Re: svn commit: r322272 - head/sys/compat/linuxkpi/common/src

On Tue, Aug 8, 2017 at 3:36 PM Alexander Motin wrote: > > Author: mav > Date: Tue Aug 8 19:36:34 2017 > New Revision: 322272 > URL: https://svnweb.freebsd.org/changeset/base/322272 > > Log: > Fix few issues of LinuxKPI workqueue. > > LinuxKPI workqueue wrappers reported "successful"

Re: svn commit: r365475 - head/sys/net/route

On Tue, Sep 8, 2020 at 5:39 PM Alexander V. Chernikov wrote: > -CTLFLAG_VNET | CTLTYPE_U32 | CTLFLAG_RWTUN | CTLFLAG_MPSAFE, NULL, 0, > +CTLFLAG_VNET | CTLTYPE_U32 | CTLFLAG_MPSAFE, NULL, 0, Don't you still need CTLFLAG_RW? I don't believe that the sysctl is usable at all without it.

Re: svn commit: r360849 - in head: share/man/man9 sys/kern

On Sat, May 9, 2020 at 11:56 AM Ed Maste wrote: > > Author: emaste > Date: Sat May 9 15:56:02 2020 > New Revision: 360849 > URL: https://svnweb.freebsd.org/changeset/base/360849 > > Log: > remove %n support from printf(9) Should we put a KASSERT in to catch if new code tries to use a %n flag

Re: svn commit: r357051 - head/sys/dev/bge

On Thu, Jan 23, 2020 at 8:25 PM Gleb Smirnoff wrote: > Because at interrupt level we can batch multiple packets in a single epoch. > This speeds up unfiltered packet forwarding performance by 5%. > > With driver level pfil hooks I would claim even more improvement, because > before > the change

Re: svn commit: r357051 - head/sys/dev/bge

On Thu, Jan 23, 2020 at 6:05 PM Gleb Smirnoff wrote: > > On Thu, Jan 23, 2020 at 02:17:33PM -0500, Ryan Stone wrote: > R> What is a driver's responsibility now for entering/leaving the net epoch > now? > > For drivers that are 'special', entering the net epoch is necessa

Re: svn commit: r357051 - head/sys/dev/bge

What is a driver's responsibility now for entering/leaving the net epoch now? ___ svn-src-all@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"

svn commit: r356444 - head/libexec/rtld-elf

Author: rstone Date: Tue Jan 7 16:03:11 2020 New Revision: 356444 URL: https://svnweb.freebsd.org/changeset/base/356444 Log: rtld: Fix segfault in direct exec mode When rtld is directly executed with arguments, it has to move the program arguments, environment and elf aux data up a few

svn commit: r353938 - head/sys/compat/linuxkpi/common/src

Author: rstone Date: Wed Oct 23 17:20:20 2019 New Revision: 353938 URL: https://svnweb.freebsd.org/changeset/base/353938 Log: Add missing M_NOWAIT flag The LinuxKPI linux_dma code calls PCTRIE_INSERT with a mutex held, but does not set M_NOWAIT when allocating nodes, leading to a

svn commit: r352926 - head/sys/conf

Author: rstone Date: Tue Oct 1 13:36:01 2019 New Revision: 352926 URL: https://svnweb.freebsd.org/changeset/base/352926 Log: Clean up duplicate entries in sys/conf/files Reviewed by: imp Modified: head/sys/conf/files Modified: head/sys/conf/files

Re: svn commit: r352778 - in head/sys: conf sys

Is this enough? Won't out-of-tree modules get compiled without EPOCH_TRACE set? If such a module is loaded on a kernel with EPOCH_TRACE set then the module will call epoch_enter_preempt() with a epoch_tracker that is too small and have its stack corrupted, won't it? On Thu, Sep 26, 2019 at 5:12

Re: svn commit: r352707 - in head/sys: conf kern net sys

We also shouldn't have ifdef's that change the size of a kernel structure that's part of the KBI. Isn't struct epoch_tracker part of the KBI? On Thu, Sep 26, 2019 at 12:46 PM Warner Losh wrote: > > But we shouldn't be including opt_foo.h files in sys/*.h files. That's the > root cause of

Re: svn commit: r352585 - head/sys/kern

Thanks! On Sat, Sep 21, 2019 at 4:03 PM Konstantin Belousov wrote: > > Author: kib > Date: Sat Sep 21 20:03:17 2019 > New Revision: 352585 > URL: https://svnweb.freebsd.org/changeset/base/352585 > > Log: > kern.elf{32,64}.pie_base sysctl: enforce page alignment. > > Requested by: rstone >

Re: svn commit: r346386 - in head/sys: dev/bge dev/pci dev/twa x86/iommu

IRC discussion on 2/28/12 that resulted in revision > r232267 > as a quick MFC'able fix, but I don't have a log of that conversation. :( I > couldn't find anything in e-mail either that was definitive for why this > might have > been inherent in PCI-e vs a few firmware writers ha

Re: svn commit: r346386 - in head/sys: dev/bge dev/pci dev/twa x86/iommu

Linux handle the issue? On Thu, Apr 25, 2019 at 3:17 PM Ryan Stone wrote: > > This change makes me *very* uncomfortable. It was originally brought > in due to issues with Adaptec RAID cards using the aac(9) driver. The > symptoms of the bug included silent corruption of data as it w

Re: svn commit: r346386 - in head/sys: dev/bge dev/pci dev/twa x86/iommu

This change makes me *very* uncomfortable. It was originally brought in due to issues with Adaptec RAID cards using the aac(9) driver. The symptoms of the bug included silent corruption of data as it was written to disk. Are we sure that this change is a good idea, given how catastrophic it is

Re: svn commit: r343525 - in head/sys/netinet: . tcp_stacks

On Mon, Jan 28, 2019 at 7:46 AM Michael Tuexen wrote: > > Author: tuexen > Date: Mon Jan 28 12:45:31 2019 > New Revision: 343525 > URL: https://svnweb.freebsd.org/changeset/base/343525 > > Log: > Fix the detection of ECN-setup SYN-ACK packets. > > RFC 3168 defines an ECN-setup SYN-ACK packet

svn commit: r339185 - head/sys/netinet

Author: rstone Date: Thu Oct 4 22:03:58 2018 New Revision: 339185 URL: https://svnweb.freebsd.org/changeset/base/339185 Log: Hold a write lock across udp_notify() With the new route cache feature udp_notify() will modify the inp when it needs to invalidate the route cache. Ensure that

Re: svn commit: r332447 - stable/11/sys/dev/ixgbe

Spinning in the kernel for a full second is a really bad idea. At minimum this is going to hold off all callouts from one of the callout threads for up to a full second as ixgbe_local_timer() waits for the core mutex. That chews up two CPU cores doing busy-wait loops (the ixgbe_stop() thread

svn commit: r329743 - in head: sbin/route sys/net

Author: rstone Date: Wed Feb 21 19:13:23 2018 New Revision: 329743 URL: https://svnweb.freebsd.org/changeset/base/329743 Log: Allow route change requests to not specify the gateway. Only require a gateway to be specified on a route add request. On a route change request that does not

svn commit: r329744 - head/sbin/route

Author: rstone Date: Wed Feb 21 19:13:27 2018 New Revision: 329744 URL: https://svnweb.freebsd.org/changeset/base/329744 Log: Fix route manpage to show correct flush syntax The current route(8) manpage shows that "flush" is an argument to the optional -n flag, rather than a separate

svn commit: r328271 - in head/sys: net netinet

Author: rstone Date: Tue Jan 23 03:15:39 2018 New Revision: 328271 URL: https://svnweb.freebsd.org/changeset/base/328271 Log: Reduce code duplication for inpcb route caching Add a new macro to clear both the L3 and L2 route caches, to hopefully prevent future instances where only the L3

svn commit: r328272 - head/sys/net

Author: rstone Date: Tue Jan 23 03:15:44 2018 New Revision: 328272 URL: https://svnweb.freebsd.org/changeset/base/328272 Log: Increment the route table gen count after a modify Increment the route table generation count after modifying a route. This signals back to TCP connections that

svn commit: r328270 - head/sys/net

Author: rstone Date: Tue Jan 23 03:15:35 2018 New Revision: 328270 URL: https://svnweb.freebsd.org/changeset/base/328270 Log: Invalidate inpcb LLE cache if cached route is invalidated When the inpcb route cache is invalidated after a change to the routing tables, we need to invalidate

svn commit: r326860 - head/sys/net

Author: rstone Date: Thu Dec 14 20:48:50 2017 New Revision: 326860 URL: https://svnweb.freebsd.org/changeset/base/326860 Log: Plug an ifaddr leak when changing a route's src If a route is modified in a way that changes the route's source address (i.e. the address used to access the

Re: svn commit: r326702 - head/sys/net

On Fri, Dec 8, 2017 at 1:43 PM, Stephen Hurd wrote: > Author: shurd > Date: Fri Dec 8 18:43:31 2017 > New Revision: 326702 > URL: https://svnweb.freebsd.org/changeset/base/326702 > + if (!M_WRITABLE(*m_head)) { > + new_head = m_dup(*m_head, M_NOWAIT); > +

Re: svn commit: r321477 - head/sys/net

On Tue, Jul 25, 2017 at 10:41 AM, Sean Bruno wrote: > Author: sbruno > Date: Tue Jul 25 14:41:50 2017 > New Revision: 321477 > URL: https://svnweb.freebsd.org/changeset/base/321477 > > Log: > Don't hold the RM lock during lagg_proto_addport() to avoid an LOR. Can

Re: svn commit: r321476 - head/sys/dev/ixgbe

On Tue, Jul 25, 2017 at 10:38 AM, Sean Bruno wrote: > Author: sbruno > Date: Tue Jul 25 14:38:30 2017 > New Revision: 321476 > URL: https://svnweb.freebsd.org/changeset/base/321476 > > Log: > Drop ixgbe RX lock during TCP_LRO processing. This eliminates a "storm" > of LOR

svn commit: r323506 - head/usr.sbin/iovctl

Author: rstone Date: Tue Sep 12 21:12:04 2017 New Revision: 323506 URL: https://svnweb.freebsd.org/changeset/base/323506 Log: Fix incorrect error message in iovctl If the iovctl command was invoked with only the -C flag, the user would receive a message claiming that they needed to also

svn commit: r321939 - head/usr.bin/calendar/calendars

+374,7 @@ 10/26 Philip M. Gollucci <pgollu...@freebsd.org> born in Silver Spring, Maryland, United States, 1979 10/27 Takanori Watanabe <takaw...@freebsd.org> born in Numazu, Shizuoka, Japan, 1972 10/31 Taras Korenko <ta...@freebsd.org> born in Cherkasy region, Ukraine,

svn commit: r321300 - svnadmin/conf

Author: rstone Date: Thu Jul 20 18:02:25 2017 New Revision: 321300 URL: https://svnweb.freebsd.org/changeset/base/321300 Log: Welcome Matt Joras (mjoras@) as a new src committer. Matt has been doing work at various layers of the network stack. I will be his mentor. Approved by:

Re: svn commit: r317755 - head/sbin/ifconfig

On Wed, May 3, 2017 at 1:39 PM, Ryan Stone <ryst...@gmail.com> wrote: > > > On Wed, May 3, 2017 at 1:21 PM, Alan Somers <asom...@freebsd.org> wrote: > >> Author: asomers >> Date: Wed May 3 17:21:01 2017 >> New Revision: 317755 >> URL: https://sv

Re: svn commit: r317755 - head/sbin/ifconfig

On Wed, May 3, 2017 at 1:21 PM, Alan Somers wrote: > Author: asomers > Date: Wed May 3 17:21:01 2017 > New Revision: 317755 > URL: https://svnweb.freebsd.org/changeset/base/317755 > > Log: > Various Coverity fixes in ifconfig(8) > > * Exit early if kldload(2) fails

svn commit: r316527 - head/sys/netinet

Author: rstone Date: Wed Apr 5 16:57:13 2017 New Revision: 316527 URL: https://svnweb.freebsd.org/changeset/base/316527 Log: Revert the optimization from r304436 r304436 attempted to optimize the handling of incoming UDP packet by only making an expensive call to in_broadcast() if the

Re: svn commit: r315333 - in stable/10/sys: conf dev/ixgbe modules/ix modules/ixv

On Thu, Mar 16, 2017 at 2:59 PM, Eric Joyner wrote: > I don't think the PCI_IOV stuff should be left in there. At least from > what I gathered in D4767, it's probably abandoned in 10. > It shouldn't hurt to leave it in, though. It should all be compiled under #ifdef PCI_IOV

Re: svn commit: r311849 - in head: . sys/amd64/conf sys/arm64/conf sys/conf sys/dev/e1000 sys/i386/conf sys/mips/conf sys/modules sys/modules/em sys/modules/igb sys/powerpc/conf

There's also the issue that running "ifconfig igb0 blah" during startup will cause if_igb to be automatically loaded by ifconfig. I guess we could add a dummy if_igb.ko that just has a dependency on if_em.ko On Fri, Mar 10, 2017 at 1:13 PM, Warner Losh wrote: > On Fri, Mar 10,

Re: svn commit: r313878 - head/sys/kern

On Tue, Feb 21, 2017 at 10:07 AM, Pedro Giffuni wrote: > Good point: it doesn't. > > The change gets ignored but it doesn't break the converter. At least not > immediately, the github converter in AOO breaks frequently but it is > probably unrelated. > It wouldn't "break" the

svn commit: r313824 - head/sys/kern

Author: rstone Date: Thu Feb 16 21:18:31 2017 New Revision: 313824 URL: https://svnweb.freebsd.org/changeset/base/313824 Log: Revert r313814 and r313816 Something evidently got mangled in my git tree in between testing and review, as an old and broken version of the patch was apparently

Re: svn commit: r313814 - head/sys/kern

Sorry about that. It's fixed in r313816. On Thu, Feb 16, 2017 at 3:31 PM, Cy Schubert <cy.schub...@komquats.com> wrote: > In message <201702161941.v1gjfdop087...@repo.freebsd.org>, Ryan Stone > writes: > > Author: rstone > > Date: Thu Feb 16 19:41:13 2017 > >

svn commit: r313816 - head/sys/kern

Author: rstone Date: Thu Feb 16 20:06:21 2017 New Revision: 313816 URL: https://svnweb.freebsd.org/changeset/base/313816 Log: Fix a typo in my previous commit Somehow in the late stages of testing my sched_ule patch, a character was accidentally deleted from the file. Correct this.

svn commit: r313814 - head/sys/kern

Author: rstone Date: Thu Feb 16 19:41:13 2017 New Revision: 313814 URL: https://svnweb.freebsd.org/changeset/base/313814 Log: Check for preemption after lowering a thread's priority When a high-priority thread is waiting for a mutex held by a low-priority thread, it temporarily lends its

svn commit: r313646 - head/sys/netinet

Author: rstone Date: Sat Feb 11 17:05:08 2017 New Revision: 313646 URL: https://svnweb.freebsd.org/changeset/base/313646 Log: Don't zero out srtt after excess retransmits If the TCP stack has retransmitted more than 1/4 of the total number of retransmits before a connection drop, it

svn commit: r313388 - in stable/11/sys: dev/ixgbe net

Author: rstone Date: Tue Feb 7 15:13:19 2017 New Revision: 313388 URL: https://svnweb.freebsd.org/changeset/base/313388 Log: MFC r312544 Fix reference to free memory in ixgbe/if_media.c When ixgbe receives an interrupt indicating that a new optical module may have been

svn commit: r313387 - in stable/10/sys: dev/ixgbe net

Author: rstone Date: Tue Feb 7 15:12:27 2017 New Revision: 313387 URL: https://svnweb.freebsd.org/changeset/base/313387 Log: MFC r312544 Fix reference to free memory in ixgbe/if_media.c When ixgbe receives an interrupt indicating that a new optical module may have been

svn commit: r312544 - in head/sys: dev/ixgbe net

Author: rstone Date: Fri Jan 20 17:16:48 2017 New Revision: 312544 URL: https://svnweb.freebsd.org/changeset/base/312544 Log: Fix reference to free memory in ixgbe/if_media.c When ixgbe receives an interrupt indicating that a new optical module may have been inserted, it discards all of

Re: svn commit: r311283 - head/sys/cam/ctl

On Wed, Jan 4, 2017 at 7:50 AM, Edward Tomasz Napierala wrote: > + refcount_release(>cs_outstanding_ctl_pdus); > Shouldn't the return value of refcount_release() be checked? ___ svn-src-all@freebsd.org mailing list

Re: svn commit: r310180 - head/sys/net

On Fri, Dec 16, 2016 at 5:39 PM, Alan Somers wrote: > > Modified: head/sys/net/if_lagg.c > > == > --- head/sys/net/if_lagg.c Fri Dec 16 22:37:16 2016(r310179) > +++

svn commit: r309414 - head/sys/sys

Author: rstone Date: Fri Dec 2 15:38:34 2016 New Revision: 309414 URL: https://svnweb.freebsd.org/changeset/base/309414 Log: Revert r309372 The bug intended to be fixed by r309372 was already addressed by r296178, so revert my change. Reported by: seph Modified:

svn commit: r309372 - head/sys/sys

Author: rstone Date: Thu Dec 1 21:08:42 2016 New Revision: 309372 URL: https://svnweb.freebsd.org/changeset/base/309372 Log: Fix a false positive in a buf_ring assert buf_ring contains an assert that checks whether an item being enqueued already exists on the ring. There is a subtle

Re: svn commit: r309194 - head/usr.sbin/syslogd

> > There is something very wrong in the code here. nextp is never > initialized, so obviously we can't access *nextp. I'm surprised that this > code can even compile with -Werror. I'd fix it but it'd not clear to me > what nextp is intended to do. > Apologies for the noise; I didn't see

Re: svn commit: r309194 - head/usr.sbin/syslogd

On Sat, Nov 26, 2016 at 10:50 AM, Baptiste Daroussin wrote: > Author: bapt > Date: Sat Nov 26 15:49:59 2016 > New Revision: 309194 > URL: https://svnweb.freebsd.org/changeset/base/309194 > > Log: > initialize *nextp which could be left uninitialized in case the >

svn commit: r309177 - stable/11/sys/net

Author: rstone Date: Sat Nov 26 01:16:33 2016 New Revision: 309177 URL: https://svnweb.freebsd.org/changeset/base/309177 Log: MFC r308580: Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and

svn commit: r308790 - head/tools/sched

@@ +#!/bin/sh +# +# Copyright (c) 2012 Ryan Stone +# All rights reserved. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# 1. Redistributions of source code must retain the above copyright

svn commit: r308580 - head/sys/net

Author: rstone Date: Sat Nov 12 19:03:23 2016 New Revision: 308580 URL: https://svnweb.freebsd.org/changeset/base/308580 Log: Don't read if_counters with if_addr_lock held Calling into an ifnet implementation with the if_addr_lock already held can cause a LOR and potentially a deadlock,

svn commit: r308562 - head/tools/tools/git

Author: rstone Date: Fri Nov 11 23:07:31 2016 New Revision: 308562 URL: https://svnweb.freebsd.org/changeset/base/308562 Log: Fix git tools when run against a worktree In a git worktree, the gitdir is in an entirely different location. In arcgit, use git rev-parse --git-dir to get the

Re: svn commit: r304437 - head/sys/netinet

There are some performance implications of this change that are unresolved in the UDP receive path. I haven't had a chance to go back and try to address that, so I can't MFC until then. On Tue, Nov 1, 2016 at 11:13 AM, Harry Schmalzbauer wrote: > Bezüglich Ryan Stone's

svn commit: r307887 - head/sys/netinet

Author: rstone Date: Mon Oct 24 22:11:33 2016 New Revision: 307887 URL: https://svnweb.freebsd.org/changeset/base/307887 Log: Fix ip_output() on point-to-point links In r304435, ip_output() was changed to use the result of the route lookup to decide whether the outgoing packet was a

svn commit: r307681 - head/tools/tools/git

Author: rstone Date: Thu Oct 20 18:28:05 2016 New Revision: 307681 URL: https://svnweb.freebsd.org/changeset/base/307681 Log: Set the executable bit on arcgit and importgit Make it possible to run these scripts directly out of svn by setting the executable property on them. Modified:

svn commit: r307680 - in head/tools/tools: . git

(r307680) @@ -0,0 +1,214 @@ +#!/bin/sh +# +# Copyright (c) 2015 Ryan Stone. All rights reserved. +# +# Redistribution and use in source and binary forms, with or without +# modification, are permitted provided that the following conditions +# are met: +# 1. Redistributions of source code must retain th

svn commit: r304606 - head/sys/netinet

Author: rstone Date: Mon Aug 22 15:27:37 2016 New Revision: 304606 URL: https://svnweb.freebsd.org/changeset/base/304606 Log: Temporarily disable the optimization from r304436 r304436 attempted to optimize the handling of incoming UDP packet by only making an expensive call to

Re: svn commit: r304548 - head/sys/netinet

Thanks! On Sat, Aug 20, 2016 at 6:12 PM, Marko Zec wrote: > Author: zec > Date: Sat Aug 20 22:12:26 2016 > New Revision: 304548 > URL: https://svnweb.freebsd.org/changeset/base/304548 > > Log: > Permit disabling net.inet.udp.require_l2_bcast in VIMAGE kernels. > > The

Re: svn commit: r304436 - in head: . sys/netinet

On Sat, Aug 20, 2016 at 2:45 PM, Slawa Olhovchenkov wrote: > You also can recive this on ethernet too, IMHO, in case of /32. > Receiving L3 broadcst in L2 unicast is legitime (IMHO) and we must be > relaxed on this. > There is no broadcast address on a /32 network. Independent

Re: svn commit: r304436 - in head: . sys/netinet

On Sat, Aug 20, 2016 at 2:09 PM, Adrian Chadd wrote: > Hi, > > So why not fix those paths? > > > > -adrian > Is there even a fix that can be done there? I wouldn't expect a point-to-point protocol like PPP to have any need to explicitly signal that a packet is a broadcast.

Re: svn commit: r304436 - in head: . sys/netinet

On Sat, Aug 20, 2016 at 1:30 PM, Slawa Olhovchenkov wrote: > For host enought have [hidden] alias with broadcsts bits. > Anyway, don't should relay on the L2 information, you can recive L3 > unicast addressed packets (with alien dst IP address) in L2 broadcas > packet. > This

Re: svn commit: r304436 - in head: . sys/netinet

, Bruce Simpson <b...@fastmail.net> wrote: > On 20/08/16 16:27, Ryan Stone wrote: > >> Can you send a broadcast packet through an L3 tunnel? I thought that a >> L2 tunnel was required. >> > > Yes. This is perfectly legal and necessary for forwarding of IPv4 &

Re: svn commit: r304436 - in head: . sys/netinet

On Sat, Aug 20, 2016 at 11:01 AM, Bruce Simpson wrote: > tun(4) on the other hand is a plain, PPP-like, IP tunnel. > Can you send a broadcast packet through an L3 tunnel? I thought that a L2 tunnel was required. But this mbuf flag is not guaranteed to be set in all

Re: svn commit: r304435 - head/sys/netinet

On Sat, Aug 20, 2016 at 6:08 AM, Kubilay Kocak wrote: > Hi Ryan, > > Could you elaborate on any of the potential performance implications of > this? > As if r304437, skipping the call to in_broadcast() means that we avoid an additional (potentially heavily contended) rlock

Re: svn commit: r304436 - in head: . sys/netinet

On Sat, Aug 20, 2016 at 7:48 AM, Bruce Simpson wrote: > It is perfectly legal for broadcast packets to be addressed to the end of > a P2P or non-Ethernet link, which may not set M_BCAST or M_MCAST. The > classic example is ATM (Non-Broadcast, Multiple Access (NBMA)) but the >

svn commit: r304437 - head/sys/netinet

Author: rstone Date: Thu Aug 18 22:59:10 2016 New Revision: 304437 URL: https://svnweb.freebsd.org/changeset/base/304437 Log: Fix unlocked access to ifnet address list in_broadcast() was iterating over the ifnet address list without first taking an IF_ADDR_RLOCK. This could cause a

svn commit: r304435 - head/sys/netinet

Author: rstone Date: Thu Aug 18 22:59:00 2016 New Revision: 304435 URL: https://svnweb.freebsd.org/changeset/base/304435 Log: Don't iterate over the ifnet addr list in ip_output() For almost every packet that is transmitted through ip_output(), a call to in_broadcast() was made to decide

svn commit: r304436 - in head: . sys/netinet

Author: rstone Date: Thu Aug 18 22:59:05 2016 New Revision: 304436 URL: https://svnweb.freebsd.org/changeset/base/304436 Log: Don't check for broadcast IPs on non-bcast pkts in_broadcast() can be quite expensive, so skip calling it if the incoming mbuf wasn't sent to a broadcast L2

svn commit: r304179 - releng/11.0/sys/dev/bxe

Author: rstone Date: Mon Aug 15 19:18:10 2016 New Revision: 304179 URL: https://svnweb.freebsd.org/changeset/base/304179 Log: MFC r304163 Don't enqueue NULL on a drbr In one corner case in the bxe TX path, a NULL mbuf could be enqueued onto a drbr queue. This could

svn commit: r304163 - stable/11/sys/dev/bxe

Author: rstone Date: Mon Aug 15 15:23:45 2016 New Revision: 304163 URL: https://svnweb.freebsd.org/changeset/base/304163 Log: MFC r303836 Don't enqueue NULL on a drbr In one corner case in the bxe TX path, a NULL mbuf could be enqueued onto a drbr queue. This could case a

svn commit: r303836 - head/sys/dev/bxe

Author: rstone Date: Mon Aug 8 16:19:24 2016 New Revision: 303836 URL: https://svnweb.freebsd.org/changeset/base/303836 Log: Don't enqueue NULL on a drbr In one corner case in the bxe TX path, a NULL mbuf could be enqueued onto a drbr queue. This could case a KASSERT to fire with

Re: svn commit: r296868 - head/sys/sys

On Mon, Mar 14, 2016 at 2:07 PM, Gleb Smirnoff wrote: > Remove useless cast in SYSCTL_ADD_COUNTER_U64 macro. > Is it useless? I believe that the point is to give a compiler error if an incompatible pointer type was passed as the ptr parameter.

svn commit: r296865 - in head: sys/dev/pci sys/sys usr.sbin/iovctl

Author: rstone Date: Mon Mar 14 17:41:17 2016 New Revision: 296865 URL: https://svnweb.freebsd.org/changeset/base/296865 Log: Clean up repeated "All rights reserved" Modified: head/sys/dev/pci/pci_iov.c head/sys/dev/pci/pci_iov_private.h head/sys/dev/pci/pci_iov_schema.c

Re: svn commit: r294327 - in head/sys: dev/cxgb dev/cxgbe dev/e1000 dev/hyperv/netvsc dev/ixgbe dev/mxge netinet sys

On Tue, Jan 19, 2016 at 3:23 PM, Adrian Chadd wrote: > Sure, but TCP fragments and non-fragments can hash to different > values, so suddenly you get interleaved packets. > If the NIC is hashing fragments and non-fragments to different values, aren't we already

Re: svn commit: r294327 - in head/sys: dev/cxgb dev/cxgbe dev/e1000 dev/hyperv/netvsc dev/ixgbe dev/mxge netinet sys

On Tue, Jan 19, 2016 at 10:33 AM, Hans Petter Selasky wrote: > > + qsort(lc->lro_mbuf_data, lc->lro_mbuf_count, sizeof(struct mbuf *), > + _lro_mbuf_compare_header); > In the worst case, qsort() can take O(n**2) time and consume O(n) stack space. Is there

Re: svn commit: r294327 - in head/sys: dev/cxgb dev/cxgbe dev/e1000 dev/hyperv/netvsc dev/ixgbe dev/mxge netinet sys

libkern's qsort() is a quicksort implementation. AFAICT it has the worst case behaviour that I describe. On Tue, Jan 19, 2016 at 11:54 AM, Hans Petter Selasky <h...@selasky.org> wrote: > On 01/19/16 17:09, Ryan Stone wrote: > >> On Tue, Jan 19, 2016 at 10:33 AM, Hans Petter

svn commit: r293208 - stable/10/sbin/route

Author: rstone Date: Tue Jan 5 16:58:07 2016 New Revision: 293208 URL: https://svnweb.freebsd.org/changeset/base/293208 Log: MFC r287920: Fix /sbin/route to never look up (invalid) interface names through DNS /sbin/route has a bug where if it is passed an interface name that does

svn commit: r287922 - stable/10/sys/vm

Author: rstone Date: Thu Sep 17 17:09:03 2015 New Revision: 287922 URL: https://svnweb.freebsd.org/changeset/base/287922 Log: MFC r286970: Prevent ticks rollover from preventing vm_lowmem event Currently vm_pageout_scan() uses a ticks-based scheme to rate-limit the number of

svn commit: r287926 - stable/10/sys/dev/puc

Author: rstone Date: Thu Sep 17 17:56:23 2015 New Revision: 287926 URL: https://svnweb.freebsd.org/changeset/base/287926 Log: MFC r263109 Add MSI support to puc(9) Add support for MSI interrupts in the puc(9) driver. By default the driver will prefer MSI interrupts to legacy

svn commit: r287929 - stable/10/sys/kern

Author: rstone Date: Thu Sep 17 18:21:47 2015 New Revision: 287929 URL: https://svnweb.freebsd.org/changeset/base/287929 Log: MFC r279410: Correct the use of an unitialized variable in sendfind_getobj() When sendfile_getobj() is called on a DTYPE_SHM file, it never initializes

svn commit: r287928 - stable/10/sys/dev/watchdog

Author: rstone Date: Thu Sep 17 18:19:55 2015 New Revision: 287928 URL: https://svnweb.freebsd.org/changeset/base/287928 Log: MFC r279413: Add a missing include of an options header. watchdog.c does an #ifdef DDB but does not #include "opt_ddb.h". Fixing this turned up a

svn commit: r287920 - head/sbin/route

Author: rstone Date: Thu Sep 17 16:56:49 2015 New Revision: 287920 URL: https://svnweb.freebsd.org/changeset/base/287920 Log: Fix /sbin/route to never look up (invalid) interface names through DNS /sbin/route has a bug where if it is passed an interface name that does not exist, it falls

svn commit: r287945 - in stable/10/sys: amd64/amd64 i386/i386 ia64/ia64 kern mips/mips powerpc/aim sparc64/sparc64 sys vm

Author: rstone Date: Thu Sep 17 23:31:44 2015 New Revision: 287945 URL: https://svnweb.freebsd.org/changeset/base/287945 Log: MFC r280957 Fix integer truncation bug in malloc(9) A couple of internal functions used by malloc(9) and uma truncated a size_t down to an int. This

svn commit: r286970 - head/sys/vm

Author: rstone Date: Thu Aug 20 20:28:51 2015 New Revision: 286970 URL: https://svnweb.freebsd.org/changeset/base/286970 Log: Prevent ticks rollover from preventing vm_lowmem event Currently vm_pageout_scan() uses a ticks-based scheme to rate-limit the number of times that the vm_lowmem

Re: svn commit: r286100 - in head/sys: net netipsec

You can't use CTASSERT in a header. You'll get a compile error if two different headers included in the same translation unit have a CTASSERT on the same line number. On Jul 30, 2015 5:23 PM, John-Mark Gurney j...@freebsd.org wrote: Author: jmg Date: Fri Jul 31 00:23:21 2015 New Revision:

Re: svn commit: r283883 - in head/sys: dev/ixgbe modules/ixv

I think that this is a conflict with r283670, which changed the interface for SR-IOV a little bit. Sorry Jack, I knew that that change was coming but it didn't occur to me that the ixgbe patch would have to be updated. I believe that this patch will fix it: diff --git a/sys/dev/ixgbe/if_ix.c

svn commit: r280957 - in head/sys: amd64/amd64 i386/i386 kern mips/mips powerpc/aim sparc64/sparc64 sys vm

Author: rstone Date: Wed Apr 1 12:42:26 2015 New Revision: 280957 URL: https://svnweb.freebsd.org/changeset/base/280957 Log: Fix integer truncation bug in malloc(9) A couple of internal functions used by malloc(9) and uma truncated a size_t down to an int. This could cause any number

Re: svn commit: r280957 - in head/sys: amd64/amd64 i386/i386 kern mips/mips powerpc/aim sparc64/sparc64 sys vm

On Wed, Apr 1, 2015 at 11:10 AM, Alan Cox a...@rice.edu wrote: On 04/01/2015 07:42, Ryan Stone wrote: Note to self: When this is MFCed, sparc64 needs the same fix. I suspect that you mean ia64 here, not sparc64. Yes? Apparently. Thanks for the correction

Re: svn commit: r280120 - head/sys/dev/wpi

On Sun, Mar 15, 2015 at 5:35 PM, Adrian Chadd adr...@freebsd.org wrote: .. promise I'm done for now. (God, it'd be nice to use git, or some web ui that lets me batch review and commit things like this.) I have a script that takes a series of git commits and makes a separate Differential

Re: svn commit: r280120 - head/sys/dev/wpi

Anyone want to review them before I commit them? (I'm going to put them in tools/tools/git/) On Sun, Mar 15, 2015 at 5:45 PM, Adrian Chadd adr...@freebsd.org wrote: On 15 March 2015 at 14:44, Ryan Stone ryst...@gmail.com wrote: On Sun, Mar 15, 2015 at 5:35 PM, Adrian Chadd adr...@freebsd.org

Re: svn commit: r280120 - head/sys/dev/wpi

I've thrown them up for review here: https://reviews.freebsd.org/D2071 Comments are welcome. I'll commit it all later this week. ___ svn-src-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any

Re: svn commit: r279932 - head/sys/vm

On Thu, Mar 12, 2015 at 2:06 PM, Ian Lepore i...@freebsd.org wrote: Nullterminate strings returned via sysctl. PR: 195668 To quote the manpage: The *sbuf* family of functions allows one to safely allocate, construct and release bounded null-terminated strings in kernel

svn commit: r279939 - head/sys/dev/hwpmc

Author: rstone Date: Thu Mar 12 23:44:28 2015 New Revision: 279939 URL: https://svnweb.freebsd.org/changeset/base/279939 Log: hwpmc: Fix event number to match enum name Differential revision:https://reviews.freebsd.org/D1592 Reviewed by: Joseph Kong MFC after:1 month

svn commit: r279868 - in head/sys: dev/pci kern sys

Author: rstone Date: Tue Mar 10 23:27:13 2015 New Revision: 279868 URL: https://svnweb.freebsd.org/changeset/base/279868 Log: Fix SR-IOV passthrough devices to allow ppt to attach A late change to the SR-IOV infrastructure broke passthrough of VFs. device_set_devclass() was being used

svn commit: r279832 - in head: lib/libpmc sys/dev/hwpmc

Author: rstone Date: Tue Mar 10 01:23:47 2015 New Revision: 279832 URL: https://svnweb.freebsd.org/changeset/base/279832 Log: Fix Sandy Bridge+ hwpmc branch counters On Sandy Bridge and later, to count branch-related events you have to or together a mask indicating the type of branch

svn commit: r279835 - in head: lib/libpmc sys/dev/hwpmc

Author: rstone Date: Tue Mar 10 01:24:08 2015 New Revision: 279835 URL: https://svnweb.freebsd.org/changeset/base/279835 Log: Fix Ivy Bridge+ MEM_UOPS_RETIRED counters The MEM_UOPS_RETIRED actually work the same way as the Sandy Bridge counters, but the counters were documented in a

svn commit: r279830 - head/sys/dev/hwpmc

Author: rstone Date: Tue Mar 10 01:23:34 2015 New Revision: 279830 URL: https://svnweb.freebsd.org/changeset/base/279830 Log: Fix various bugs in Haswell counter definitions 1) The WALK_COMPLETED_2M_4M event incorrectly referenced 4K pages. 2) The umask for RING0 and RING123 events was

svn commit: r279833 - head/lib/libpmc

Author: rstone Date: Tue Mar 10 01:23:55 2015 New Revision: 279833 URL: https://svnweb.freebsd.org/changeset/base/279833 Log: Use the correct event table for Haswell Xeon events Differential Revision:https://reviews.freebsd.org/D1588 MFC after:1 month Sponsored by: Sandvine

  1   2   3   >