Author: glebius
Date: Thu Feb 7 03:55:05 2019
New Revision: 343858
URL: https://svnweb.freebsd.org/changeset/base/343858
Log:
Now that there is only one way to allocate a slab, remove uz_slab method.
Discussed with: jeff
Modified:
head/sys/vm/uma_core.c
head/sys/vm/uma_int.h
Mo
Author: glebius
Date: Thu Feb 7 03:32:45 2019
New Revision: 343857
URL: https://svnweb.freebsd.org/changeset/base/343857
Log:
Report cache zones in UMA stats sysctl, that 'vmstat -z' uses. This
should had been part of r251826.
Modified:
head/sys/vm/uma_core.c
Modified: head/sys/vm/uma_cor
Author: glebius
Date: Sun Feb 3 08:28:02 2019
New Revision: 343707
URL: https://svnweb.freebsd.org/changeset/base/343707
Log:
Teach pfil_ioctl() about VIMAGE.
Submitted by: gallatin
Modified:
head/sys/net/pfil.c
Modified: head/sys/net/pfil.c
Author: glebius
Date: Sat Feb 2 05:49:05 2019
New Revision: 343678
URL: https://svnweb.freebsd.org/changeset/base/343678
Log:
Return PFIL_CONSUMED if packet was consumed. While here gather all
the identical endings of pf_check_*() into single function.
PR: 235411
Modified:
Author: glebius
Date: Fri Feb 1 08:10:26 2019
New Revision: 343646
URL: https://svnweb.freebsd.org/changeset/base/343646
Log:
Repair siftr(4): PFIL_IN and PFIL_OUT are defines of some value, relying
on them having particular values can break things.
Modified:
head/sys/netinet/siftr.c
Modi
Author: glebius
Date: Fri Feb 1 07:48:37 2019
New Revision: 343645
URL: https://svnweb.freebsd.org/changeset/base/343645
Log:
Unbreak call to ipf_check(): it expects the out parameter to be 0 or 1.
Pointy hat to:glebius
Reported by: cy
Modified:
head/sys/contrib/ipfilter/neti
Author: glebius
Date: Fri Feb 1 00:34:18 2019
New Revision: 343636
URL: https://svnweb.freebsd.org/changeset/base/343636
Log:
Hopefully fix compilation by other compilers.
Modified:
head/sbin/pfilctl/pfilctl.c
Modified: head/sbin/pfilctl/pfilctl.c
===
Author: glebius
Date: Fri Feb 1 00:33:17 2019
New Revision: 343635
URL: https://svnweb.freebsd.org/changeset/base/343635
Log:
Fix build without INET6.
Modified:
head/sys/netpfil/ipfw/ip_fw_pfil.c
Modified: head/sys/netpfil/ipfw/ip_fw_pfil.c
==
==
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/sbin/pfilctl/pfilctl.8 Thu Jan 31 23:01:03 2019(r343631)
@@ -0,0 +1,117 @@
+.\" Copyright (c) 2019 Gleb Smirnoff
+.\"
+.\" Redistribution and use in source a
Author: glebius
Date: Thu Jan 31 22:37:28 2019
New Revision: 343627
URL: https://svnweb.freebsd.org/changeset/base/343627
Log:
Add new m_ext type for data for M_NOFREE mbufs, which doesn't actually do
anything except several assertions. This type is going to be used for
temporary on stack m
Author: glebius
Date: Thu Jan 31 21:04:50 2019
New Revision: 343619
URL: https://svnweb.freebsd.org/changeset/base/343619
Log:
Revert r316461: Remove "IPFW static rules" rmlock, and use pfil's global lock.
The pfil(9) system is about to be converted to epoch(9) synchronization, so
we need
Author: glebius
Date: Thu Jan 31 17:52:48 2019
New Revision: 343616
URL: https://svnweb.freebsd.org/changeset/base/343616
Log:
In zone_alloc_bucket() max argument was calculated based on uz_count.
Then bucket_alloc() also selects bucket size based on uz_count. However,
since zone lock is dro
Author: glebius
Date: Wed Jan 16 20:20:38 2019
New Revision: 343101
URL: https://svnweb.freebsd.org/changeset/base/343101
Log:
Fix mistake in r343030: move nswbuf calculation back to
kern_vfs_bio_buffer_alloc(), because in init_param2() nbuf
isn't really initialized yet.
Pointed out by:
Author: glebius
Date: Wed Jan 16 20:14:16 2019
New Revision: 343100
URL: https://svnweb.freebsd.org/changeset/base/343100
Log:
Do not reserve KVA for paging bufs in vm_ksubmap_init(), since now
they allocate it in pbuf_init(). This should have been done together
with r343030.
Modified:
he
Author: glebius
Date: Wed Jan 16 04:02:08 2019
New Revision: 343080
URL: https://svnweb.freebsd.org/changeset/base/343080
Log:
Whitespace.
Modified:
head/sys/vm/uma_int.h
Modified: head/sys/vm/uma_int.h
==
--- head/s
On Tue, Jan 15, 2019 at 01:20:14PM -0700, Ian Lepore wrote:
I> On Tue, 2019-01-15 at 12:06 -0800, Gleb Smirnoff wrote:
I> > On Tue, Jan 15, 2019 at 01:46:23PM -0600, Justin Hibbits wrote:
I> > J> Why not #include counter.h in the relevant vm_machdep.c files
I> > instead?
les just need counter_u64_t size,
and they don't use counter(9) KPI.
The fact that vm_machdep or Linux KPI want to look into internal
type uma_zone_t is also ugly.
--
Gleb Smirnoff
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org
Author: glebius
Date: Tue Jan 15 19:33:47 2019
New Revision: 343058
URL: https://svnweb.freebsd.org/changeset/base/343058
Log:
Fix compilation failures on different arches that have vm_machdep.c not
aware of counter_u64_t by including counter.h into uma_int.h. I'm not
happy about this inclus
Author: glebius
Date: Tue Jan 15 18:50:11 2019
New Revision: 343055
URL: https://svnweb.freebsd.org/changeset/base/343055
Log:
style(9): break long line.
Modified:
head/sys/vm/uma_core.c
Modified: head/sys/vm/uma_core.c
Author: glebius
Date: Tue Jan 15 18:49:31 2019
New Revision: 343054
URL: https://svnweb.freebsd.org/changeset/base/343054
Log:
Remove harmless leftover from code that cycles over zone's kegs. Just use +
instead of +=. There is no functional change.
Modified:
head/sys/vm/uma_core.c
Modified
Author: glebius
Date: Tue Jan 15 18:47:19 2019
New Revision: 343053
URL: https://svnweb.freebsd.org/changeset/base/343053
Log:
This was missed in r343051: make uz_allocs, uz_frees and uz_fails counter(9).
Modified:
head/lib/libmemstat/memstat_uma.c
Modified: head/lib/libmemstat/memstat_uma.c
Author: glebius
Date: Tue Jan 15 18:32:26 2019
New Revision: 343052
URL: https://svnweb.freebsd.org/changeset/base/343052
Log:
Only do uz_items accounting for zones that have a limit set in uz_max_items.
This reduces amount of locking required for these zones.
Also, for cache only zones (
Author: glebius
Date: Tue Jan 15 18:24:34 2019
New Revision: 343051
URL: https://svnweb.freebsd.org/changeset/base/343051
Log:
Make uz_allocs, uz_frees and uz_fails counter(9). This removes some
atomic updates and reduces amount of data protected by zone lock.
During startup point these f
he finish line!
P> >
P> It appears to be very impressive! Plans for MFC?
Nope. I'm very conservative about stable branch being stable branch :)
--
Gleb Smirnoff
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinf
Author: glebius
Date: Tue Jan 15 03:43:46 2019
New Revision: 343037
URL: https://svnweb.freebsd.org/changeset/base/343037
Log:
Fix compilation on 32-bit.
Modified:
head/sys/vm/uma_core.c
Modified: head/sys/vm/uma_core.c
Author: glebius
Date: Tue Jan 15 01:19:29 2019
New Revision: 343031
URL: https://svnweb.freebsd.org/changeset/base/343031
Log:
Remove pbuf(9) manual page that is now nearly 100% outdated.
Deleted:
head/share/man/man9/pbuf.9
Modified:
head/ObsoleteFiles.inc
head/share/man/man9/Makefile
Mo
Author: glebius
Date: Tue Jan 15 01:02:16 2019
New Revision: 343030
URL: https://svnweb.freebsd.org/changeset/base/343030
Log:
Allocate pager bufs from UMA instead of 80-ish mutex protected linked list.
o In vm_pager_bufferinit() create pbuf_zone and start accounting on how many
pbufs a
Author: glebius
Date: Tue Jan 15 00:35:19 2019
New Revision: 343027
URL: https://svnweb.freebsd.org/changeset/base/343027
Log:
Add flag LK_NEW for lockinit() that is converted to LO_NEW and passed
down to lock_init(). This allows for lockinit() on a not prezeroed
memory.
Modified:
head/s
Author: glebius
Date: Tue Jan 15 00:02:06 2019
New Revision: 343026
URL: https://svnweb.freebsd.org/changeset/base/343026
Log:
o Move zone limit from keg level up to zone level. This means that now
two zones sharing a keg may have different limits. Now this is going
to work:
zone
(r343005)
J> @@ -886,6 +886,8 @@ solisten_wakeup(struct socket *sol)
J> }
J> SOLISTEN_UNLOCK(sol);
J> wakeup_one(&sol->sol_comp);
J> +if ((sol->so_state & SS_ASYNC) && sol->so_sigio != NULL)
J> +pgsigio(&s
Author: glebius
Date: Fri Jan 11 00:45:22 2019
New Revision: 342923
URL: https://svnweb.freebsd.org/changeset/base/342923
Log:
Remove recursive NET_EPOCH_ENTER() from sysctl_ifmalist(), missed in r342872.
Modified:
head/sys/net/rtsock.c
Modified: head/sys/net/rtsock.c
===
Author: glebius
Date: Thu Jan 10 23:27:29 2019
New Revision: 342922
URL: https://svnweb.freebsd.org/changeset/base/342922
Log:
Remove support for FreeBSD 9 kernel, which used to change byte order
of packet headers.
Modified:
head/sys/contrib/ipfilter/netinet/ip_fil_freebsd.c
Modified: head
mits/pfil
Soon to reach FreeBSD head :)
--
Gleb Smirnoff
___
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"
Author: glebius
Date: Thu Jan 10 00:52:06 2019
New Revision: 342907
URL: https://svnweb.freebsd.org/changeset/base/342907
Log:
Remove unnecessary ifdef. With INVARIANTS all KASSERTs are empty statements,
so won't be compiled in.
Modified:
head/sys/netinet6/in6_mcast.c
Modified: head/sys/ne
Author: glebius
Date: Thu Jan 10 00:37:14 2019
New Revision: 342906
URL: https://svnweb.freebsd.org/changeset/base/342906
Log:
Bring the comment up to date.
Modified:
head/sys/net/if_vlan.c
Modified: head/sys/net/if_vlan.c
=
Author: glebius
Date: Thu Jan 10 00:25:12 2019
New Revision: 342905
URL: https://svnweb.freebsd.org/changeset/base/342905
Log:
Simplify sosetopt() so that function has single return point. No
functional change.
Modified:
head/sys/kern/uipc_socket.c
Modified: head/sys/kern/uipc_socket.c
===
Author: glebius
Date: Wed Jan 9 05:53:59 2019
New Revision: 342877
URL: https://svnweb.freebsd.org/changeset/base/342877
Log:
Timer interval is correctly counted in ticks (1/100 s) in
mibif_reset_hc_timer(). Multiplication by 10 is erroneous
and is probably a blind copy and paste from next
Author: glebius
Date: Wed Jan 9 01:11:19 2019
New Revision: 342872
URL: https://svnweb.freebsd.org/changeset/base/342872
Log:
Mechanical cleanup of epoch(9) usage in network stack.
- Remove macros that covertly create epoch_tracker on thread stack. Such
macros a quite unsafe, e.g. will
Author: glebius
Date: Wed Jan 9 00:38:16 2019
New Revision: 342870
URL: https://svnweb.freebsd.org/changeset/base/342870
Log:
Remove part of comment that doesn't match reality.
Modified:
head/sys/net/if.c
Modified: head/sys/net/if.c
==
Author: glebius
Date: Mon Jan 7 23:17:09 2019
New Revision: 342854
URL: https://svnweb.freebsd.org/changeset/base/342854
Log:
Fix regression in r331368, that broke dumping of UMA startup pages
when WITNESS is present.
Discussed with: markj
Modified:
head/sys/vm/vm_page.c
Modifi
Author: glebius
Date: Tue Dec 18 19:38:13 2018
New Revision: 342197
URL: https://svnweb.freebsd.org/changeset/base/342197
Log:
Use mbufq instead of ifqueue to queue mbufs.
Modified:
head/sys/dev/usb/net/uhso.c
Modified: head/sys/dev/usb/net/uhso.c
Author: glebius
Date: Wed Nov 28 19:54:02 2018
New Revision: 341163
URL: https://svnweb.freebsd.org/changeset/base/341163
Log:
Fix yet another edge case in uma_startup_count(). If zone size fits into
several pages, but leaves no space for struct uma_slab at the end we
miscalculate number of
Author: glebius
Date: Wed Nov 28 19:17:27 2018
New Revision: 341162
URL: https://svnweb.freebsd.org/changeset/base/341162
Log:
For not offpage zones the slab is placed at the end of page. Keg's uk_pgoff
is calculated to guarantee that struct uma_slab is placed at pointer size
alignment. Calc
On Wed, Nov 14, 2018 at 10:27:48PM +0200, Konstantin Belousov wrote:
K> On Wed, Nov 14, 2018 at 08:28:31AM -0800, Gleb Smirnoff wrote:
K> > On Wed, Nov 14, 2018 at 11:06:38AM +0100, Hans Petter Selasky wrote:
K> > H> On 11/14/18 10:33 AM, Cy Schubert wrote:
K> > H> >
Author: glebius
Date: Wed Nov 14 19:10:35 2018
New Revision: 340443
URL: https://svnweb.freebsd.org/changeset/base/340443
Log:
Initialize compatibility epoch tracker for thread0. Fixes
panics for drivers that call if_maddr_lock() during startup.
Reported by: cy
Modified:
head/sys/kern
On Wed, Nov 14, 2018 at 11:06:38AM +0100, Hans Petter Selasky wrote:
H> On 11/14/18 10:33 AM, Cy Schubert wrote:
H> > + epoch_thread_init(td);
H>
H> Did you forget to call epoch_thread_init() for thread0 ?
Yes, this is my guess. I'm preparing patch for Cy to test
Author: glebius
Date: Wed Nov 14 00:33:03 2018
New Revision: 340423
URL: https://svnweb.freebsd.org/changeset/base/340423
Log:
Fix build on some architectures after r340413. On amd64 epoch.h
appeared to be included implicitly.
Modified:
head/sys/kern/kern_thread.c
Modified: head/sys/kern/k
Author: glebius
Date: Tue Nov 13 23:57:34 2018
New Revision: 340419
URL: https://svnweb.freebsd.org/changeset/base/340419
Log:
style(9), mostly adjusting overly long lines.
Modified:
head/sys/kern/subr_epoch.c
Modified: head/sys/kern/subr_epoch.c
=
Author: glebius
Date: Tue Nov 13 23:45:38 2018
New Revision: 340417
URL: https://svnweb.freebsd.org/changeset/base/340417
Log:
With epoch not inlined, there is no point in using _lite KPI. While here,
remove some unnecessary casts.
Modified:
head/sys/kern/genoffset.c
head/sys/kern/subr_ep
Author: glebius
Date: Tue Nov 13 23:44:27 2018
New Revision: 340416
URL: https://svnweb.freebsd.org/changeset/base/340416
Log:
Plug build break after r340411.
Modified:
head/sbin/dump/dump.h
head/sbin/dump/main.c
head/sbin/dump/traverse.c
Modified: head/sbin/dump/dump.h
=
Author: glebius
Date: Tue Nov 13 23:20:55 2018
New Revision: 340415
URL: https://svnweb.freebsd.org/changeset/base/340415
Log:
The dualism between epoch_tracker and epoch_thread is fragile and
unnecessary. So, expose CK types to kernel and use a single normal
structure for epoch_tracker.
Author: glebius
Date: Tue Nov 13 22:58:38 2018
New Revision: 340413
URL: https://svnweb.freebsd.org/changeset/base/340413
Log:
For compatibility KPI functions like if_addr_rlock() that used to have
mutexes but now are converted to epoch(9) use thread-private epoch_tracker.
Embedding tracker
Author: glebius
Date: Tue Nov 13 19:02:11 2018
New Revision: 340404
URL: https://svnweb.freebsd.org/changeset/base/340404
Log:
Uninline epoch(9) entrance and exit. There is no proof that modern
processors would benefit from avoiding a function call, but bloating
code. In fact, clang created
Author: glebius
Date: Wed Oct 31 23:05:44 2018
New Revision: 339976
URL: https://svnweb.freebsd.org/changeset/base/339976
Log:
MFhead r339643:
Fix ipw_start(), where logic was reverted in r287197.
PR: 232554
Modified:
stable/11/sys/dev/ipw/if_ipw.c
Directory Properties:
Author: glebius
Date: Wed Oct 31 23:02:59 2018
New Revision: 339975
URL: https://svnweb.freebsd.org/changeset/base/339975
Log:
MFhead r339643:
Fix ipw_start(), where logic was reverted in r287197.
PR232554
Approved by: re (kib)
Modified:
stable/12/sys/dev/ipw/if_ipw
Author: glebius
Date: Wed Oct 31 19:59:20 2018
New Revision: 339970
URL: https://svnweb.freebsd.org/changeset/base/339970
Log:
Remove unnecessary include from libstand.
Modified:
head/stand/libsa/ip.c
head/stand/libsa/net.c
head/stand/libsa/udp.c
Modified: head/stand/libsa/ip.c
=
Author: glebius
Date: Wed Oct 31 19:37:11 2018
New Revision: 339969
URL: https://svnweb.freebsd.org/changeset/base/339969
Log:
Define QMD_SAVELINK() only for QUEUE_MACRO_DEBUG_TRASH case. Otherwise
with QUEUE_MACRO_DEBUG_TRACE compilation fails due to unused variable.
Modified:
head/sys/sys
Author: glebius
Date: Wed Oct 31 18:01:02 2018
New Revision: 339963
URL: https://svnweb.freebsd.org/changeset/base/339963
Log:
MFhead r339596:
If we lost race or were migrated during bucket allocation for the per-CPU
cache, then we put new bucket on generic bucket cache. However, code
Author: glebius
Date: Mon Oct 29 22:10:52 2018
New Revision: 339889
URL: https://svnweb.freebsd.org/changeset/base/339889
Log:
Add a note that epoch(9) may change, to untie our hands for any future MFCs.
Modified:
head/share/man/man9/epoch.9
Modified: head/share/man/man9/epoch.9
Author: glebius
Date: Wed Oct 24 20:49:51 2018
New Revision: 339695
URL: https://svnweb.freebsd.org/changeset/base/339695
Log:
Remove code that is dead since r287197. Today wlan(4) interfaces aren't
children of some other interface. Creation happens only in wlan_up().
Modified:
head/libexec
Author: glebius
Date: Tue Oct 23 12:53:09 2018
New Revision: 339643
URL: https://svnweb.freebsd.org/changeset/base/339643
Log:
Fix ipw_start(), where logic was reverted in r287197.
PR: 232554
Submitted by: gl0...@mail.ru
Modified:
head/sys/dev/ipw/if_ipw.c
Modified: head/sys
Author: glebius
Date: Mon Oct 22 15:48:07 2018
New Revision: 339596
URL: https://svnweb.freebsd.org/changeset/base/339596
Log:
If we lost race or were migrated during bucket allocation for the per-CPU
cache, then we put new bucket on generic bucket cache. However, code didn't
honor UMA_ZONE_
Author: glebius
Date: Sun Oct 21 12:39:00 2018
New Revision: 339532
URL: https://svnweb.freebsd.org/changeset/base/339532
Log:
Fix exiting an epoch(9) we never entered. May happen only with MAC.
Modified:
head/sys/net/if_gif.c
head/sys/net/if_gre.c
head/sys/net/if_me.c
Modified: head/sys
Author: glebius
Date: Tue Oct 16 15:57:16 2018
New Revision: 339379
URL: https://svnweb.freebsd.org/changeset/base/339379
Log:
Plug sendfile(2) on a listening socket with proper error code.
Reported by: ngie
Reviewed by: ngie
Approved by: re (delphij)
Modified:
head/sys/kern/kern_
Author: glebius
Date: Mon Oct 15 20:20:57 2018
New Revision: 339370
URL: https://svnweb.freebsd.org/changeset/base/339370
Log:
Avoid OOB reads in memmem(3).
commit 51bdcdc424bd7169c8cccdc2de7cad17f5ea0f70
Author: Alexander Monakov
Date: Fri Jun 30 00:35:33 2017 +0300
fix OOB
s, up to 50%.
Here is script:
#!/usr/sbin/dtrace -s
fbt::cpu_set_syscall_retval:entry
{
@[args[1]] = count();
}
I would be interested if anybody reports results on a busy
web server running nginx.
So, I doubt that using __predict_true() is an optimisation here.
--
Gl
Author: glebius
Date: Wed Oct 3 17:40:04 2018
New Revision: 339170
URL: https://svnweb.freebsd.org/changeset/base/339170
Log:
In PR 227259, a user is reporting that they have code which is using
shutdown() to wakeup another thread blocked on a stream listen socket.
This code is failing, whi
Author: glebius
Date: Fri Aug 24 18:47:50 2018
New Revision: 338302
URL: https://svnweb.freebsd.org/changeset/base/338302
Log:
Either "free" or "allocated" is misleading here, since an item
in a bucket is free from perspective of UMA consumer, and it is
allocated from perspective of keg.
Author: glebius
Date: Thu Aug 23 23:24:28 2018
New Revision: 338284
URL: https://svnweb.freebsd.org/changeset/base/338284
Log:
Fix comment. The actual meaning of ub_cnt is the opposite.
Modified:
head/sys/vm/uma_int.h
Modified: head/sys/vm/uma_int.h
==
profile->depth = (u_int)TBR_UNSCALE(tbr->tbr_depth);
P> -}
P> -IFQ_UNLOCK(ifq);
P> -return (0);
P> -}
P> -
P> -/*
P> * attach a discipline to the interface. if one already exists, it is
P> * overridden.
P> * Locking is done in the discipline sp
On Tue, Aug 21, 2018 at 02:28:27PM -0700, Navdeep Parhar wrote:
N> On 8/21/18 2:26 PM, Gleb Smirnoff wrote:
N> > Navdeep,
N> >
N> > On Thu, Aug 16, 2018 at 11:46:38PM +, Navdeep Parhar wrote:
N> > N> Author: np
N> > N> Date: Thu Aug 16 23:46:38 2018
N
t
N> +vlan_pcp(struct ifnet *ifp, uint16_t *pcpp)
N> +{
N> +struct ifvlan *ifv;
N> +
N> +if (ifp->if_type != IFT_L2VLAN)
N> +return (EINVAL);
N> +ifv = ifp->if_softc;
N> +*pcpp = ifv->ifv_pcp;
N> +return (0);
N> +}
Is there
; if_freemulti(ll_ifma);
M> }
Coverity suggested there is a cut and paste mistake here, and it is compilable.
After quick glance I tend to agree. Looks like flag is cleared on wrong ifma.
--
Gleb Smirnoff
___
svn-src-all
E> int soopt_getm(struct sockopt *sopt, struct mbuf **mp);
E> int soopt_mcopyin(struct sockopt *sopt, struct mbuf *m);
E> int soopt_mcopyout(struct sockopt *sopt, struct mbuf *m);
This comment was a good mark to note that these three functions
need to go away. But IPv6 still uses them.
On Sat, Aug 04, 2018 at 10:48:17AM +0300, Konstantin Belousov wrote:
K> On Fri, Aug 03, 2018 at 03:27:39PM -0700, Gleb Smirnoff wrote:
K> > Hi Konstantin,
K> >
K> > On Fri, Jul 06, 2018 at 07:50:44PM +, Konstantin Belousov wrote:
K> > K> Author: kib
K> >
Author: glebius
Date: Sat Aug 4 00:03:21 2018
New Revision: 337279
URL: https://svnweb.freebsd.org/changeset/base/337279
Log:
Now that after r335979 the kernel addresses in API structures are
fixed size, there is no reason left for the unions.
Discussed with: brooks
Modified:
he
no
K> background school girl laughing: bob
For FreeBSD it would also make sense to bump hash sizes along with this change.
--
Gleb Smirnoff
___
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"
Author: glebius
Date: Fri Aug 3 22:55:58 2018
New Revision: 337278
URL: https://svnweb.freebsd.org/changeset/base/337278
Log:
Use if_tunnel_check_nesting() for ng_iface(4).
Modified:
head/sys/netgraph/ng_iface.c
head/sys/netgraph/ng_iface.h
Modified: head/sys/netgraph/ng_iface.c
=
rtual addresses.
Thanks for this change. After it the anonymous unions in xinpcb, unpxb
are no longer needed. Do you agree? I added them to make sizes on
32-bit and 64-bit ABIs to be identical. I prefer your new code,
but didn't the unions work? Anyway, now they can be removed.
--
Gleb Smirnoff
Since this has been broken already several times as people who edit
surrounding code don't understand the magic, and since now we round
the struct pcpu size to PAGE_SIZE anyways, what about stop carefully
craft the padding and just change alignment
ute path? e.g. kern.corefile=/var/log/cores/%N.core
Looks like the vn_fullpath_global needs to be fixed rather than problem
workarounded.
--
Gleb Smirnoff
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To u
Author: glebius
Date: Wed Jun 27 22:01:59 2018
New Revision: 335749
URL: https://svnweb.freebsd.org/changeset/base/335749
Log:
Check the inp_flags under inp lock. Looks like the race was hidden
before, the conversion of tcbinfo to CK_LIST have uncovered it.
Modified:
head/sys/netinet/tcp_su
Author: glebius
Date: Wed Jun 27 22:00:50 2018
New Revision: 335748
URL: https://svnweb.freebsd.org/changeset/base/335748
Log:
Correct r335242. Use unsigned cast instead of abs(). Using abs() gives
incorrect result when ticks has already wrapped, and are about to reach
the cr_ticks value (cr
Author: glebius
Date: Sat Jun 23 06:53:53 2018
New Revision: 335577
URL: https://svnweb.freebsd.org/changeset/base/335577
Log:
Revert r334843, and partially revert r335180.
tcp_outflags[] were defined since 4BSD and are defined nowadays in
all its descendants. Removing them breaks third p
Author: glebius
Date: Fri Jun 15 21:36:16 2018
New Revision: 335242
URL: https://svnweb.freebsd.org/changeset/base/335242
Log:
Since 'ticks' is an int, it may wrap around and cr_ticks at a certain
counter_rate will be greater than ticks, resulting in counter_ratecheck()
failure. To fix this
- non NULL controlp is not an error, returning EINVAL
M> would cause X forwarding to fail
Can you please provide reference for a code that uses
control messages with a stream socket?
The removed return was in action since FreeBSD 7.0 and
everything with X forwarding was work
Author: glebius
Date: Thu Jun 14 22:25:10 2018
New Revision: 335180
URL: https://svnweb.freebsd.org/changeset/base/335180
Log:
TCPOUTFLAGS no longer exists since r334843.
Modified:
head/sys/netinet/tcp_hpts.c
head/sys/netinet/tcp_offload.c
head/sys/netinet/tcp_output.c
head/sys/netinet/
han the implementation?
K>
K> Or 'sample frequency'
Interval definitely doesn't fit here. Frequency is closer, but still not the
right term, IMHO. Native speaker required here to judge. I am okay if anyone
who is confident changes wording here.
--
Gleb Smirnoff
;t yet fix the problem. This won't
happen if commit was reviewed.
--
Gleb Smirnoff
___
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"
is positive feedback like this which
M> keeps me going!
Btw, what was the point of checking flags || result? Most places in kernel
ignore flags and just test result regerdless of M_WAITOK/M_NOWAIT.
The result is already in a register, why do you think checking for absense
of M_WAITOK is faster that
t compilation time. If they
M> contain
M> M_WAITOK, the flag check will get elided along (we know for a fact it
M> passes)
M> and subsequently the NULL check will be short circuited, iow for known
M> M_WAITOK|M_ZERO flags this is:
M>
M> _malloc_item = malloc(_size, type, fla
Author: glebius
Date: Fri Jun 8 00:15:08 2018
New Revision: 334819
URL: https://svnweb.freebsd.org/changeset/base/334819
Log:
UMA memory debugging enabled with INVARIANTS consists of two things:
trashing freed memory and checking that allocated memory is properly
trashed, and also of keepin
Author: glebius
Date: Thu Jun 7 22:38:40 2018
New Revision: 334817
URL: https://svnweb.freebsd.org/changeset/base/334817
Log:
Add new functionality and syntax to cron(1) to allow to run jobs at a
given interval, which is counted in seconds since exit of the previous
invocation of the job. E
Author: glebius
Date: Tue May 29 21:45:15 2018
New Revision: 334356
URL: https://svnweb.freebsd.org/changeset/base/334356
Log:
Revert second chunk of r333860. The warning from gcc is false positive. The
npages won't be ever used in no space case.
Modified:
head/sys/kern/kern_sendfile.c
Mod
76b.
O> This brings us the renaming of fields in ck_queue, so that our own
O> LIST/SLIST/TAILQ/etc won't accidentally work with them.
Indeed, this broke kernel build :)
--
Gleb Smirnoff
___
svn-src-all@freebsd.org mailing list
https://li
ontrib/ because the change is already in contrib
for years. The vendor change came to vendor from us.
--
Gleb Smirnoff
___
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"
t; If you want to _actually_ be helpful fix these:
M> https://people.freebsd.org/~mmacy/gcc8logs/GENERIC-NODEBUG.log
M>
M> https://people.freebsd.org/~mmacy/gcc8logs/GENERIC.log
M>
M> On Wed, May 23, 2018 at 9:42 PM, Gleb Smirnoff wrote:
M> > Let me repeat again. The warning is a
On Wed, May 23, 2018 at 10:13:25PM -0700, Matthew Macy wrote:
M> On Wed, May 23, 2018 at 10:07 PM, Gleb Smirnoff wrote:
M> > Can you please explain the bug supposed to be fixed by r333860 QUESTION
MARK
M>
M> Did I say it fixed a bug? Or are you saying we should just turn
ash _later_ than production kernels,
since instead of uma_junk it places clean zeroes.
May be changes like that deserve more than a 30 minute time frame for review?
--
Gleb Smirnoff
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/m
d be added with
a comment /* pacify gcc */.
On Wed, May 23, 2018 at 03:59:33PM -0700, Matthew Macy wrote:
M> On Wed, May 23, 2018 at 3:57 PM, Gleb Smirnoff wrote:
M> > The initialization isn't useful.
M>
M> It silences a gcc warning. So yes it is. It's this exchange which
401 - 500 of 2945 matches
Mail list logo