Author: glebius
Date: Tue Dec 8 23:54:09 2020
New Revision: 368468
URL: https://svnweb.freebsd.org/changeset/base/368468
Log:
Fixup r368446 with KERN_TLS.
Modified:
head/sys/net/if_lagg.c
Modified: head/sys/net/if_lagg.c
==
Author: glebius
Date: Tue Dec 8 16:46:00 2020
New Revision: 368448
URL: https://svnweb.freebsd.org/changeset/base/368448
Log:
The list of ports in configuration path shall be protected by locks,
epoch shall be used only for fast path. Thus use LAGG_XLOCK() in
lagg_[un]register_vlan. This
Author: glebius
Date: Tue Dec 8 16:36:46 2020
New Revision: 368446
URL: https://svnweb.freebsd.org/changeset/base/368446
Log:
Convert LAGG_RLOCK() to NET_EPOCH_ENTER(). No functional changes.
Modified:
head/sys/net/if_lagg.c
Modified: head/sys/net/if_lagg.c
=
Author: glebius
Date: Wed Dec 2 21:53:28 2020
New Revision: 368287
URL: https://svnweb.freebsd.org/changeset/base/368287
Log:
Fix r368197: suppress error printing for the "check" command.
Reviewed by: kevans
Modified:
head/sbin/bectl/bectl.c
Modified: head/sbin/bectl/bectl.c
=
root != NULL ? root : "");
+ if (!cmd->silent)
+ fprintf(stderr, "libbe_init(\"%s\") failed.\n",
+ root != NULL ? root : "");
return (-1);
}
--
Gleb Smirnoff
___
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"
Failure of the library is something different to failed
check of current system. It is permanent failure, meaning that command
is being with incorrect root argument or ZFS is missing at all. Pretty
much the same as using bectl with incorrect arguments or options.
--
Gleb Smirnoff
___
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"
Author: glebius
Date: Mon Nov 30 21:05:31 2020
New Revision: 368197
URL: https://svnweb.freebsd.org/changeset/base/368197
Log:
Print at least something when failing.
Modified:
head/sbin/bectl/bectl.c
Modified: head/sbin/bectl/bectl.c
==
Author: glebius
Date: Thu Nov 19 02:20:38 2020
New Revision: 367824
URL: https://svnweb.freebsd.org/changeset/base/367824
Log:
Add '-u' switch that would uncompress cores that were compressed by
kernel during dump time.
A real life scenario is that cores are compressed to reduce
size of
Author: glebius
Date: Tue Nov 3 22:04:32 2020
New Revision: 367307
URL: https://svnweb.freebsd.org/changeset/base/367307
Log:
Style, not functional changes:
- Improve spelling of a false check [1]
- A missing line from r367150.
Submitted by: kib
Modified:
head/sbin/savecore/savecore
Author: glebius
Date: Thu Oct 29 23:15:11 2020
New Revision: 367150
URL: https://svnweb.freebsd.org/changeset/base/367150
Log:
Convert flags from int to bool. Some (compress) were already used in
comparisons with bool values. No functional changes.
Modified:
head/sbin/savecore/savecore.c
Author: glebius
Date: Thu Oct 22 18:00:07 2020
New Revision: 366945
URL: https://svnweb.freebsd.org/changeset/base/366945
Log:
Fix typo
Modified:
head/share/man/man9/pfil.9
Modified: head/share/man/man9/pfil.9
==
---
and would
K> probably allow us to drop a __FreeBSD_version split of dependencies in
K> the short-to-mid term from OpenZFS if we don't need to pretend that
K> OpenZFS will build on 11. None of it looked too invasive for stable
K> from a cursory glance.
My apologi
r tree before doing a merge and
there will be no conflict.
Example commit: https://svnweb.freebsd.org/base?view=revision&revision=360579
--
Gleb Smirnoff
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-
Author: glebius
Date: Wed Sep 9 16:13:33 2020
New Revision: 365504
URL: https://svnweb.freebsd.org/changeset/base/365504
Log:
In r354148 the goal was to check THREAD_CAN_SLEEP() only once for the
purpose of epoch_trace() and for calling subsequent panic, but to keep
code fully under INVARIA
On Fri, Aug 28, 2020 at 02:31:30PM -0700, Oleksandr Tymoshenko wrote:
O> Gleb Smirnoff (gleb...@freebsd.org) wrote:
O> > Author: glebius
O> > Date: Thu Oct 10 23:42:55 2019
O> > New Revision: 353419
O> > URL: https://svnweb.freebsd.org/changeset/base/353419
O> >
Author: glebius
Date: Sun Aug 30 17:13:04 2020
New Revision: 364975
URL: https://svnweb.freebsd.org/changeset/base/364975
Log:
Followup on r364922. Old comment said that the only reason to put
the hook at queue mode was that mn_rx_intr() doesn't run at splnet
level. In today's netgraph the o
Author: glebius
Date: Thu Aug 20 20:31:47 2020
New Revision: 364441
URL: https://svnweb.freebsd.org/changeset/base/364441
Log:
When we have a command returned by zfs_nextboot() that is longer
than command in the loader.conf, the latter needs to be nul terminated,
otherwise garbage trailer le
Author: glebius
Date: Mon Aug 17 15:37:08 2020
New Revision: 364310
URL: https://svnweb.freebsd.org/changeset/base/364310
Log:
With INVARIANTS panic immediately if M_WAITOK is requested in a
non-sleepable context. Previously only _sleep() would panic.
This will catch misuse of M_WAITOK at d
Author: glebius
Date: Tue Aug 11 15:46:22 2020
New Revision: 364117
URL: https://svnweb.freebsd.org/changeset/base/364117
Log:
ipfw: make the "frag" keyword accept additional options "mf",
"df", "rf" and "offset". This allows to match on specific
bits of ip_off field.
For compatibility
(&inm_free_task, 0, inm_release_task, NULL);
-}
-SYSINIT(inm_init, SI_SUB_TASKQ, SI_ORDER_ANY, inm_init, NULL);
-
void
inm_release_wait(void *arg __unused)
{
Same for inm6.
--
Gleb Smirnoff
___
svn-src-head@freebsd.org mailing list
https://li
Author: glebius
Date: Tue Jul 7 02:43:53 2020
New Revision: 362982
URL: https://svnweb.freebsd.org/changeset/base/362982
Log:
Fixup r362981: remove gzipped manual pages.
Pointy hat to:glebius
Modified:
head/ObsoleteFiles.inc
Modified: head/ObsoleteFiles.inc
Author: glebius
Date: Tue Jul 7 02:41:51 2020
New Revision: 362981
URL: https://svnweb.freebsd.org/changeset/base/362981
Log:
Fixup for r360574: install new mlinks for sglist(9) and remove old ones.
Modified:
head/ObsoleteFiles.inc
head/share/man/man9/Makefile
Modified: head/ObsoleteFiles
On Tue, Jun 23, 2020 at 01:20:01AM +0300, Yuri Pankov wrote:
Y> Gleb Smirnoff wrote:
Y> >Yuri, Zhihao,
Y> >
Y> > this commit totally broke Russian input for me in nvi. After
Y> > exiting edit mode, nvi immediately converts all text to ???.
Y> >
Y> &g
et(), 0);
Y> +else
Y> +o_set(sp, O_FILEENCODING, OS_STRDUP, "iso8859-1", 0);
Y>
Y> conv_enc(sp, O_FILEENCODING, 0);
Y> #endif
Y> ___
Y> svn-src-...@freebsd.org mailing list
Y> https://lists.freebsd.or
Author: glebius
Date: Mon Jun 8 17:40:39 2020
New Revision: 361931
URL: https://svnweb.freebsd.org/changeset/base/361931
Log:
Move MPASS() macros to systm.h. They are widely used all over
the kernel and aren't contained only to the locking code.
Reviewed by: kib, mjg
Differential Rev
Author: glebius
Date: Sun May 3 00:37:16 2020
New Revision: 360583
URL: https://svnweb.freebsd.org/changeset/base/360583
Log:
Step 4.2: start divorce of M_EXT and M_EXTPG
They have more differencies than similarities. For now there is lots
of code that would check for M_EXT only and work
Author: glebius
Date: Sun May 3 00:27:41 2020
New Revision: 360582
URL: https://svnweb.freebsd.org/changeset/base/360582
Log:
Mechanically rename MBUF_EXT_PGS_ASSERT() to M_ASSERTEXTPG() to match
classical M_ASSERTPKTHDR.
Reviewed by: gallatin
Differential Revision:https://rev
Author: glebius
Date: Sun May 3 00:21:11 2020
New Revision: 360581
URL: https://svnweb.freebsd.org/changeset/base/360581
Log:
Step 4.1: mechanically rename M_NOMAP to M_EXTPG
Reviewed by: gallatin
Differential Revision:https://reviews.freebsd.org/D24598
Modified:
head/sys/dev
Author: glebius
Date: Sun May 3 00:12:56 2020
New Revision: 360579
URL: https://svnweb.freebsd.org/changeset/base/360579
Log:
Step 3: anonymize struct mbuf_ext_pgs and move all its fields into mbuf
within m_epg namespace.
All edits except the 'struct mbuf' declaration and mb_dupcl()
Author: glebius
Date: Sun May 3 00:08:05 2020
New Revision: 360578
URL: https://svnweb.freebsd.org/changeset/base/360578
Log:
Step 2.5: Stop using 'struct mbuf_ext_pgs' in the kernel itself.
Reviewed by: gallatin
Differential Revision:https://reviews.freebsd.org/D24598
Modified
Author: glebius
Date: Sun May 3 00:03:39 2020
New Revision: 360577
URL: https://svnweb.freebsd.org/changeset/base/360577
Log:
Make MBUF_EXT_PGS_ASSERT_SANITY() a macro, so that it prints file:line.
While here, stop using struct mbuf_ext_pgs.
Reviewed by: gallatin
Differential Revision
Author: glebius
Date: Sat May 2 23:58:20 2020
New Revision: 360576
URL: https://svnweb.freebsd.org/changeset/base/360576
Log:
Step 2.4: Stop using 'struct mbuf_ext_pgs' in drivers.
Reviewed by: gallatin, hselasky
Differential Revision:https://reviews.freebsd.org/D24598
Modified
Author: glebius
Date: Sat May 2 23:52:35 2020
New Revision: 360575
URL: https://svnweb.freebsd.org/changeset/base/360575
Log:
Step 2.3: Rename mbuf_ext_pg_len() to m_epg_pagelen() that
uses mbuf argument.
Reviewed by: gallatin
Differential Revision:https://reviews.fr
Author: glebius
Date: Sat May 2 23:46:29 2020
New Revision: 360574
URL: https://svnweb.freebsd.org/changeset/base/360574
Log:
Step 2.2:
o Shrink sglist(9) functions to work with multipage mbufs down from
four functions to two.
o Don't use 'struct mbuf_ext_pgs *' as argument, use struct
Author: glebius
Date: Sat May 2 23:38:13 2020
New Revision: 360573
URL: https://svnweb.freebsd.org/changeset/base/360573
Log:
Step 2.1: Build TLS workqueue from mbufs, not struct mbuf_ext_pgs.
Reviewed by: gallatin
Differential Revision:https://reviews.freebsd.org/D24598
Modifi
Author: glebius
Date: Sat May 2 22:56:22 2020
New Revision: 360572
URL: https://svnweb.freebsd.org/changeset/base/360572
Log:
Get rid of the mbuf self-pointing pointer.
Reviewed by: gallatin
Differential Revision:https://reviews.freebsd.org/D24598
Modified:
head/sys/kern/uipc
Author: glebius
Date: Sat May 2 22:49:14 2020
New Revision: 360571
URL: https://svnweb.freebsd.org/changeset/base/360571
Log:
Start moving into EPG_/epg_ namespace. There is only one flag, but
next commit brings in second flag, so let them already be in the
future namespace.
Reviewed
Author: glebius
Date: Sat May 2 22:44:23 2020
New Revision: 360570
URL: https://svnweb.freebsd.org/changeset/base/360570
Log:
In mb_unmapped_compress() we don't need mbuf structure to keep data,
but we need buffer of MLEN bytes. This isn't just a simplification,
but important fixup, becaus
Author: glebius
Date: Sat May 2 22:39:26 2020
New Revision: 360569
URL: https://svnweb.freebsd.org/changeset/base/360569
Log:
Continuation of multi page mbuf redesign from r359919.
The following series of patches addresses three things:
Now that array of pages is embedded into mbuf, w
Author: glebius
Date: Mon Apr 20 23:32:49 2020
New Revision: 360138
URL: https://svnweb.freebsd.org/changeset/base/360138
Log:
Fix immediate crash when snmpd is bound to a specific IP address.
The code that sets up msghdr must first fully fill in the msghdr
itself, and only then use CMSG_
Author: glebius
Date: Fri Apr 17 06:05:08 2020
New Revision: 360037
URL: https://svnweb.freebsd.org/changeset/base/360037
Log:
Make ZFS depend on xdr.ko only. It doesn't need kernel RPC.
Differential Revision:https://reviews.freebsd.org/D24408
Modified:
head/sys/cddl/contrib/ope
Author: glebius
Date: Fri Apr 17 06:04:20 2020
New Revision: 360036
URL: https://svnweb.freebsd.org/changeset/base/360036
Log:
Split XDR into separate kernel module. Make krpc depend on xdr.
Reviewed by: rmacklem
Differential Revision:https://reviews.freebsd.org/D24408
Added:
Author: glebius
Date: Fri Apr 17 06:02:13 2020
New Revision: 360035
URL: https://svnweb.freebsd.org/changeset/base/360035
Log:
Move M_RPC malloc type into XDR. Both RPC and XDR libraries use
this type, but since RPC depends on XDR (not vice versa) we need
it defined in XDR to make the modul
Author: glebius
Date: Fri Apr 17 05:59:38 2020
New Revision: 360034
URL: https://svnweb.freebsd.org/changeset/base/360034
Log:
Don't initialize m->m_data to m->m_pktdat, this is already done by the
mbuf allocator. That was the last remnant of such code in the kernel.
Modified:
head/sys/net
'make delete-old' times, and
trimming the default file definitely makes sense. However, what about
keeping the full ObsoleteFiles.inc version, at least for documenting
purposes?
--
Gleb Smirnoff
___
svn-src-head@freebsd.org mailing list
https:/
tls = NULL;
IMHO, such changes must always be accompanied by a comment. Otherwise, I
can easily imagine someone in couple of years "optimizing" it back with
commit message "Remove extraneous check. m->m_ext.ext_pgs->tls is NULL when tls
is off"
--
Gleb Smirnoff
er reach EOF, or have enough data. In that
D> case, a buffer with the size of if_msghdr is also enough, since you
D> never need to read beyond that.
Sorry for delayed answer. The routing socket is a datagram socket,
it isn't like TCP, it can't deliver par
Author: glebius
Date: Thu Mar 5 21:01:47 2020
New Revision: 358686
URL: https://svnweb.freebsd.org/changeset/base/358686
Log:
Align the buffer to the alignment of the structure we expect.
Submitted by: Slawa Olhovchenkov
Modified:
head/sbin/mount_nfs/mount_nfs.c
Modified: head/sbin/mo
ning.
S> > >
S> > > Thanks, Slawa! I think this is the most elegant solution.
S> >
S> > Why don't just declare the buffer as:
S> >
S> > struct if_msghdr buf;
S> >
S> > and then do:
S> >
S> > nread = read(s, &buf, si
t; just through (void *) cast?
S> >
S> > Copy to aligned buffer or got SIGBUS on some architectures?
S>
S> char buf[2048] __aligned(__alignof(struct if_msghdr));
S>
S> resolve this watning.
Thanks, Slawa! I think this is the most elegant solution.
--
Gleb Smirnoff
On Thu, Mar 05, 2020 at 03:29:23PM +0100, Dimitry Andric wrote:
D> On 2020-03-04 23:27, Gleb Smirnoff wrote:
D> > Author: glebius
D> > Date: Wed Mar 4 22:27:16 2020
D> > New Revision: 358655
D> > URL: https://svnweb.freebsd.org/changeset/base/358655
D> >
D>
ys/vm/vm_page.h
J>
==
J> --- head/sys/vm/vm_page.hFri Feb 28 21:31:40 2020(r358450)
J> +++ head/sys/vm/vm_page.hFri Feb 28 21:42:48 2020(r358451)
J> @@ -653,6 +653,7 @@ void vm_page_reference(vm_page_t m);
J> #define VPR_NOREUSE 0x02
J> void vm_page_release(vm_page_t m, int flags);
J> void vm_page_release_locked(vm_page_t m, int flags);
J> +vm_page_t vm_page_relookup(vm_object_t, vm_pindex_t);
J> bool vm_page_remove(vm_page_t);
J> bool vm_page_remove_xbusy(vm_page_t);
J> int vm_page_rename(vm_page_t, vm_object_t, vm_pindex_t);
J> ___
J> svn-src-...@freebsd.org mailing list
J> https://lists.freebsd.org/mailman/listinfo/svn-src-all
J> To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
--
Gleb Smirnoff
___
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"
Author: glebius
Date: Wed Mar 4 23:49:20 2020
New Revision: 358658
URL: https://svnweb.freebsd.org/changeset/base/358658
Log:
Add a missing bktr header.
Modified:
head/ObsoleteFiles.inc
Modified: head/ObsoleteFiles.inc
Author: glebius
Date: Wed Mar 4 22:32:40 2020
New Revision: 358657
URL: https://svnweb.freebsd.org/changeset/base/358657
Log:
Fix spelling of "dropped".
Submitted by: Lutz Donnerhacke
Differential Revision:https://reviews.freebsd.org/D23954
Modified:
head/share/man/man4/ng_car
Author: glebius
Date: Wed Mar 4 22:31:41 2020
New Revision: 358656
URL: https://svnweb.freebsd.org/changeset/base/358656
Log:
Remove unused function.
Modified:
head/sys/dev/wtap/if_wtap.c
head/sys/dev/wtap/if_wtapvar.h
Modified: head/sys/dev/wtap/if_wtap.c
Author: glebius
Date: Wed Mar 4 22:27:16 2020
New Revision: 358655
URL: https://svnweb.freebsd.org/changeset/base/358655
Log:
When a machine boots the NFS mounting script is executed after
interfaces are configured, but for many interfaces (e.g. all Intel)
ifconfig causes link renegotiation
gus pages in the middle of a request. We never substitute
first or last.
--
Gleb Smirnoff
___
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"
Author: glebius
Date: Tue Feb 25 19:29:05 2020
New Revision: 358320
URL: https://svnweb.freebsd.org/changeset/base/358320
Log:
Generalize resources freeing in sendfile with different scenarios.
Now we execute sendfile_iodone() in all possible cases, which
guarantees that vm_object_pip_wakeup
Author: glebius
Date: Tue Feb 25 19:26:40 2020
New Revision: 358319
URL: https://svnweb.freebsd.org/changeset/base/358319
Log:
Make ktls_frame() never fail. Caller must supply correct mbufs.
This makes sendfile code a bit simplier.
Modified:
head/sys/kern/kern_sendfile.c
head/sys/kern/ui
Author: glebius
Date: Tue Feb 25 19:11:20 2020
New Revision: 358317
URL: https://svnweb.freebsd.org/changeset/base/358317
Log:
When sendfile_swapin() sweeps through pages in search for a bogus page
skip first and last pages. This is a micro optimisation.
Modified:
head/sys/kern/kern_sendfi
Author: glebius
Date: Mon Feb 24 21:07:30 2020
New Revision: 358301
URL: https://svnweb.freebsd.org/changeset/base/358301
Log:
Although most of the NIC drivers are epoch ready, due to peer pressure
switch over to opt-in instead of opt-out for epoch.
Instead of IFF_NEEDSEPOCH, provide IFF_
Author: glebius
Date: Fri Feb 21 04:18:15 2020
New Revision: 358194
URL: https://svnweb.freebsd.org/changeset/base/358194
Log:
Rework second part of r357558. Unroll the macro and allocate memory in
sleepable manner before entering the epoch for the send.
Modified:
head/sys/netgraph/ng_sock
Author: glebius
Date: Fri Feb 21 04:10:41 2020
New Revision: 358193
URL: https://svnweb.freebsd.org/changeset/base/358193
Log:
Revert one half of previous change r357558. Don't enter the epoch on
sends to control socket. Control socket messages can run constructors
of nodes and other stuff
ontext() were clearly
designed to avoid dereferencing an ifnet pointer after a packet has been
queued and dequeued on IGMP/MLD internal queue.
This patch now replicates the exactly same problem but with netisr
queue. Of course netisr not always queues, someti
mov0x4f02fd(%rip),%rax# 0x80c01788
M> addq $0x1,%gs:(%rax)
M>
M> Reviewed by: jeff
M> Differential Revision: https://reviews.freebsd.org/D23570
Neat optimization! Thanks. Why didn't we do it back when created counter?
--
Gleb Smirnoff
__
Author: glebius
Date: Tue Feb 11 20:59:41 2020
New Revision: 357779
URL: https://svnweb.freebsd.org/changeset/base/357779
Log:
Remove assertion from TASK_INIT() macro, since some users of
sys/taskqueue.h may not have includes that define MPASS. It
was useful during testing of r357771, but ca
bool, but we should use it in any new code.
Some people may still drop ints into new code, and this can be
tolerable, but intentionally converting from bools to ints is a
move backwards and this isn't okay.
Thanks for redoing it with via ifdefs for legacy platf
Author: glebius
Date: Tue Feb 11 19:13:34 2020
New Revision: 357773
URL: https://svnweb.freebsd.org/changeset/base/357773
Log:
Mark lio taskqueue as requiring network epoch.
Modified:
head/sys/dev/liquidio/base/lio_droq.c
Modified: head/sys/dev/liquidio/base/lio_droq.c
==
Author: glebius
Date: Tue Feb 11 18:57:07 2020
New Revision: 357772
URL: https://svnweb.freebsd.org/changeset/base/357772
Log:
Use NET_TASK_INIT() and NET_GROUPTASK_INIT() for drivers that process
incoming packets in taskqueue context.
Reviewed by: hselasky
Differential Revision:
Author: glebius
Date: Tue Feb 11 18:48:07 2020
New Revision: 357771
URL: https://svnweb.freebsd.org/changeset/base/357771
Log:
Add flag to struct task to mark the task as requiring network epoch.
When processing a taskqueue and a task has associated epoch, then
enter for duration of the t
re fine with getting a value from a different CPU.
However, can't imagine a situation where migrating during a replace
operation is acceptable.
--
Gleb Smirnoff
___
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"
Author: glebius
Date: Wed Feb 5 03:07:20 2020
New Revision: 357559
URL: https://svnweb.freebsd.org/changeset/base/357559
Log:
I doubt anybody in the world uses ng_device, but its write method should
also enter the network epoch when sending data from user level to netgraph.
Modified:
head/
Author: glebius
Date: Wed Feb 5 03:06:29 2020
New Revision: 357558
URL: https://svnweb.freebsd.org/changeset/base/357558
Log:
Enter the network epoch when ng_socket sends data or control from user
land to the netgraph and potentially further down the network stack.
Modified:
head/sys/netgr
Author: glebius
Date: Wed Feb 5 02:53:40 2020
New Revision: 357557
URL: https://svnweb.freebsd.org/changeset/base/357557
Log:
netgraph(4) callouts need to be executed in the network epoch.
Modified:
head/sys/netgraph/ng_base.c
Modified: head/sys/netgraph/ng_base.c
==
ne, don't we? I
mean we don't expect 3rd party device drivers to use it. So may be
it is fine that if_clone.h exposes those functions? If no, then
we probably should hide both if_clone_addif and if_clone_destroyif
away from if_clone.h and declare them as extern in epair and net80211.
--
G
Author: glebius
Date: Mon Feb 3 20:48:57 2020
New Revision: 357466
URL: https://svnweb.freebsd.org/changeset/base/357466
Log:
Couple protocol drain routines (frag6_drain and sctp_drain) may send
packets. An unexpected behaviour for memory reclamation routine.
Anyway, we need enter the netw
On Thu, Jan 30, 2020 at 10:51:23AM -1000, Jeff Roberson wrote:
J> On Thu, 30 Jan 2020, Gleb Smirnoff wrote:
J>
J> > On Thu, Jan 30, 2020 at 10:28:01AM +, Hans Petter Selasky wrote:
J> > H> Author: hselasky
J> > H> Date: Thu Jan 30 10:28:01 2020
J> > H>
t;
H> This issue was introduced by r357004. This is a preparation step for
H> replacing the functionality provided by r357004.
H>
H> Found by: kib@
H> Sponsored by: Mellanox Technologies
What?! NETISR_RLOCK() which in turn ends up calling msleep()? Ca
e fixed by D23408:
bwn, malo, mwl, wtap
P.S. A funny note about wtap. You modified even a function that
is a dead code - wtap_rx_deliver(). Gives some clue on quality
of your sweep over all drivers.
--
Gleb Smirnoff
___
svn-src-head@freebsd.org mailing
>
H> Pointy hat:glebius@
H> Sponsored by: Mellanox Technologies
Just looking at the very first driver in the patch - bwi.
Why do you call it "broken"? It doesn't need any "unbreaking after r357004".
Can you please show me a callgraph where epoch_
truct ifnet *, struct if_clone *);
K>
K> -/* The below interface used only by epair(4). */
K> +/* The below interfaces are used only by epair(4). */
K> +voidif_clone_addif(struct if_clone *, struct ifnet *);
K> int if_clone_destroyif(struct if_clone *, struct ifnet *
Author: glebius
Date: Thu Jan 30 00:18:00 2020
New Revision: 357279
URL: https://svnweb.freebsd.org/changeset/base/357279
Log:
Fix text format definition for kern.maxvnodes, vfs.wantfreevnodes. This
is a regression from r356642, r356645.
Modified:
head/sys/kern/vfs_subr.c
Modified: head/s
Author: glebius
Date: Wed Jan 29 22:48:18 2020
New Revision: 357276
URL: https://svnweb.freebsd.org/changeset/base/357276
Log:
Fix missing NET_EPOCH_ENTER() when compiled with TCP_OFFLOAD.
Reported by: Coverity
CID: 1413162
Modified:
head/sys/netinet/tcp_usrreq.c
Modified: h
nice features of FreeBSD kernel
that nginx supports?
Don't read me as if I'm against adding TCP_CORK support to linux(4).
--
Gleb Smirnoff
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubs
the garbage collection can be batched at
level of UMA. Jeff is working on that.
2) Use separate epoch for PCBs, leaving the network epoch for long
lived structures only. Don't hold the PCB epoch long.
3) Just don't use epoch for PCBs. Dec
workload or configuration?
Can you imagine a workload where calling epoch_enter/exit more often but
for a shorter period of execution would be beneficial that calling it
once for a packet?
--
Gleb Smirnoff
___
svn-src-head@freebsd.org mailing list
http
OCH's. There might
H> be realtime constraints for EPOCH's aswell.
Epoch isn't a lock, so it can't become the new Giant.
--
Gleb Smirnoff
___
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"
Author: glebius
Date: Sat Jan 25 00:06:18 2020
New Revision: 357102
URL: https://svnweb.freebsd.org/changeset/base/357102
Log:
Enter the network epoch in RX processing taskqueue.
Modified:
head/sys/dev/mlx4/mlx4_en/mlx4_en_rx.c
Modified: head/sys/dev/mlx4/mlx4_en/mlx4_en_rx.c
===
Author: glebius
Date: Fri Jan 24 21:56:10 2020
New Revision: 357094
URL: https://svnweb.freebsd.org/changeset/base/357094
Log:
Enter the network epoch when rack_output() is called in setsockopt(2).
Modified:
head/sys/netinet/tcp_stacks/rack.c
Modified: head/sys/netinet/tcp_stacks/rack.c
Author: glebius
Date: Fri Jan 24 21:04:33 2020
New Revision: 357093
URL: https://svnweb.freebsd.org/changeset/base/357093
Log:
Enter the network epoch in USB WiFi drivers when processing input
mbuf queues.
Submitted by: Idwer Vollering
Modified:
head/sys/dev/otus/if_otus.c
head/sys/
al changes.
Just out of curiosity: what Linux binaries that do unmount may
somebody run on FreeBSD for production or just for fun?
--
Gleb Smirnoff
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To
Author: glebius
Date: Fri Jan 24 17:24:02 2020
New Revision: 357090
URL: https://svnweb.freebsd.org/changeset/base/357090
Log:
re(4) uses taskqueue to process input packets. Enter network epoch
in there.
Modified:
head/sys/dev/re/if_re.c
Modified: head/sys/dev/re/if_re.c
=
Author: glebius
Date: Fri Jan 24 17:11:54 2020
New Revision: 357088
URL: https://svnweb.freebsd.org/changeset/base/357088
Log:
ath(4) processing input packets in taskqueue. Enter network epoch
before calling ieee80211_input_mimo().
Modified:
head/sys/dev/ath/if_ath_rx.c
Modified: head/sys
interrupt
handlers. Complexity and performance impact of using a pointer are
noted by Drew in D23347.
--
Gleb Smirnoff
___
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"
gt; link in all of them.
I probably now on will also include in every patch in a serie. I just
dislike that first one (usually a preparation change) closes the review.
--
Gleb Smirnoff
___
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/m
preemption, etc. this
J> could be 100s of ms of PCBs held.
We also are concerned about that theoretically. Haven't yet seen effect
in practice, but our sessions are mostly longer living. First we have the
tunable to limit batching. Second, there are some ideas on how to improve
the garbage
uxkpi/common/include/linux/interrupt.h?revision=352205&view=markup#l103
Well, it is not my fault that a video driver requests INTR_TYPE_NET
interrupt. I mean, you can't put this as a rationale against using
network epoch for all interrrupts that declare theirselves as network
interrupts. H
On Thu, Jan 23, 2020 at 08:17:46PM -0500, Ryan Stone wrote:
R> On Thu, Jan 23, 2020 at 6:05 PM Gleb Smirnoff wrote:
R> >
R> > On Thu, Jan 23, 2020 at 02:17:33PM -0500, Ryan Stone wrote:
R> > R> What is a driver's responsibility now for entering/leaving the net
epo
On Thu, Jan 23, 2020 at 08:17:46PM -0500, Ryan Stone wrote:
R> On Thu, Jan 23, 2020 at 6:05 PM Gleb Smirnoff wrote:
R> >
R> > On Thu, Jan 23, 2020 at 02:17:33PM -0500, Ryan Stone wrote:
R> > R> What is a driver's responsibility now for entering/leaving the net
epo
On Thu, Jan 23, 2020 at 04:08:32PM -0700, Ian Lepore wrote:
I> On Thu, 2020-01-23 at 15:05 -0800, Gleb Smirnoff wrote:
I> > On Thu, Jan 23, 2020 at 02:17:33PM -0500, Ryan Stone wrote:
I> > R> What is a driver's responsibility now for entering/leaving the net
epoch now?
I>
1 - 100 of 2552 matches
Mail list logo