys/mips/conf/TP-MR3020 == --- head/sys/mips/conf/TP-MR3020Sat Oct 10 17:32:06 2015 (r289116) +++ head/sys/mips/conf/TP-MR3020Sat Oct 10 19:08:34 2015 (r289117) @@ -50,16 +50,6 @@ device geom_uncompress # Used for the static uboot partition map device geom_map -# Boot off of the rootfs, as defined in the geom_map setup. -# Probably, this should be a USB device as the memory available -# compressed rootfs is simply too small for FreeBSD -#options ROOTDEVNAME=\"ufs:map/rootfs.uncompress\" - -# Boot off of a uboot tftp ramdisk kernel image. Because the flash -# on this unit is so small, this is the only way to do dev work. -# For full deployment, you will *have* to use a usb storage device -# as a rootfs and use the flash to hold the kernel only. -optionsMD_ROOT # md device usable as a potential root device -optionsMD_ROOT_SIZE=10240 -#makeoptions MFS_IMAGE=/tftpboot/mfsroot-tl-mr3020.img.ulzma -optionsROOTDEVNAME=\"ufs:md0.uncompress\" +# With only 4MB of flash, we are stuck using USB +# for the rootfs. +optionsROOTDEVNAME=\"ufs:da0\" ___ 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: r289118 - head/sys/dev/iicbus
Author: ian
Date: Sat Oct 10 19:51:00 2015
New Revision: 289118
URL: https://svnweb.freebsd.org/changeset/base/289118
Log:
Replace a local sx lock that allowed only one client at a time to access
an eeprom device with iicbus_request/release_bus(), which achieves the
same effect and also keeps other i2c slave drivers from clashing on the bus.
Modified:
head/sys/dev/iicbus/icee.c
Modified: head/sys/dev/iicbus/icee.c
==
--- head/sys/dev/iicbus/icee.c Sat Oct 10 19:08:34 2015(r289117)
+++ head/sys/dev/iicbus/icee.c Sat Oct 10 19:51:00 2015(r289118)
@@ -48,7 +48,7 @@ __FBSDID("$FreeBSD$");
struct icee_softc {
device_tsc_dev; /* Myself */
- struct sx sc_lock;/* basically a perimeter lock */
+ device_tsc_busdev; /* Parent bus */
struct cdev *cdev; /* user interface */
int addr;
int size; /* How big am I? */
@@ -57,12 +57,6 @@ struct icee_softc {
int wr_sz; /* What's the write page size */
};
-#define ICEE_LOCK(_sc) sx_xlock(&(_sc)->sc_lock)
-#defineICEE_UNLOCK(_sc)sx_xunlock(&(_sc)->sc_lock)
-#define ICEE_LOCK_INIT(_sc)sx_init(&_sc->sc_lock, "icee")
-#define ICEE_LOCK_DESTROY(_sc) sx_destroy(&_sc->sc_lock);
-#define ICEE_ASSERT_LOCKED(_sc)sx_assert(&_sc->sc_lock, SA_XLOCKED);
-#define ICEE_ASSERT_UNLOCKED(_sc) sx_assert(&_sc->sc_lock, SA_UNLOCKED);
#define CDEV2SOFTC(dev)((dev)->si_drv1)
/* cdev routines */
@@ -97,6 +91,7 @@ icee_attach(device_t dev)
int dunit, err;
sc->sc_dev = dev;
+ sc->sc_busdev = device_get_parent(sc->sc_dev);
sc->addr = iicbus_get_addr(dev);
err = 0;
dname = device_get_name(dev);
@@ -117,7 +112,6 @@ icee_attach(device_t dev)
goto out;
}
sc->cdev->si_drv1 = sc;
- ICEE_LOCK_INIT(sc);
out:
return (err);
}
@@ -155,7 +149,9 @@ icee_read(struct cdev *dev, struct uio *
return (EIO);
if (sc->type != 8 && sc->type != 16)
return (EINVAL);
- ICEE_LOCK(sc);
+ error = iicbus_request_bus(sc->sc_busdev, sc->sc_dev, IIC_INTRWAIT);
+ if (error!= 0)
+ return (iic2errno(error));
slave = error = 0;
while (uio->uio_resid > 0) {
if (uio->uio_offset >= sc->size)
@@ -188,7 +184,7 @@ icee_read(struct cdev *dev, struct uio *
if (error)
break;
}
- ICEE_UNLOCK(sc);
+ iicbus_release_bus(sc->sc_busdev, sc->sc_dev);
return (error);
}
@@ -216,7 +212,10 @@ icee_write(struct cdev *dev, struct uio
return (EIO);
if (sc->type != 8 && sc->type != 16)
return (EINVAL);
- ICEE_LOCK(sc);
+
+ error = iicbus_request_bus(sc->sc_busdev, sc->sc_dev, IIC_INTRWAIT);
+ if (error!= 0)
+ return (iic2errno(error));
slave = error = 0;
while (uio->uio_resid > 0) {
if (uio->uio_offset >= sc->size)
@@ -256,7 +255,7 @@ icee_write(struct cdev *dev, struct uio
break;
}
}
- ICEE_UNLOCK(sc);
+ iicbus_release_bus(sc->sc_busdev, sc->sc_dev);
return error;
}
___
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: r283526 - in head: . contrib/llvm/include/llvm/Target contrib/llvm/lib/Analysis contrib/llvm/lib/CodeGen contrib/llvm/lib/CodeGen/SelectionDAG contrib/llvm/lib/ExecutionEngine/RuntimeD
On Mon, 25 May 2015 13:43:03 + (UTC) Dimitry Andric wrote: > Author: dim > Date: Mon May 25 13:43:03 2015 > New Revision: 283526 > URL: https://svnweb.freebsd.org/changeset/base/283526 > > Log: > Upgrade our copy of clang and llvm to 3.6.1 release. > Hello. After update i have next building problem: IF i use CPUTYPE?=native flag in /etc/make.conf - compilation fails: ... --- crypto.o --- /usr/local/libexec/ccache/world/cc -target x86_64-unknown-freebsd11.0 --sysroot=/media/da0s1/obj/usr/src/tmp -B/media/da0s1/obj/usr/src/tmp/usr/bin -O2 -pipe -march=native -I/usr/src/usr.bin/svn/lib/libsvn_subr/../../../../contrib/subversion/subversion/include -I/usr/src/usr.bin/svn/lib/libsvn_subr/../../../../contrib/subversion/subversion -I/usr/src/usr.bin/svn/lib/libsvn_subr/../.. -I/usr/src/usr.bin/svn/lib/libsvn_subr/../libapr -I/usr/src/usr.bin/svn/lib/libsvn_subr/../../../../contrib/apr/include/arch/unix -I/usr/src/usr.bin/svn/lib/libsvn_subr/../../../../contrib/apr/include -I/usr/src/usr.bin/svn/lib/libsvn_subr/../libapr_util -I/usr/src/usr.bin/svn/lib/libsvn_subr/../../../../contrib/apr-util/include/private -I/usr/src/usr.bin/svn/lib/libsvn_subr/../../../../contrib/apr-util/include -I/usr/src/usr.bin/svn/lib/libsvn_subr/../../../../contrib/sqlite3 -std=gnu99 -fstack-protector-strong -Wno-pointer-sign -Wno-empty-body -Wno-string-plus-int -Wno-unused-const-variable -W no-tautological-compare -Wno-unused-value -Wno-parentheses-equality -Wno-unused-function -Wno-enum-conversion -Wno-unused-local-typedef -Wno-switch -Wno-switch-enum -Wno-knr-promoted-parameter -Wno-parentheses -Qunused-arguments -c /usr/src/usr.bin/svn/lib/libsvn_subr/../../../../contrib/subversion/subversion/libsvn_subr/crypto.c -o crypto.o --- usr.sbin.all__D --- --- ieee802_11_common.o --- distcc[67017] ERROR: compile (null) on localhost failed with exit code 254 Cannot emit physreg copy instruction UNREACHABLE executed at /usr/src/lib/clang/libllvmx86codegen/../../../contrib/llvm/lib/Target/X86/X86InstrInfo.cpp:3935! cc: error: unable to execute command: Abort trap cc: error: clang frontend command failed due to signal (use -v to see invocation) FreeBSD clang version 3.7.0 (tags/RELEASE_370/final 246257) 20150906 Target: x86_64-unknown-freebsd11.0 Thread model: posix cc: note: diagnostic msg: PLEASE submit a bug report to https://bugs.freebsd.org/submit/ and include the crash backtrace, preprocessed source, and associated run script. cc: note: diagnostic msg: PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT: Preprocessed source(s) and associated run script(s) are located at: cc: note: diagnostic msg: /tmp/ieee802_11_common-974cbf.c cc: note: diagnostic msg: /tmp/ieee802_11_common-974cbf.sh cc: note: diagnostic msg: *** [ieee802_11_common.o] Error code 254 make[5]: stopped in /usr/src/usr.sbin/wpa/wpa_supplicant 1 error make[5]: stopped in /usr/src/usr.sbin/wpa/wpa_supplicant *** [all_subdir_wpa_supplicant] Error code 2 make[4]: stopped in /usr/src/usr.sbin/wpa 1 error make[4]: stopped in /usr/src/usr.sbin/wpa *** [all_subdir_wpa] Error code 2 make[3]: stopped in /usr/src/usr.sbin 1 error make[3]: stopped in /usr/src/usr.sbin *** [usr.sbin.all__D] Error code 2 make[2]: stopped in /usr/src --- usr.bin.all__D --- A failure has been detected in another branch of the parallel make make[6]: stopped in /usr/src/usr.bin/svn/lib/libsvn_subr *** [_sub.all] Error code 2 make[5]: stopped in /usr/src/usr.bin/svn/lib 1 error make[5]: stopped in /usr/src/usr.bin/svn/lib *** [all_subdir_lib] Error code 2 make[4]: stopped in /usr/src/usr.bin/svn 1 error make[4]: stopped in /usr/src/usr.bin/svn *** [all_subdir_svn] Error code 2 make[3]: stopped in /usr/src/usr.bin 1 error make[3]: stopped in /usr/src/usr.bin *** [usr.bin.all__D] Error code 2 make[2]: stopped in /usr/src 2 errors make[2]: stopped in /usr/src *** [everything] Error code 2 make[1]: stopped in /usr/src 1 error make[1]: stopped in /usr/src *** [buildworld] Error code 2 make: stopped in /usr/src 1 error make: stopped in /usr/src If remove CPUTYPE?=native - compilation is successful. I have sandybridge CPU. Thanks. ___ 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: r289123 - in head: contrib/dma contrib/dma/debian libexec/dma
Author: bapt
Date: Sat Oct 10 23:31:47 2015
New Revision: 289123
URL: https://svnweb.freebsd.org/changeset/base/289123
Log:
Update Dragonfly Mail Agent to v0.10
Deleted:
head/contrib/dma/BSDmakefile
head/contrib/dma/debian/
Modified:
head/contrib/dma/LICENSE
head/contrib/dma/README.markdown
head/contrib/dma/TODO
head/contrib/dma/VERSION
head/contrib/dma/aliases_parse.y
head/contrib/dma/aliases_scan.l
head/contrib/dma/conf.c
head/contrib/dma/dma-mbox-create.c
head/contrib/dma/dma.8
head/contrib/dma/dma.c
head/contrib/dma/dma.h
head/contrib/dma/dns.c
head/contrib/dma/get-version.sh
head/contrib/dma/local.c
head/contrib/dma/mail.c
head/contrib/dma/net.c
head/contrib/dma/spool.c
head/contrib/dma/util.c
head/libexec/dma/Makefile.inc
Directory Properties:
head/contrib/dma/ (props changed)
Modified: head/contrib/dma/LICENSE
==
--- head/contrib/dma/LICENSESat Oct 10 23:22:55 2015(r289122)
+++ head/contrib/dma/LICENSESat Oct 10 23:31:47 2015(r289123)
@@ -1,5 +1,5 @@
+Copyright (c) 2008-2014, Simon Schubert <2...@0x2c.org>.
Copyright (c) 2008 The DragonFly Project.
-Copyright (c) 2008-2011, Simon Schubert <2...@0x2c.org>.
All rights reserved.
This code is derived from software contributed to The DragonFly Project
@@ -37,7 +37,7 @@ OF THE USE OF THIS SOFTWARE, EVEN IF ADV
SUCH DAMAGE.
-Copyright (c) 1995-2001 Kungliga Tekniska H�gskolan
+Copyright (c) 1995-2001 Kungliga Tekniska Högskolan
(Royal Institute of Technology, Stockholm, Sweden).
All rights reserved.
Modified: head/contrib/dma/README.markdown
==
--- head/contrib/dma/README.markdownSat Oct 10 23:22:55 2015
(r289122)
+++ head/contrib/dma/README.markdownSat Oct 10 23:31:47 2015
(r289123)
@@ -15,8 +15,13 @@ incoming connections.
Building
+In Linux:
+
make
+In BSD:
+
+ cd bsd && make
Installation
Modified: head/contrib/dma/TODO
==
--- head/contrib/dma/TODO Sat Oct 10 23:22:55 2015(r289122)
+++ head/contrib/dma/TODO Sat Oct 10 23:31:47 2015(r289123)
@@ -1,5 +1,4 @@
- unquote/handle quoted local recipients
-- use proper sysexit codes
- handle/use ESMTP extensions
- .forward support
- suggest way to run a queue flush on boot
Modified: head/contrib/dma/VERSION
==
--- head/contrib/dma/VERSIONSat Oct 10 23:22:55 2015(r289122)
+++ head/contrib/dma/VERSIONSat Oct 10 23:31:47 2015(r289123)
@@ -1 +1 @@
-v0.9
+v0.10
Modified: head/contrib/dma/aliases_parse.y
==
--- head/contrib/dma/aliases_parse.ySat Oct 10 23:22:55 2015
(r289122)
+++ head/contrib/dma/aliases_parse.ySat Oct 10 23:31:47 2015
(r289123)
@@ -7,8 +7,6 @@
extern int yylineno;
static void yyerror(const char *);
-int yywrap(void);
-int yylex(void);
static void
yyerror(const char *msg)
Modified: head/contrib/dma/aliases_scan.l
==
--- head/contrib/dma/aliases_scan.l Sat Oct 10 23:22:55 2015
(r289122)
+++ head/contrib/dma/aliases_scan.l Sat Oct 10 23:31:47 2015
(r289123)
@@ -1,11 +1,11 @@
%{
#include
+#include "dma.h"
#include "aliases_parse.h"
#define YY_NO_INPUT
-int yylex(void);
%}
%option yylineno
Modified: head/contrib/dma/conf.c
==
--- head/contrib/dma/conf.c Sat Oct 10 23:22:55 2015(r289122)
+++ head/contrib/dma/conf.c Sat Oct 10 23:31:47 2015(r289123)
@@ -64,7 +64,7 @@ trim_line(char *line)
if (line[0] == '.') {
if ((linelen + 2) > 1000) {
syslog(LOG_CRIT, "Cannot escape leading dot. Buffer
overflow");
- exit(1);
+ exit(EX_DATAERR);
}
memmove((line + 1), line, (linelen + 1));
line[0] = '.';
@@ -101,7 +101,7 @@ parse_authfile(const char *path)
a = fopen(path, "r");
if (a == NULL) {
- errlog(1, "can not open auth file `%s'", path);
+ errlog(EX_NOINPUT, "can not open auth file `%s'", path);
/* NOTREACHED */
}
@@ -121,7 +121,7 @@ parse_authfile(const char *path)
au = calloc(1, sizeof(*au));
if (au == NULL)
- errlog(1, "calloc failed");
+ errlog(EX_OSERR, NULL);
data = strdup(line);
au->login = strsep(&data, "|");
@@ -131,8 +131,7 @@ parse_authfile(const char *pa
Re: svn commit: r283526 - in head: . contrib/llvm/include/llvm/Target contrib/llvm/lib/Analysis contrib/llvm/lib/CodeGen contrib/llvm/lib/CodeGen/SelectionDAG contrib/llvm/lib/ExecutionEngine/RuntimeD
On 10 Oct 2015, at 23:32, Ivan Klymenko wrote: > On Mon, 25 May 2015 13:43:03 + (UTC) > Dimitry Andric wrote: > >> Author: dim >> Date: Mon May 25 13:43:03 2015 >> New Revision: 283526 >> URL: https://svnweb.freebsd.org/changeset/base/283526 >> >> Log: >> Upgrade our copy of clang and llvm to 3.6.1 release. This is some time ago already, you probably meant the upgrade to 3.7.0? :-) > Cannot emit physreg copy instruction > UNREACHABLE executed at > /usr/src/lib/clang/libllvmx86codegen/../../../contrib/llvm/lib/Target/X86/X86InstrInfo.cpp:3935! > cc: error: unable to execute command: Abort trap > cc: error: clang frontend command failed due to signal (use -v to see > invocation) > FreeBSD clang version 3.7.0 (tags/RELEASE_370/final 246257) 20150906 > Target: x86_64-unknown-freebsd11.0 > Thread model: posix > cc: note: diagnostic msg: PLEASE submit a bug report to > https://bugs.freebsd.org/submit/ and include the crash backtrace, > preprocessed source, and associated run script. > cc: note: diagnostic msg: > > > PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT: > Preprocessed source(s) and associated run script(s) are located at: > cc: note: diagnostic msg: /tmp/ieee802_11_common-974cbf.c > cc: note: diagnostic msg: /tmp/ieee802_11_common-974cbf.sh Yes, this error has already been reported; please see the thread starting here: https://lists.freebsd.org/pipermail/freebsd-current/2015-October/057701.html It has been submitted as an upstream bug, and a fix is being worked on. For now, you can work around it by unsetting CPUTYPE. -Dimitry signature.asc Description: Message signed with OpenPGP using GPGMail
svn commit: r289124 - head/sys/dev/wpi
Author: adrian
Date: Sun Oct 11 01:31:18 2015
New Revision: 289124
URL: https://svnweb.freebsd.org/changeset/base/289124
Log:
wpi(4): use more correct types.
This change fixes some amount of -Wsign-conversion and -Wconversion warnings
and sets correct sizes for some variables (as a result, some loop counters
were touched too).
Submitted by:
Differential Revision:https://reviews.freebsd.org/D3763
Modified:
head/sys/dev/wpi/if_wpi.c
head/sys/dev/wpi/if_wpireg.h
head/sys/dev/wpi/if_wpivar.h
Modified: head/sys/dev/wpi/if_wpi.c
==
--- head/sys/dev/wpi/if_wpi.c Sat Oct 10 23:31:47 2015(r289123)
+++ head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:31:18 2015(r289124)
@@ -156,7 +156,7 @@ static void wpi_update_rx_ring_ps(struct
static voidwpi_reset_rx_ring(struct wpi_softc *);
static voidwpi_free_rx_ring(struct wpi_softc *);
static int wpi_alloc_tx_ring(struct wpi_softc *, struct wpi_tx_ring *,
- int);
+ uint8_t);
static voidwpi_update_tx_ring(struct wpi_softc *, struct wpi_tx_ring *);
static voidwpi_update_tx_ring_ps(struct wpi_softc *,
struct wpi_tx_ring *);
@@ -165,15 +165,14 @@ static void wpi_free_tx_ring(struct wpi_
static int wpi_read_eeprom(struct wpi_softc *,
uint8_t macaddr[IEEE80211_ADDR_LEN]);
static uint32_twpi_eeprom_channel_flags(struct wpi_eeprom_chan *);
-static voidwpi_read_eeprom_band(struct wpi_softc *, int);
-static int wpi_read_eeprom_channels(struct wpi_softc *, int);
+static voidwpi_read_eeprom_band(struct wpi_softc *, uint8_t);
+static int wpi_read_eeprom_channels(struct wpi_softc *, uint8_t);
static struct wpi_eeprom_chan *wpi_find_eeprom_channel(struct wpi_softc *,
struct ieee80211_channel *);
static int wpi_setregdomain(struct ieee80211com *,
struct ieee80211_regdomain *, int,
struct ieee80211_channel[]);
-static int wpi_read_eeprom_group(struct wpi_softc *, int);
-static int wpi_add_node_entry_adhoc(struct wpi_softc *);
+static int wpi_read_eeprom_group(struct wpi_softc *, uint8_t);
static struct ieee80211_node *wpi_node_alloc(struct ieee80211vap *,
const uint8_t mac[IEEE80211_ADDR_LEN]);
static voidwpi_node_free(struct ieee80211_node *);
@@ -210,7 +209,8 @@ static void wpi_watchdog_rfkill(void *);
static voidwpi_scan_timeout(void *);
static voidwpi_tx_timeout(void *);
static voidwpi_parent(struct ieee80211com *);
-static int wpi_cmd(struct wpi_softc *, int, const void *, size_t, int);
+static int wpi_cmd(struct wpi_softc *, uint8_t, const void *, uint16_t,
+ int);
static int wpi_mrr_setup(struct wpi_softc *);
static int wpi_add_node(struct wpi_softc *, struct ieee80211_node *);
static int wpi_add_broadcast_node(struct wpi_softc *, int);
@@ -257,7 +257,8 @@ static int wpi_key_set(struct ieee80211v
static int wpi_key_delete(struct ieee80211vap *,
const struct ieee80211_key *);
static int wpi_post_alive(struct wpi_softc *);
-static int wpi_load_bootcode(struct wpi_softc *, const uint8_t *, int);
+static int wpi_load_bootcode(struct wpi_softc *, const uint8_t *,
+ uint32_t);
static int wpi_load_firmware(struct wpi_softc *);
static int wpi_read_firmware(struct wpi_softc *);
static voidwpi_unload_firmware(struct wpi_softc *);
@@ -327,7 +328,8 @@ wpi_attach(device_t dev)
{
struct wpi_softc *sc = (struct wpi_softc *)device_get_softc(dev);
struct ieee80211com *ic;
- int i, error, rid;
+ uint8_t i;
+ int error, rid;
#ifdef WPI_DEBUG
int supportsa = 1;
const struct wpi_ident *ident;
@@ -387,10 +389,11 @@ wpi_attach(device_t dev)
sc->sc_st = rman_get_bustag(sc->mem);
sc->sc_sh = rman_get_bushandle(sc->mem);
- i = 1;
- rid = 0;
- if (pci_alloc_msi(dev, &i) == 0)
+ rid = 1;
+ if (pci_alloc_msi(dev, &rid) == 0)
rid = 1;
+ else
+ rid = 0;
/* Install interrupt handler. */
sc->irq = bus_alloc_resource_any(dev, SYS_RES_IRQ, &rid, RF_ACTIVE |
(rid != 0 ? 0 : RF_SHAREABLE));
@@ -677,7 +680,7 @@ wpi_detach(device_t dev)
{
struct wpi_softc *sc = device_get_softc(dev);
struct ieee80211com *ic = &sc->sc_ic;
- int qid;
+ uint8_t qid;
DPRINTF(sc, WPI_DEBUG_TRACE, TRACE_STR_BEGIN, __func__);
@@ -828,9 +831,9 @@ wpi_prph_clrbits(struct wpi_softc *sc, u
static __inline void
wpi_prph_write_region_4(struct wpi_softc *sc, uint32_t addr,
-const uint32_t *data, int count)
+const uint32_t *data, uint32_t count)
{
- for (; count > 0; count--, data++, addr += 4)
+ for (; count != 0; count--, data++, addr += 4)
svn commit: r289125 - head/sys/dev/wpi
Author: adrian
Date: Sun Oct 11 01:50:54 2015
New Revision: 289125
URL: https://svnweb.freebsd.org/changeset/base/289125
Log:
wpi(4): do not override iv_recv_mgmt() in non-adhoc modes.
Tested:
* Tested with Intel 3945BG, IBSS and STA modes
Submitted by:
Differential Revision:https://reviews.freebsd.org/D3764
Modified:
head/sys/dev/wpi/if_wpi.c
Modified: head/sys/dev/wpi/if_wpi.c
==
--- head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:31:18 2015(r289124)
+++ head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:50:54 2015(r289125)
@@ -176,7 +176,7 @@ static int wpi_read_eeprom_group(struct
static struct ieee80211_node *wpi_node_alloc(struct ieee80211vap *,
const uint8_t mac[IEEE80211_ADDR_LEN]);
static voidwpi_node_free(struct ieee80211_node *);
-static voidwpi_recv_mgmt(struct ieee80211_node *, struct mbuf *, int,
+static voidwpi_ibss_recv_mgmt(struct ieee80211_node *, struct mbuf *, int,
const struct ieee80211_rx_stats *,
int, int);
static voidwpi_restore_node(void *, struct ieee80211_node *);
@@ -640,8 +640,10 @@ wpi_vap_create(struct ieee80211com *ic,
/* Override with driver methods. */
vap->iv_key_set = wpi_key_set;
vap->iv_key_delete = wpi_key_delete;
- wvp->wv_recv_mgmt = vap->iv_recv_mgmt;
- vap->iv_recv_mgmt = wpi_recv_mgmt;
+ if (opmode == IEEE80211_M_IBSS) {
+ wvp->wv_recv_mgmt = vap->iv_recv_mgmt;
+ vap->iv_recv_mgmt = wpi_ibss_recv_mgmt;
+ }
wvp->wv_newstate = vap->iv_newstate;
vap->iv_newstate = wpi_newstate;
vap->iv_update_beacon = wpi_update_beacon;
@@ -1664,7 +1666,7 @@ wpi_check_bss_filter(struct wpi_softc *s
}
static void
-wpi_recv_mgmt(struct ieee80211_node *ni, struct mbuf *m, int subtype,
+wpi_ibss_recv_mgmt(struct ieee80211_node *ni, struct mbuf *m, int subtype,
const struct ieee80211_rx_stats *rxs,
int rssi, int nf)
{
@@ -1675,8 +1677,7 @@ wpi_recv_mgmt(struct ieee80211_node *ni,
wvp->wv_recv_mgmt(ni, m, subtype, rxs, rssi, nf);
- if (vap->iv_opmode == IEEE80211_M_IBSS &&
- vap->iv_state == IEEE80211_S_RUN &&
+ if (vap->iv_state == IEEE80211_S_RUN &&
(subtype == IEEE80211_FC0_SUBTYPE_BEACON ||
subtype == IEEE80211_FC0_SUBTYPE_PROBE_RESP)) {
ni_tstamp = le64toh(ni->ni_tstamp.tsf);
___
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: r289126 - head/sys/dev/wpi
Author: adrian
Date: Sun Oct 11 01:53:51 2015
New Revision: 289126
URL: https://svnweb.freebsd.org/changeset/base/289126
Log:
wpi(4): do not allocate space for unused rings.
Tested:
* Tested with Intel 3945BG, STA mode
Submitted by:
Differential Revision:https://reviews.freebsd.org/D3765
Modified:
head/sys/dev/wpi/if_wpi.c
head/sys/dev/wpi/if_wpivar.h
Modified: head/sys/dev/wpi/if_wpi.c
==
--- head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:50:54 2015(r289125)
+++ head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:53:51 2015(r289126)
@@ -425,7 +425,7 @@ wpi_attach(device_t dev)
}
/* Allocate TX rings - 4 for QoS purposes, 1 for commands. */
- for (i = 0; i < WPI_NTXQUEUES; i++) {
+ for (i = 0; i < WPI_DRV_NTXQUEUES; i++) {
if ((error = wpi_alloc_tx_ring(sc, &sc->txq[i], i)) != 0) {
device_printf(dev,
"could not allocate TX ring %d, error %d\n", i,
@@ -713,7 +713,7 @@ wpi_detach(device_t dev)
if (sc->txq[0].data_dmat) {
/* Free DMA resources. */
- for (qid = 0; qid < WPI_NTXQUEUES; qid++)
+ for (qid = 0; qid < WPI_DRV_NTXQUEUES; qid++)
wpi_free_tx_ring(sc, &sc->txq[qid]);
wpi_free_rx_ring(sc);
@@ -1186,16 +1186,6 @@ wpi_alloc_tx_ring(struct wpi_softc *sc,
bus_dmamap_sync(sc->shared_dma.tag, sc->shared_dma.map,
BUS_DMASYNC_PREWRITE);
- /*
-* We only use rings 0 through 4 (4 EDCA + cmd) so there is no need
-* to allocate commands space for other rings.
-* XXX Do we really need to allocate descriptors for other rings?
-*/
- if (qid > WPI_CMD_QUEUE_NUM) {
- DPRINTF(sc, WPI_DEBUG_TRACE, TRACE_STR_END, __func__);
- return 0;
- }
-
size = WPI_TX_RING_COUNT * sizeof (struct wpi_tx_cmd);
error = wpi_dma_contig_alloc(sc, &ring->cmd_dma, (void **)&ring->cmd,
size, 4);
@@ -5326,7 +5316,7 @@ wpi_hw_stop(struct wpi_softc *sc)
wpi_reset_rx_ring(sc);
/* Reset all TX rings. */
- for (qid = 0; qid < WPI_NTXQUEUES; qid++)
+ for (qid = 0; qid < WPI_DRV_NTXQUEUES; qid++)
wpi_reset_tx_ring(sc, &sc->txq[qid]);
if (wpi_nic_lock(sc) == 0) {
Modified: head/sys/dev/wpi/if_wpivar.h
==
--- head/sys/dev/wpi/if_wpivar.hSun Oct 11 01:50:54 2015
(r289125)
+++ head/sys/dev/wpi/if_wpivar.hSun Oct 11 01:53:51 2015
(r289126)
@@ -176,7 +176,7 @@ struct wpi_softc {
struct wpi_dma_info shared_dma;
struct wpi_shared *shared;
- struct wpi_tx_ring txq[WPI_NTXQUEUES];
+ struct wpi_tx_ring txq[WPI_DRV_NTXQUEUES];
struct mtx txq_mtx;
struct mtx txq_state_mtx;
___
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: r289127 - head/sys/dev/wpi
Author: adrian
Date: Sun Oct 11 01:56:46 2015
New Revision: 289127
URL: https://svnweb.freebsd.org/changeset/base/289127
Log:
wpi(4): fix possible race between TX/RX threads.
Tested:
* Tested with Intel 3945BG, STA mode
Submitted by:
Differential Revision:https://reviews.freebsd.org/D3766
Modified:
head/sys/dev/wpi/if_wpi.c
Modified: head/sys/dev/wpi/if_wpi.c
==
--- head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:53:51 2015(r289126)
+++ head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:56:46 2015(r289127)
@@ -1093,6 +1093,7 @@ wpi_update_rx_ring_ps(struct wpi_softc *
return;
}
+ WPI_TXQ_LOCK(sc);
WPI_SETBITS(sc, WPI_GP_CNTRL, WPI_GP_CNTRL_MAC_ACCESS_REQ);
if (WPI_READ(sc, WPI_GP_CNTRL) & WPI_GP_CNTRL_SLEEP) {
DPRINTF(sc, WPI_DEBUG_PWRSAVE, "%s: wakeup request\n",
@@ -1102,6 +1103,7 @@ wpi_update_rx_ring_ps(struct wpi_softc *
wpi_update_rx_ring(sc);
WPI_CLRBITS(sc, WPI_GP_CNTRL, WPI_GP_CNTRL_MAC_ACCESS_REQ);
}
+ WPI_TXQ_UNLOCK(sc);
}
static void
___
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: r289128 - head/sys/dev/wpi
Author: adrian
Date: Sun Oct 11 01:58:50 2015
New Revision: 289128
URL: https://svnweb.freebsd.org/changeset/base/289128
Log:
wpi(4): move error handling upper (wpi_tx_data()/wpi_cmd2() ->
wpi_raw_xmit()/wpi_transmit()).
Tested:
* Tested with Intel 3945BG, STA mode
Submitted by:
Differential Revision:https://reviews.freebsd.org/D3767
Modified:
head/sys/dev/wpi/if_wpi.c
Modified: head/sys/dev/wpi/if_wpi.c
==
--- head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:56:46 2015(r289127)
+++ head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:58:50 2015(r289128)
@@ -2595,7 +2595,7 @@ wpi_cmd2(struct wpi_softc *sc, struct wp
if (__predict_false(sc->sc_running == 0)) {
/* wpi_stop() was called */
error = ENETDOWN;
- goto fail;
+ goto end;
}
wh = mtod(buf->m, struct ieee80211_frame *);
@@ -2604,7 +2604,7 @@ wpi_cmd2(struct wpi_softc *sc, struct wp
if (__predict_false(totlen < sizeof(struct ieee80211_frame_min))) {
error = EINVAL;
- goto fail;
+ goto end;
}
if (hdrlen & 3) {
@@ -2635,7 +2635,7 @@ wpi_cmd2(struct wpi_softc *sc, struct wp
if (error != 0 && error != EFBIG) {
device_printf(sc->sc_dev,
"%s: can't map mbuf (error %d)\n", __func__, error);
- goto fail;
+ goto end;
}
if (error != 0) {
/* Too many DMA segments, linearize mbuf. */
@@ -2644,17 +2644,29 @@ wpi_cmd2(struct wpi_softc *sc, struct wp
device_printf(sc->sc_dev,
"%s: could not defrag mbuf\n", __func__);
error = ENOBUFS;
- goto fail;
+ goto end;
}
buf->m = m1;
error = bus_dmamap_load_mbuf_sg(ring->data_dmat, data->map,
buf->m, segs, &nsegs, BUS_DMA_NOWAIT);
if (__predict_false(error != 0)) {
+ /* XXX fix this (applicable to the iwn(4) too) */
+ /*
+* NB: Do not return error;
+* original mbuf does not exist anymore.
+*/
device_printf(sc->sc_dev,
"%s: can't map mbuf (error %d)\n", __func__,
error);
- goto fail;
+ if (ring->qid < WPI_CMD_QUEUE_NUM) {
+ if_inc_counter(buf->ni->ni_vap->iv_ifp,
+ IFCOUNTER_OERRORS, 1);
+ ieee80211_free_node(buf->ni);
+ }
+ m_freem(buf->m);
+ error = 0;
+ goto end;
}
}
@@ -2698,19 +2710,12 @@ wpi_cmd2(struct wpi_softc *sc, struct wp
WPI_TXQ_STATE_UNLOCK(sc);
}
- DPRINTF(sc, WPI_DEBUG_TRACE, TRACE_STR_END, __func__);
-
- WPI_TXQ_UNLOCK(sc);
-
- return 0;
-
-fail: m_freem(buf->m);
-
- DPRINTF(sc, WPI_DEBUG_TRACE, TRACE_STR_END_ERR, __func__);
+end: DPRINTF(sc, WPI_DEBUG_TRACE, error ? TRACE_STR_END_ERR : TRACE_STR_END,
+ __func__);
WPI_TXQ_UNLOCK(sc);
- return error;
+ return (error);
}
/*
@@ -2731,7 +2736,7 @@ wpi_tx_data(struct wpi_softc *sc, struct
uint32_t flags;
uint16_t ac, qos;
uint8_t tid, type, rate;
- int error, swcrypt, ismcast, totlen;
+ int swcrypt, ismcast, totlen;
wh = mtod(m, struct ieee80211_frame *);
type = wh->i_fc[0] & IEEE80211_FC0_TYPE_MASK;
@@ -2771,10 +2776,9 @@ wpi_tx_data(struct wpi_softc *sc, struct
if (wh->i_fc[1] & IEEE80211_FC1_PROTECTED) {
/* Retrieve key for TX. */
k = ieee80211_crypto_encap(ni, m);
- if (k == NULL) {
- error = ENOBUFS;
- goto fail;
- }
+ if (k == NULL)
+ return (ENOBUFS);
+
swcrypt = k->wk_flags & IEEE80211_KEY_SWCRYPT;
/* 802.11 header may have moved. */
@@ -2843,8 +2847,7 @@ wpi_tx_data(struct wpi_softc *sc, struct
if (wn->id == WPI_ID_UNDEFINED) {
device_printf(sc->sc_dev,
"%s: undefined node id\n", __func__);
- error = EINVAL;
- goto fail;
+ return (EINVAL);
}
tx->id = wn->id;
@@ -2880,9 +2883,6 @@ wpi_tx_data(struct wpi_softc *sc, struct
tx_data.ac = ac;
return wpi_cmd2(sc, &tx_data);
-
-fail: m_freem(m);
- return error;
}
static int
@@ -2923,10 +2923,9 @@ w
svn commit: r289129 - head/sys/dev/wpi
Author: adrian
Date: Sun Oct 11 02:00:08 2015
New Revision: 289129
URL: https://svnweb.freebsd.org/changeset/base/289129
Log:
wpi(4): do not count failures twice for ic_raw_xmit().
Tested:
* Tested with Intel 3945BG, STA mode
Submitted by:
Differential Revision:https://reviews.freebsd.org/D3768
Modified:
head/sys/dev/wpi/if_wpi.c
Modified: head/sys/dev/wpi/if_wpi.c
==
--- head/sys/dev/wpi/if_wpi.c Sun Oct 11 01:58:50 2015(r289128)
+++ head/sys/dev/wpi/if_wpi.c Sun Oct 11 02:00:08 2015(r289129)
@@ -3001,14 +3001,6 @@ wpi_tx_ring_is_full(struct wpi_softc *sc
return retval;
}
-static __inline void
-wpi_handle_tx_failure(struct ieee80211_node *ni)
-{
- /* NB: m is reclaimed on tx failure */
- if_inc_counter(ni->ni_vap->iv_ifp, IFCOUNTER_OERRORS, 1);
- ieee80211_free_node(ni);
-}
-
static int
wpi_raw_xmit(struct ieee80211_node *ni, struct mbuf *m,
const struct ieee80211_bpf_params *params)
@@ -3046,8 +3038,8 @@ wpi_raw_xmit(struct ieee80211_node *ni,
unlock:WPI_TX_UNLOCK(sc);
if (error != 0) {
- wpi_handle_tx_failure(ni);
m_freem(m);
+ ieee80211_free_node(ni);
DPRINTF(sc, WPI_DEBUG_TRACE, TRACE_STR_END_ERR, __func__);
return error;
@@ -3085,7 +3077,8 @@ wpi_transmit(struct ieee80211com *ic, st
error = 0;
ni = (struct ieee80211_node *)m->m_pkthdr.rcvif;
if (wpi_tx_data(sc, m, ni) != 0) {
- wpi_handle_tx_failure(ni);
+ if_inc_counter(ni->ni_vap->iv_ifp, IFCOUNTER_OERRORS, 1);
+ ieee80211_free_node(ni);
m_freem(m);
}
___
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: r289130 - head/sys/mips/conf
Author: sbruno Date: Sun Oct 11 03:31:11 2015 New Revision: 289130 URL: https://svnweb.freebsd.org/changeset/base/289130 Log: Use machine specific values cleaned from openwrt for the mac address location on the TP link mr3020 Modified: head/sys/mips/conf/TP-MR3020.hints Modified: head/sys/mips/conf/TP-MR3020.hints == --- head/sys/mips/conf/TP-MR3020.hints Sun Oct 11 02:00:08 2015 (r289129) +++ head/sys/mips/conf/TP-MR3020.hints Sun Oct 11 03:31:11 2015 (r289130) @@ -34,8 +34,10 @@ hint.arge.1.fduplex=1 # Where the ART is - last 64k in the flash # 0x9fff1000 ? -hint.ath.0.eepromaddr=0x1fff +hint.ath.0.eepromaddr=0x1fff1000 hint.ath.0.eepromsize=16384 + +hint.ar71xx.0.eeprom_mac_addr=0x1f01fc00 # The board 16MiB flash layout in uboot env: # ___ 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: r289130 - head/sys/mips/conf
Hm, its an AR9300, right? It should be getting ART from 0x1fff like the rest of the boards and the AR9300 HAL will add 0x1000 for you. Don't ask me why. What's the MAC address of the wlan0 interface you created? Does it actually all work? :) -a On 10 October 2015 at 20:31, Sean Bruno wrote: > Author: sbruno > Date: Sun Oct 11 03:31:11 2015 > New Revision: 289130 > URL: https://svnweb.freebsd.org/changeset/base/289130 > > Log: > Use machine specific values cleaned from openwrt for the mac address > location on the TP link mr3020 > > Modified: > head/sys/mips/conf/TP-MR3020.hints > > Modified: head/sys/mips/conf/TP-MR3020.hints > == > --- head/sys/mips/conf/TP-MR3020.hints Sun Oct 11 02:00:08 2015 > (r289129) > +++ head/sys/mips/conf/TP-MR3020.hints Sun Oct 11 03:31:11 2015 > (r289130) > @@ -34,8 +34,10 @@ hint.arge.1.fduplex=1 > > # Where the ART is - last 64k in the flash > # 0x9fff1000 ? > -hint.ath.0.eepromaddr=0x1fff > +hint.ath.0.eepromaddr=0x1fff1000 > hint.ath.0.eepromsize=16384 > + > +hint.ar71xx.0.eeprom_mac_addr=0x1f01fc00 > > # The board 16MiB flash layout in uboot env: > # > ___ 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"
