svn commit: r368743 - head/tests/sys/net

Author: kp Date: Thu Dec 17 21:54:25 2020 New Revision: 368743 URL: https://svnweb.freebsd.org/changeset/base/368743 Log: net tests: Re-enable most if_clone tests All but one of these (tap_ipv6_up_stress) currently pass, so we should enable them so we don't regress. Modified:

svn commit: r368666 - stable/12/tests/sys/netpfil/pf

Author: kp Date: Tue Dec 15 16:15:25 2020 New Revision: 368666 URL: https://svnweb.freebsd.org/changeset/base/368666 Log: MFC r368239: pf tests: Re-enable panicing tests We've fixed the vnet/epair cleanup race, so it is now safe to re-enable these tests. Sponsored by: Modirum

svn commit: r368665 - stable/12/sys/net

Author: kp Date: Tue Dec 15 16:13:32 2020 New Revision: 368665 URL: https://svnweb.freebsd.org/changeset/base/368665 Log: MFC r368238: net: Revert vnet/epair cleanup race mitigation Revert the mitigation code for the vnet/epair cleanup race (done in r365457). r368237 introduced a

svn commit: r368663 - stable/12/sys/net

Author: kp Date: Tue Dec 15 15:33:28 2020 New Revision: 368663 URL: https://svnweb.freebsd.org/changeset/base/368663 Log: MFC r368237: if: Fix panic when destroying vnet and epair simultaneously When destroying a vnet and an epair (with one end in the vnet) we often panicked. This

svn commit: r368656 - stable/12/sys/netpfil/pf

Author: kp Date: Tue Dec 15 08:29:45 2020 New Revision: 368656 URL: https://svnweb.freebsd.org/changeset/base/368656 Log: MFC r368588: pf: Allow net.pf.request_maxcount to be set from loader.conf Mark request_maxcount as RWTUN so we can set it both at runtime and from loader.conf.

svn commit: r368588 - head/sys/netpfil/pf

Author: kp Date: Sat Dec 12 20:14:39 2020 New Revision: 368588 URL: https://svnweb.freebsd.org/changeset/base/368588 Log: pf: Allow net.pf.request_maxcount to be set from loader.conf Mark request_maxcount as RWTUN so we can set it both at runtime and from loader.conf. This avoids usings

svn commit: r368554 - stable/12/sys/net

Author: kp Date: Fri Dec 11 15:39:22 2020 New Revision: 368554 URL: https://svnweb.freebsd.org/changeset/base/368554 Log: MFC r368020, r368025: if: Protect V_ifnet in vnet_if_return() When we terminate a vnet (i.e. jail) we move interfaces back to their home vnet. We need to protect

svn commit: r368552 - stable/12/sys/net

Author: kp Date: Fri Dec 11 14:11:41 2020 New Revision: 368552 URL: https://svnweb.freebsd.org/changeset/base/368552 Log: MFC r368015: if: Remove ifnet_rwlock It no longer serves any purpose, as evidenced by the fact that we never take it without ifnet_sxlock. This differs

svn commit: r368488 - stable/12/tests/sys/netpfil/pf

Author: kp Date: Wed Dec 9 17:17:45 2020 New Revision: 368488 URL: https://svnweb.freebsd.org/changeset/base/368488 Log: MFC r368277: pf tests: Test case for bug #251414 Changing a table from not having counters to having counters (or vice versa) may trigger panics. PR:

svn commit: r368277 - head/tests/sys/netpfil/pf

Author: kp Date: Wed Dec 2 16:33:23 2020 New Revision: 368277 URL: https://svnweb.freebsd.org/changeset/base/368277 Log: pf tests: Test case for bug #251414 Changing a table from not having counters to having counters (or vice versa) may trigger panics. PR: 251414 MFC

svn commit: r368239 - head/tests/sys/netpfil/pf

Author: kp Date: Tue Dec 1 16:44:36 2020 New Revision: 368239 URL: https://svnweb.freebsd.org/changeset/base/368239 Log: pf tests: Re-enable panicing tests We've fixed the vnet/epair cleanup race, so it is now safe to re-enable these tests. MFC after:2 weeks Sponsored by:

svn commit: r368238 - head/sys/net

Author: kp Date: Tue Dec 1 16:34:43 2020 New Revision: 368238 URL: https://svnweb.freebsd.org/changeset/base/368238 Log: net: Revert vnet/epair cleanup race mitigation Revert the mitigation code for the vnet/epair cleanup race (done in r365457). r368237 introduced a more reliable fix.

svn commit: r368237 - head/sys/net

Author: kp Date: Tue Dec 1 16:23:59 2020 New Revision: 368237 URL: https://svnweb.freebsd.org/changeset/base/368237 Log: if: Fix panic when destroying vnet and epair simultaneously When destroying a vnet and an epair (with one end in the vnet) we often panicked. This was the result of

svn commit: r368235 - stable/12/tests/sys/netpfil/pf/ioctl

Author: kp Date: Tue Dec 1 16:02:52 2020 New Revision: 368235 URL: https://svnweb.freebsd.org/changeset/base/368235 Log: MFC r367990: pf tests: provoke tag hashing panic Attempt to provoke the panic fixed in r367987. Modified: stable/12/tests/sys/netpfil/pf/ioctl/validation.c

svn commit: r368101 - stable/12/tests/sys/netpfil/pf

is newly added) +++ stable/12/tests/sys/netpfil/pf/src_track.sh Fri Nov 27 11:12:59 2020 (r368101, copy of r367869, head/tests/sys/netpfil/pf/src_track.sh) @@ -0,0 +1,66 @@ +# $FreeBSD$ +# +# SPDX-License-Identifier: BSD-2-Clause-FreeBSD +# +# Copyright (c) 2020 Kristof Provost

svn commit: r368100 - stable/12/sys/netpfil/pf

Author: kp Date: Fri Nov 27 11:12:10 2020 New Revision: 368100 URL: https://svnweb.freebsd.org/changeset/base/368100 Log: MFC r367867: pf: Fix incorrect assertion We never set PFRULE_RULESRCTRACK when calling pf_insert_src_node(). We do set PFRULE_SRCTRACK, so update the assertion

svn commit: r368054 - stable/12/sys/net

Author: kp Date: Thu Nov 26 10:17:56 2020 New Revision: 368054 URL: https://svnweb.freebsd.org/changeset/base/368054 Log: Add missing NET_EPOCH_EXIT() to if_bridge In r367706 one return in bridge_input() did not NET_EPOCH_EXIT(), which caused seemingly random panics on the next use of

svn commit: r368025 - head/sys/net

Author: kp Date: Wed Nov 25 17:15:24 2020 New Revision: 368025 URL: https://svnweb.freebsd.org/changeset/base/368025 Log: if: Fix non-VIMAGE build if_link_ifnet() and if_unlink_ifnet() are needed even when VIMAGE is not enabled. MFC after:2 weeks Sponsored by: Modirum MDPay

svn commit: r368020 - head/sys/net

Author: kp Date: Wed Nov 25 15:07:22 2020 New Revision: 368020 URL: https://svnweb.freebsd.org/changeset/base/368020 Log: if: Protect V_ifnet in vnet_if_return() When we terminate a vnet (i.e. jail) we move interfaces back to their home vnet. We need to protect our access to the V_ifnet

svn commit: r368015 - head/sys/net

Author: kp Date: Wed Nov 25 10:56:38 2020 New Revision: 368015 URL: https://svnweb.freebsd.org/changeset/base/368015 Log: if: Remove ifnet_rwlock It no longer serves any purpose, as evidenced by the fact that we never take it without ifnet_sxlock. Sponsored by: Modirum MDPay

svn commit: r367990 - head/tests/sys/netpfil/pf/ioctl

Author: kp Date: Tue Nov 24 17:32:49 2020 New Revision: 367990 URL: https://svnweb.freebsd.org/changeset/base/367990 Log: pf tests: provoke tag hashing panic Attempt to provoke the panic fixed in r367987. MFC after:1 week Differential Revision:

svn commit: r367869 - head/tests/sys/netpfil/pf

@@ +# $FreeBSD$ +# +# SPDX-License-Identifier: BSD-2-Clause-FreeBSD +# +# Copyright (c) 2020 Kristof Provost +# +# 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

svn commit: r367867 - head/sys/netpfil/pf

Author: kp Date: Fri Nov 20 10:08:33 2020 New Revision: 367867 URL: https://svnweb.freebsd.org/changeset/base/367867 Log: pf: Fix incorrect assertion We never set PFRULE_RULESRCTRACK when calling pf_insert_src_node(). We do set PFRULE_SRCTRACK, so update the assertion to match. MFC

svn commit: r367706 - stable/12/sys/net

Author: kp Date: Sun Nov 15 11:56:16 2020 New Revision: 367706 URL: https://svnweb.freebsd.org/changeset/base/367706 Log: MFC r366500: bridge: call member interface ioctl() without NET_EPOCH We're not allowed to hold NET_EPOCH while sleeping, so when we call ioctl() handlers for

svn commit: r367705 - stable/12/sys/net

Author: kp Date: Sun Nov 15 11:46:44 2020 New Revision: 367705 URL: https://svnweb.freebsd.org/changeset/base/367705 Log: bridge: epoch-ification Undo the revert (r363568). This commit still violates epoch rules by sleeping within NET_EPOCH. That will be resolved in the following commit.

Re: svn commit: r367321 - head/sys/amd64/linux

On 4 Nov 2020, at 11:45, Gordon Bergling wrote: Shouldn't such a commit be approved by someone? The committer’s guide seems clear that doc committers may fix comments: https://www.freebsd.org/doc/en_US.ISO8859-1/articles/committers-guide/committer.types.html doc committers may commit

svn commit: r367078 - head/sys/riscv/riscv

Author: kp Date: Tue Oct 27 12:44:49 2020 New Revision: 367078 URL: https://svnweb.freebsd.org/changeset/base/367078 Log: riscv: Minor cleanup in startup code - remove setting of register value which is not used until the next value is set - Use the L2_SHIFT constant when setting

svn commit: r367058 - stable/12/tests/sys/netpfil/pf

Author: kp Date: Mon Oct 26 13:24:20 2020 New Revision: 367058 URL: https://svnweb.freebsd.org/changeset/base/367058 Log: MFC r366648: pf tests: Test that 'set skip on ' works on new group members There's a know issue where new group members don't get the 'set skip on' applied until

svn commit: r367057 - stable/12/sys/netpfil/pf

Author: kp Date: Mon Oct 26 13:23:40 2020 New Revision: 367057 URL: https://svnweb.freebsd.org/changeset/base/367057 Log: MFC r37: pf: do not remove kifs that are referenced by rules Even if a kif doesn't have an ifp or if_group pointer we still can't delete it if it's

svn commit: r367056 - stable/12/sys/netpfil/pf

Author: kp Date: Mon Oct 26 13:22:55 2020 New Revision: 367056 URL: https://svnweb.freebsd.org/changeset/base/367056 Log: MFC r366647: pf: create a kif for flags If userspace tries to set flags (e.g. 'set skip on ') and doesn't exist we should create a kif so that we apply the

svn commit: r366667 - head/sys/netpfil/pf

Author: kp Date: Tue Oct 13 11:04:00 2020 New Revision: 37 URL: https://svnweb.freebsd.org/changeset/base/37 Log: pf: do not remove kifs that are referenced by rules Even if a kif doesn't have an ifp or if_group pointer we still can't delete it if it's referenced by a rule. In

svn commit: r366648 - head/tests/sys/netpfil/pf

Author: kp Date: Mon Oct 12 12:41:10 2020 New Revision: 366648 URL: https://svnweb.freebsd.org/changeset/base/366648 Log: pf tests: Test that 'set skip on ' works on new group members There's a know issue where new group members don't get the 'set skip on' applied until the rules are

svn commit: r366647 - head/sys/netpfil/pf

Author: kp Date: Mon Oct 12 12:39:37 2020 New Revision: 366647 URL: https://svnweb.freebsd.org/changeset/base/366647 Log: pf: create a kif for flags If userspace tries to set flags (e.g. 'set skip on ') and doesn't exist we should create a kif so that we apply the flags when the does

svn commit: r366500 - head/sys/net

Author: kp Date: Tue Oct 6 19:19:56 2020 New Revision: 366500 URL: https://svnweb.freebsd.org/changeset/base/366500 Log: bridge: call member interface ioctl() without NET_EPOCH We're not allowed to hold NET_EPOCH while sleeping, so when we call ioctl() handlers for member interfaces we

svn commit: r366461 - head/sbin/devfs

Author: kp Date: Mon Oct 5 19:26:54 2020 New Revision: 366461 URL: https://svnweb.freebsd.org/changeset/base/366461 Log: devfs.rules: unhide pf in vnet jails /dev/pf is usable in vnet jails, so don't hide the node there. We shouldn't expose /dev/pf in regular jails, as that gives

svn commit: r366355 - head/sys/riscv/riscv

Author: kp Date: Fri Oct 2 07:30:11 2020 New Revision: 366355 URL: https://svnweb.freebsd.org/changeset/base/366355 Log: riscv: handle access faults in user mode Access faults in user mode are treated like TLB misses, which leads to an endless loop of faults. It's less serious than the

svn commit: r366315 - in head/sys/riscv: include riscv

Author: kp Date: Thu Oct 1 15:04:55 2020 New Revision: 366315 URL: https://svnweb.freebsd.org/changeset/base/366315 Log: riscv: Add memmmap so we can mmap /dev/mem Reviewed by: mhorne Sponsored by: Axiado Differential Revision:https://reviews.freebsd.org/D26622 Modified:

svn commit: r366284 - head/sys/riscv/riscv

Author: kp Date: Wed Sep 30 08:23:43 2020 New Revision: 366284 URL: https://svnweb.freebsd.org/changeset/base/366284 Log: riscv: Panic on PMP errors Load/store/fetch access exceptions always indicate a violation of a PMP rule. We can't treat those as page faults, because updating the

svn commit: r365669 - releng/12.2/sys/net

Author: kp Date: Sat Sep 12 18:58:36 2020 New Revision: 365669 URL: https://svnweb.freebsd.org/changeset/base/365669 Log: MFC r365457: net: mitigate vnet / epair cleanup races There's a race where dying vnets move their interfaces back to their original vnet, and if_epair cleanup

svn commit: r365659 - stable/12/sys/net

Author: kp Date: Sat Sep 12 12:45:31 2020 New Revision: 365659 URL: https://svnweb.freebsd.org/changeset/base/365659 Log: MFC r365457: net: mitigate vnet / epair cleanup races There's a race where dying vnets move their interfaces back to their original vnet, and if_epair cleanup

svn commit: r365626 - in head/sys/cddl/dev: dtrace/riscv fbt/riscv

Author: kp Date: Fri Sep 11 09:15:49 2020 New Revision: 365626 URL: https://svnweb.freebsd.org/changeset/base/365626 Log: dtrace: fix fbt return probes on RISC-V Return values are passed in a0, so read it from there. We also pass a1 through to userspace, as the ABI allows small structs

svn commit: r365457 - head/sys/net

Author: kp Date: Tue Sep 8 14:54:10 2020 New Revision: 365457 URL: https://svnweb.freebsd.org/changeset/base/365457 Log: net: mitigate vnet / epair cleanup races There's a race where dying vnets move their interfaces back to their original vnet, and if_epair cleanup (where deleting one

svn commit: r365246 - head/sys/netgraph

Author: kp Date: Wed Sep 2 11:49:22 2020 New Revision: 365246 URL: https://svnweb.freebsd.org/changeset/base/365246 Log: ng_ether: Enter NET_EPOCH where required We must enter NET_EPOCH before calling ether_output_frame(). Several of the functions it calls (pfil_run_hooks, if_transmit)

svn commit: r365245 - head/sys/riscv/riscv

Author: kp Date: Wed Sep 2 11:33:31 2020 New Revision: 365245 URL: https://svnweb.freebsd.org/changeset/base/365245 Log: riscv: very large dma mappings can cause integer overflow Fix the return type for _bus_dmamap_addseg(). Based on the same fix done for arm64 in r348571.

svn commit: r364456 - stable/12/sys/netpfil/pf

Author: kp Date: Fri Aug 21 13:11:33 2020 New Revision: 364456 URL: https://svnweb.freebsd.org/changeset/base/364456 Log: MFC r355744: pf: Make request_maxcount runtime adjustable There's no reason for this to be a tunable. It's perfectly safe to change this at runtime. Modified:

svn commit: r363796 - in head: lib/libc/riscv/gen sys/riscv/include

/dev/null 00:00:00 1970 (empty, because file is newly added) +++ head/lib/libc/riscv/gen/fpgetmask.c Mon Aug 3 12:48:51 2020 (r363796) @@ -0,0 +1,41 @@ +/*- + * Copyright (c) 2020 Axiado + * All rights reserved. + * + * This software was developed by Kristof Provost under + * sponso

Re: svn commit: r363568 - stable/12/sys/net

On 26 Jul 2020, at 21:21, mike tancsa wrote: Hi Kristof,     First off, thank you for all your efforts in pf and if_bridge.  I have trying to track down a problem with a golang app (sysutils/zrepl) that started acting up around the time the if_bridge stuff was commited (june 26th).  The

svn commit: r363568 - stable/12/sys/net

Author: kp Date: Sun Jul 26 17:44:03 2020 New Revision: 363568 URL: https://svnweb.freebsd.org/changeset/base/363568 Log: Revert bridge epochification Revert r363492, r363491, r363430, r363429 and r362650. The introduction of epoch in the network stack is incomplete in stable/12, and

svn commit: r363491 - stable/12/sys/net

Author: kp Date: Fri Jul 24 20:09:52 2020 New Revision: 363491 URL: https://svnweb.freebsd.org/changeset/base/363491 Log: bridge: Fix mismerges from r360345 In r362650 we merged r360345. This required manual changes due to the differences in EPOCH macros between head and stable/12, and

svn commit: r363492 - stable/12/sys/net

Author: kp Date: Fri Jul 24 20:10:27 2020 New Revision: 363492 URL: https://svnweb.freebsd.org/changeset/base/363492 Log: bridge: Enter epoch for bridge_transmit() Just like the change done for bridge_input()/bridge_output() in r363430 we must enter epoch ourselves, because its coverage

svn commit: r363430 - stable/12/sys/net

Author: kp Date: Wed Jul 22 20:13:12 2020 New Revision: 363430 URL: https://svnweb.freebsd.org/changeset/base/363430 Log: bridge: Enter epoch for bridge_input()/bridge_output() In stable/12 epoch is not as wide as it is in head. The network stack isn't yet in epoch when

svn commit: r363429 - stable/12/sys/net

Author: kp Date: Wed Jul 22 19:43:55 2020 New Revision: 363429 URL: https://svnweb.freebsd.org/changeset/base/363429 Log: MFC r363308: bridge: Don't sleep during epoch While it doesn't trigger INVARIANTS or WITNESS on head it does in stable/12. There's also no reason for it, as we

Re: svn commit: r363308 - head/sys/net

On 18 Jul 2020, at 20:47, Konstantin Belousov wrote: On Sat, Jul 18, 2020 at 12:43:11PM +, Kristof Provost wrote: Author: kp Date: Sat Jul 18 12:43:11 2020 New Revision: 363308 URL: https://svnweb.freebsd.org/changeset/base/363308 Log: bridge: Don't sleep during epoch While

svn commit: r363308 - head/sys/net

Author: kp Date: Sat Jul 18 12:43:11 2020 New Revision: 363308 URL: https://svnweb.freebsd.org/changeset/base/363308 Log: bridge: Don't sleep during epoch While it doesn't trigger INVARIANTS or WITNESS on head it does in stable/12. There's also no reason for it, as we can easily report

svn commit: r362977 - head/sys/riscv/riscv

Author: kp Date: Mon Jul 6 21:29:50 2020 New Revision: 362977 URL: https://svnweb.freebsd.org/changeset/base/362977 Log: riscv plic: Do not complete interrupts until the interrupt handler has run We cannot complete the interrupt (i.e. write to the claims/complete register until the

svn commit: r362853 - in head/sys/riscv: include riscv

Author: kp Date: Wed Jul 1 19:15:43 2020 New Revision: 362853 URL: https://svnweb.freebsd.org/changeset/base/362853 Log: riscv pmap: zero reserved pte bits in ppn The top 10 bits of a pte are reserved by specification[1] and are not part of the PPN. [1] 'Volume II: RISC-V

svn commit: r362852 - head/sys/riscv/riscv

Author: kp Date: Wed Jul 1 19:12:47 2020 New Revision: 362852 URL: https://svnweb.freebsd.org/changeset/base/362852 Log: riscv locore.S: load constant prior to loop A very minor micro-optimization; t0 is not clobbered between the loop top and bottom and there appear to be no other

svn commit: r362851 - head/sys/riscv/riscv

Author: kp Date: Wed Jul 1 19:11:02 2020 New Revision: 362851 URL: https://svnweb.freebsd.org/changeset/base/362851 Log: riscv: Log missing registers in dump_regs() If we panic we dump the registers for debugging. This is very useful, but it missed several registers (ra, sp, gp and tp).

svn commit: r362650 - stable/12/sys/net

Author: kp Date: Fri Jun 26 12:11:22 2020 New Revision: 362650 URL: https://svnweb.freebsd.org/changeset/base/362650 Log: MFC r360345: bridge: epoch-ification Run the bridge datapath under epoch, rather than under the BRIDGE_LOCK(). We still take the BRIDGE_LOCK() whenever we

svn commit: r362649 - stable/12/sys/net

Author: kp Date: Fri Jun 26 10:08:57 2020 New Revision: 362649 URL: https://svnweb.freebsd.org/changeset/base/362649 Log: MFC r359641: bridge: Change lists to CK_LIST as a peparation for epochification Prepare the ground for a rework of the bridge locking approach. We will use an

svn commit: r362648 - stable/12/sys/net

Author: kp Date: Fri Jun 26 09:52:43 2020 New Revision: 362648 URL: https://svnweb.freebsd.org/changeset/base/362648 Log: MFC r358325: bridge: Move locking defines into if_bridge.c The locking defines for if_bridge used to live in if_bridgevar.h, but they're only ever used by the

svn commit: r362235 - head/contrib/llvm-project/clang/lib/Driver/ToolChains/Arch

Author: kp Date: Tue Jun 16 18:39:56 2020 New Revision: 362235 URL: https://svnweb.freebsd.org/changeset/base/362235 Log: llvm: Default to -mno-relax on RISC-V Compiling on a RISC-V system fails with 'relocation R_RISCV_ALIGN requires unimplemented linker relaxation; recompile with

Re: svn commit: r362217 - head/stand/common

On 16 Jun 2020, at 19:11, Ed Maste wrote: On Tue, 16 Jun 2020 at 13:01, Ian Lepore wrote: As much as I prefer doing it this way, style(9) doesn't allow for variable declarations inside a for() statement (or even inside a local block, which is just too 1980s for me, but it is still our

svn commit: r362064 - stable/12/sys/netpfil/pf

Author: kp Date: Thu Jun 11 16:51:13 2020 New Revision: 362064 URL: https://svnweb.freebsd.org/changeset/base/362064 Log: MFC r357061: pf: Apply kif flags to new group members If we have a 'set skip on ' rule this flag it set on the group kif, but must also be set on all members.

svn commit: r361893 - head/tests/sys/net

Author: kp Date: Sun Jun 7 13:53:02 2020 New Revision: 361893 URL: https://svnweb.freebsd.org/changeset/base/361893 Log: bridge tests: Re-enable STP test This test should no longer provoke large amounts of traffic, which can overwhelm single-core systems, preventing them from making

svn commit: r361761 - stable/12/sys/dev/bnxt

Author: kp Date: Wed Jun 3 18:09:31 2020 New Revision: 361761 URL: https://svnweb.freebsd.org/changeset/base/361761 Log: MFC r361279: bnxt: isc_nrxd_max and isc_ntxd_max must be powers of two Modified: stable/12/sys/dev/bnxt/bnxt.h stable/12/sys/dev/bnxt/if_bnxt.c Directory

svn commit: r361762 - stable/11/sys/dev/bnxt

Author: kp Date: Wed Jun 3 18:09:32 2020 New Revision: 361762 URL: https://svnweb.freebsd.org/changeset/base/361762 Log: MFC r361279: bnxt: isc_nrxd_max and isc_ntxd_max must be powers of two Modified: stable/11/sys/dev/bnxt/bnxt.h stable/11/sys/dev/bnxt/if_bnxt.c Modified:

svn commit: r361701 - head/tests/sys/net

Author: kp Date: Mon Jun 1 19:26:16 2020 New Revision: 361701 URL: https://svnweb.freebsd.org/changeset/base/361701 Log: bridge tests: Avoid building a switching loop Enable STP before bringing the bridges up. This avoids a switching loop, which has a tendency to drown out progress in

svn commit: r361279 - head/sys/dev/bnxt

Author: kp Date: Wed May 20 16:07:37 2020 New Revision: 361279 URL: https://svnweb.freebsd.org/changeset/base/361279 Log: bnxt: isc_nrxd_max and isc_ntxd_max must be powers of two Reviewed by: gallatin, rpokala MFC after:2 weeks Differential Revision:

Re: svn commit: r347418 - head/sys/net

On 19 May 2020, at 17:02, Andrew Gallatin wrote: On 2020-05-19 04:21, Kristof Provost wrote: The if_bnxt driver initialises |.isc_nrxd_max = {INT32_MAX, INT32_MAX, INT32_MAX},|, so presumably that’s the cause. I don’t know what a sane value would be though. I’ve defaulted to 4096 (because

Re: svn commit: r347418 - head/sys/net

On 10 May 2019, at 2:41, Eric Joyner wrote: Author: erj Date: Fri May 10 00:41:42 2019 New Revision: 347418 URL: https://svnweb.freebsd.org/changeset/base/347418 Log: iflib: use default ntxd and nrxd when user value is not power of 2 From Jake: A user may set a sysctl to override the

svn commit: r360935 - head/sys/opencrypto

Author: kp Date: Mon May 11 21:42:19 2020 New Revision: 360935 URL: https://svnweb.freebsd.org/changeset/base/360935 Log: opencrypto: Add missing ioctl exit SDTs The opencrypto ioctl code has very useful probe points at the various exit points. These allow us to figure out exactly why a

svn commit: r360867 - stable/11/sys/netpfil/pf

Author: kp Date: Sun May 10 09:50:43 2020 New Revision: 360867 URL: https://svnweb.freebsd.org/changeset/base/360867 Log: MFC r360609: pf: Improve DIOCADDRULE validation We expect the addrwrap.p.dyn value to be set to NULL (and assert such), but do not verify it on input.

svn commit: r360868 - stable/12/sys/netpfil/pf

Author: kp Date: Sun May 10 09:50:44 2020 New Revision: 360868 URL: https://svnweb.freebsd.org/changeset/base/360868 Log: MFC r360609: pf: Improve DIOCADDRULE validation We expect the addrwrap.p.dyn value to be set to NULL (and assert such), but do not verify it on input.

svn commit: r360800 - stable/11/lib/libc/net

Author: kp Date: Thu May 7 21:14:12 2020 New Revision: 360800 URL: https://svnweb.freebsd.org/changeset/base/360800 Log: MFC r360231: libc: Shortcut if_indextoname() if index == 0 If the index we're trying to convert is 0 we can avoid a potentially expensive call to getifaddrs().

svn commit: r360799 - stable/12/lib/libc/net

Author: kp Date: Thu May 7 21:14:11 2020 New Revision: 360799 URL: https://svnweb.freebsd.org/changeset/base/360799 Log: MFC r360231: libc: Shortcut if_indextoname() if index == 0 If the index we're trying to convert is 0 we can avoid a potentially expensive call to getifaddrs().

svn commit: r360609 - head/sys/netpfil/pf

Author: kp Date: Sun May 3 16:09:35 2020 New Revision: 360609 URL: https://svnweb.freebsd.org/changeset/base/360609 Log: pf: Improve DIOCADDRULE validation We expect the addrwrap.p.dyn value to be set to NULL (and assert such), but do not verify it on input. Reported-by:

svn commit: r360608 - stable/11/sys/netpfil/pf

Author: kp Date: Sun May 3 16:06:23 2020 New Revision: 360608 URL: https://svnweb.freebsd.org/changeset/base/360608 Log: MFC r360344: pf: Improve input validation If we pass an anchor name which doesn't exist pfr_table_count() returns -1, which leads to an overflow in mallocarray()

svn commit: r360607 - stable/12/sys/netpfil/pf

Author: kp Date: Sun May 3 16:06:17 2020 New Revision: 360607 URL: https://svnweb.freebsd.org/changeset/base/360607 Log: MFC r360344: pf: Improve input validation If we pass an anchor name which doesn't exist pfr_table_count() returns -1, which leads to an overflow in mallocarray()

svn commit: r360347 - head/sys/netpfil/pf

Author: kp Date: Sun Apr 26 16:30:00 2020 New Revision: 360347 URL: https://svnweb.freebsd.org/changeset/base/360347 Log: pf: Virtualise pf_frag_mtx The pf_frag_mtx mutex protects the fragments queue. The fragments queue is virtualised already (i.e. per-vnet) so it makes no sense to

svn commit: r360346 - head/tests/sys/net

Author: kp Date: Sun Apr 26 16:27:03 2020 New Revision: 360346 URL: https://svnweb.freebsd.org/changeset/base/360346 Log: bridge tests: Test for #216510 We used to have an issue with recursive locking with net.link.bridge.inherit_mac. This causes us to send an ARP request while we hold

svn commit: r360345 - head/sys/net

Author: kp Date: Sun Apr 26 16:22:35 2020 New Revision: 360345 URL: https://svnweb.freebsd.org/changeset/base/360345 Log: bridge: epoch-ification Run the bridge datapath under epoch, rather than under the BRIDGE_LOCK(). We still take the BRIDGE_LOCK() whenever we insert or delete

svn commit: r360344 - head/sys/netpfil/pf

Author: kp Date: Sun Apr 26 16:16:39 2020 New Revision: 360344 URL: https://svnweb.freebsd.org/changeset/base/360344 Log: pf: Improve input validation If we pass an anchor name which doesn't exist pfr_table_count() returns -1, which leads to an overflow in mallocarray() and thus a panic.

svn commit: r360343 - stable/11/sys/netpfil/pf

Author: kp Date: Sun Apr 26 16:13:51 2020 New Revision: 360343 URL: https://svnweb.freebsd.org/changeset/base/360343 Log: MFC r360098: pf: Improve ioctl() input validation Both DIOCCHANGEADDR and DIOCADDADDR take a struct pf_pooladdr from userspace. They failed to validate the dyn

svn commit: r360342 - stable/12/sys/netpfil/pf

Author: kp Date: Sun Apr 26 16:13:50 2020 New Revision: 360342 URL: https://svnweb.freebsd.org/changeset/base/360342 Log: MFC r360098: pf: Improve ioctl() input validation Both DIOCCHANGEADDR and DIOCADDADDR take a struct pf_pooladdr from userspace. They failed to validate the dyn

svn commit: r360341 - stable/12/sbin/pfctl

Author: kp Date: Sun Apr 26 16:06:09 2020 New Revision: 360341 URL: https://svnweb.freebsd.org/changeset/base/360341 Log: MFC r360096: pfctl: Remove unused variable Submitted by: Nick Rogers MFC after:1 week Sponsored by: RG Nets Modified:

svn commit: r360299 - in stable/11/sys: kern net sys

Author: kp Date: Sat Apr 25 12:49:48 2020 New Revision: 360299 URL: https://svnweb.freebsd.org/changeset/base/360299 Log: MFC r360068: ethersubr: Make the mac address generation more robust If we create two (vnet) jails and create a bridge interface in each we end up with the same

svn commit: r360298 - in stable/12/sys: kern net sys

Author: kp Date: Sat Apr 25 12:49:48 2020 New Revision: 360298 URL: https://svnweb.freebsd.org/changeset/base/360298 Log: MFC r360068: ethersubr: Make the mac address generation more robust If we create two (vnet) jails and create a bridge interface in each we end up with the same

svn commit: r360268 - stable/11/sys/netpfil/pf

Author: kp Date: Fri Apr 24 15:27:56 2020 New Revision: 360268 URL: https://svnweb.freebsd.org/changeset/base/360268 Log: MFC r360042: pf: Do not allow negative ps_len in DIOCGETSTATES Userspace may pass a negative ps_len value to us, which causes an assertion failure in malloc().

svn commit: r360267 - stable/12/sys/netpfil/pf

Author: kp Date: Fri Apr 24 15:27:55 2020 New Revision: 360267 URL: https://svnweb.freebsd.org/changeset/base/360267 Log: MFC r360042: pf: Do not allow negative ps_len in DIOCGETSTATES Userspace may pass a negative ps_len value to us, which causes an assertion failure in malloc().

svn commit: r360231 - head/lib/libc/net

Author: kp Date: Thu Apr 23 21:16:51 2020 New Revision: 360231 URL: https://svnweb.freebsd.org/changeset/base/360231 Log: libc: Shortcut if_indextoname() if index == 0 If the index we're trying to convert is 0 we can avoid a potentially expensive call to getifaddrs(). No interface has an

Re: svn commit: r360068 - in head/sys: kern net sys

On 21 Apr 2020, at 4:34, Kyle Evans wrote: On Mon, Apr 20, 2020 at 9:14 PM Kyle Evans wrote: On Mon, Apr 20, 2020 at 8:15 PM Eric van Gyzen wrote: + sz = asprintf(, M_TEMP, "%s-%s-%s", uuid, if_name(ifp), + jailname); + if (sz < 0) { + /* Fall back to a random mac

svn commit: r360099 - head/tests/sys/net

Author: kp Date: Sun Apr 19 16:30:49 2020 New Revision: 360099 URL: https://svnweb.freebsd.org/changeset/base/360099 Log: bridge tests: Ensure that bridges in different jails get different MAC addresses We used to have a problem where bridges created in different vnet jails would end up

svn commit: r360098 - head/sys/netpfil/pf

Author: kp Date: Sun Apr 19 16:10:20 2020 New Revision: 360098 URL: https://svnweb.freebsd.org/changeset/base/360098 Log: pf: Improve ioctl() input validation Both DIOCCHANGEADDR and DIOCADDADDR take a struct pf_pooladdr from userspace. They failed to validate the dyn pointer contained

svn commit: r360097 - head/sbin/pfctl

Author: kp Date: Sun Apr 19 15:37:13 2020 New Revision: 360097 URL: https://svnweb.freebsd.org/changeset/base/360097 Log: pfctl: Call ifa_load() before ifa_grouplookup() ifa_grouplookup() uses the data loaded in ifa_load() (through is_a_group()), so we must call ifa_load() before we can

svn commit: r360096 - head/sbin/pfctl

Author: kp Date: Sun Apr 19 15:32:14 2020 New Revision: 360096 URL: https://svnweb.freebsd.org/changeset/base/360096 Log: pfctl: Remove unused variable Submitted by: Nick Rogers MFC after:1 week Sponsored by: RG Nets Modified: head/sbin/pfctl/pfctl_parser.c Modified:

Re: svn commit: r360068 - in head/sys: kern net sys

On 19 Apr 2020, at 15:33, Ronald Klop wrote: On Sat, 18 Apr 2020 09:50:30 +0200, Kristof Provost wrote: Author: kp Date: Sat Apr 18 07:50:30 2020 New Revision: 360068 URL: https://svnweb.freebsd.org/changeset/base/360068 Log: ethersubr: Make the mac address generation more robust If we

svn commit: r360069 - head/sys/net

Author: kp Date: Sat Apr 18 08:00:58 2020 New Revision: 360069 URL: https://svnweb.freebsd.org/changeset/base/360069 Log: bridge: Simplify mac address generation Unconditionally use ether_gen_addr() to generate bridge mac addresses. This function is now less likely to generate duplicate

svn commit: r360068 - in head/sys: kern net sys

Author: kp Date: Sat Apr 18 07:50:30 2020 New Revision: 360068 URL: https://svnweb.freebsd.org/changeset/base/360068 Log: ethersubr: Make the mac address generation more robust If we create two (vnet) jails and create a bridge interface in each we end up with the same mac address on both

svn commit: r360043 - head/tests/sys/net

Author: kp Date: Fri Apr 17 14:57:15 2020 New Revision: 360043 URL: https://svnweb.freebsd.org/changeset/base/360043 Log: bridge tests: Test deleting a bridge with members Reviewed by: philip, emaste Sponsored by: The FreeBSD Foundation Differential Revision:

  1   2   3   4   5   6   7   8   >