Author: gallatin
Date: Sat Dec 19 22:04:46 2020
New Revision: 368819
URL: https://svnweb.freebsd.org/changeset/base/368819
Log:
Filter TCP connections to SO_REUSEPORT_LB listen sockets by NUMA domain
In order to efficiently serve web traffic on a NUMA
machine, one must avoid as many NUMA
Author: gallatin
Date: Sat Dec 19 21:46:09 2020
New Revision: 368818
URL: https://svnweb.freebsd.org/changeset/base/368818
Log:
Optionally bind ktls threads to NUMA domains
When ktls_bind_thread is 2, we pick a ktls worker thread that is
bound to the same domain as the TCP connection
On 12/17/20 2:49 PM, Kyle Evans wrote:
On Thu, Dec 17, 2020 at 1:47 PM Andrew Gallatin wrote:
On 12/17/20 12:02 PM, Warner Losh wrote:
Author: imp
Date: Thu Dec 17 17:02:09 2020
New Revision: 368721
URL:
https://urldefense.com/v3/__https://svnweb.freebsd.org/changeset/base/368721__;!!OToaGQ
On 12/17/20 12:02 PM, Warner Losh wrote:
Author: imp
Date: Thu Dec 17 17:02:09 2020
New Revision: 368721
URL:
https://urldefense.com/v3/__https://svnweb.freebsd.org/changeset/base/368721__;!!OToaGQ!5c1mLnhtRtEV6Cv_MTWpzXWaGZEYYDp4TJ6wVDzjVZiehAItts7ZWC15uNnQYRa5Fg$
Log:
Drop
Author: gallatin
Date: Wed Nov 18 14:55:49 2020
New Revision: 367797
URL: https://svnweb.freebsd.org/changeset/base/367797
Log:
LACP: When suppressing distributing, return ENOBUFS
When links come and go, lacp goes into a "suppress distributing" mode
where it drops traffic for 3 seconds.
On 11/2/20 8:19 PM, Conrad Meyer wrote:
Log:
linux(4): Emulate Linux SOL_SOCKET:SO_PASSCRED
This is required by some major linux applications, such as Chrome and
Firefox. (As well as Electron-using applications, which are essentially
a bundled version of Chrome.)
On 2020-09-02 22:42, Alexey Dokuchaev wrote:
I want to understand which rules have to be followed (and why).
In general, FreeBSD code we write should follow style(9); it specifically
mentions "do not add whitespace at the end of a line" and "... followed by
one blank line" but doesn't go as
Author: gallatin
Date: Fri Sep 4 17:36:15 2020
New Revision: 365331
URL: https://svnweb.freebsd.org/changeset/base/365331
Log:
ktls: Check for a NULL send tag in ktls_cleanup()
When using ifnet ktls, and when ktls_reset_send_tag()
fails to allocate a replacement tag, it leaves
the tls
Author: gallatin
Date: Mon Aug 31 13:53:14 2020
New Revision: 364986
URL: https://svnweb.freebsd.org/changeset/base/364986
Log:
make m_getm2() resilient to zone_jumbop exhaustion
When the zone_jumbop is exhausted, most things using
using sosend* (like sshd) will eventually
fail or
On 2020-08-24 18:48, Matt Macy wrote:
Log:
Vendor import of openzfs master @ 184df27eef0abdc7ab2105b21257f753834b936b
Sponsored by: iX Systems, Inc.
Just wanted to say "thank you" to you and iX for this. I'm glad this has
landed!
Drew
Author: gallatin
Date: Fri Aug 21 18:31:57 2020
New Revision: 364460
URL: https://svnweb.freebsd.org/changeset/base/364460
Log:
uma: record allocation failures due to zone limits
The zone limit mechanism was recently reworked, and
allocation failures due to limits being exceeded
were
Author: gallatin
Date: Wed Aug 19 17:59:06 2020
New Revision: 364405
URL: https://svnweb.freebsd.org/changeset/base/364405
Log:
TCP: remove special treatment for hardware (ifnet) TLS
Remove most special treatment for ifnet TLS in the TCP stack, except
for code to avoid mixing handshakes
Author: gallatin
Date: Thu Jul 9 17:10:18 2020
New Revision: 363049
URL: https://svnweb.freebsd.org/changeset/base/363049
Log:
MFC r362112
x86: Bump default msi/msix vector limit to 2048
Modified:
stable/12/sys/x86/x86/msi.c
Directory Properties:
stable/12/ (props changed)
Modified:
Author: gallatin
Date: Mon Jun 29 21:35:50 2020
New Revision: 362789
URL: https://svnweb.freebsd.org/changeset/base/362789
Log:
Fix a panic when unloading firmware
LIST_FOREACH_SAFE() is not safe in the presence
of other threads removing list entries when a
mutex is released.
This
Author: gallatin
Date: Fri Jun 12 18:41:12 2020
New Revision: 362112
URL: https://svnweb.freebsd.org/changeset/base/362112
Log:
x86: Bump default msi/msix vector limit to 2048
Given that 64c/128t CPUs are currently available, and that many
devices (nvme, many NICs) desire to map 1 MSI-X
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 that’s what some other iflib users seems to do) for
Author: gallatin
Date: Tue May 12 17:18:44 2020
New Revision: 360982
URL: https://svnweb.freebsd.org/changeset/base/360982
Log:
IPv6: Fix a panic in the nd6 code with unmapped mbufs.
If the neighbor entry for an IPv6 TCP session using unmapped
mbufs times out, IPv6 will send an icmp6
Author: gallatin
Date: Tue May 12 14:01:12 2020
New Revision: 360961
URL: https://svnweb.freebsd.org/changeset/base/360961
Log:
IPv6: sync IP_NO_SND_TAG_RL support from IPv4
The IP_NO_SND_TAG_RL flag to ip{,6}_output() means that the packets
being sent should bypass hardware rate
Author: gallatin
Date: Mon May 11 21:23:22 2020
New Revision: 360930
URL: https://svnweb.freebsd.org/changeset/base/360930
Log:
Fix the build
Back out the IPv6 portion of r360903, as the stamp_tag param
is apparently not supported in upstream FreeBSD.
Sponsored by: Netflix
Pointy
Author: gallatin
Date: Mon May 11 19:17:33 2020
New Revision: 360914
URL: https://svnweb.freebsd.org/changeset/base/360914
Log:
Ktls: never skip stamping tags for NIC TLS
The newer RACK and BBR TCP stacks have added a mechanism
to disable hardware packet pacing for TCP retransmits.
Author: gallatin
Date: Tue Apr 14 14:48:00 2020
New Revision: 359920
URL: https://svnweb.freebsd.org/changeset/base/359920
Log:
Bump FreeBSD version after r359919 (KTLS / unmapped mbuf changes)
The above changes mbufs, and any module using unmapped mbufs
would need to be re-compiled.
Author: gallatin
Date: Tue Apr 14 14:46:06 2020
New Revision: 359919
URL: https://svnweb.freebsd.org/changeset/base/359919
Log:
KTLS: Re-work unmapped mbufs to carry ext_pgs in the mbuf itself.
While the original implementation of unmapped mbufs was a large
step forward in terms of
Author: gallatin
Date: Mon Apr 13 23:06:56 2020
New Revision: 359908
URL: https://svnweb.freebsd.org/changeset/base/359908
Log:
lagg: stop double-counting output errors and counting drops as errors
Before this change, lagg double-counted errors from lagg members, and counted
every drop
Author: gallatin
Date: Mon Mar 30 23:29:53 2020
New Revision: 359474
URL: https://svnweb.freebsd.org/changeset/base/359474
Log:
KTLS: Coalesce adjacent TLS trailers & headers to improve PCIe bus efficiency
KTLS uses the embedded header and trailer fields of unmapped
mbufs. This can lead
Author: gallatin
Date: Mon Mar 16 14:03:27 2020
New Revision: 359016
URL: https://svnweb.freebsd.org/changeset/base/359016
Log:
Avoid a cache miss accessing an mbuf ext_pgs pointer when doing SW kTLS.
For a Netflix 90Gb/s 100% TLS software kTLS workload, this reduces
the CPI of
On 2020-03-09 09:44, Andrew Gallatin wrote:
Author: gallatin
Date: Mon Mar 9 13:44:51 2020
New Revision: 358808
URL:
https://urldefense.com/v3/__https://svnweb.freebsd.org/changeset/base/358808__;!!OToaGQ!5mmRl2ROq7G4c4x2Xe2uHppYyETGlCRsREj-jHw0ZWcNqt3GhQju3BHBkM_vsrVvkQ$
Log:
make lacp's
Author: gallatin
Date: Mon Mar 9 13:44:51 2020
New Revision: 358808
URL: https://svnweb.freebsd.org/changeset/base/358808
Log:
make lacp's use_numa hashing aware of send tags
When I did the use_numa support, I missed the fact that there is
a separate hash function for send tag nic
Author: gallatin
Date: Sat Jan 18 18:25:37 2020
New Revision: 356866
URL: https://svnweb.freebsd.org/changeset/base/356866
Log:
pcpu_page_alloc: guard against empty NUMA domains
Some systems, such as higher end Threadripper, may have
NUMA domains with no physical memory, Don't allocate
Author: gallatin
Date: Thu Nov 7 19:54:24 2019
New Revision: 354470
URL: https://svnweb.freebsd.org/changeset/base/354470
Log:
hwpmc : fix AMD perf counter MSR access
- amd_intr() does not account for the offset (0x200) in the counter
MSR address and ends up accessing invalid regions
Author: gallatin
Date: Mon Nov 4 19:30:19 2019
New Revision: 354338
URL: https://svnweb.freebsd.org/changeset/base/354338
Log:
Add tunable to allow interrupts on hyperthreaded cores
Enabling interrupts on htt cores has benefits to workloads which are primarily
interrupt driven by
Author: gallatin
Date: Thu Oct 24 18:39:05 2019
New Revision: 354029
URL: https://svnweb.freebsd.org/changeset/base/354029
Log:
Add a tunable to set the pgcache zone's maxcache
When it is set to 0 (the default), a heavy Netflix-style web workload
suffers from heavy lock contention on the
Author: gallatin
Date: Fri Sep 27 20:08:19 2019
New Revision: 352816
URL: https://svnweb.freebsd.org/changeset/base/352816
Log:
kTLS: Fix a bug where we would not encrypt anon data inplace.
Software Kernel TLS needs to allocate a new destination crypto
buffer when encrypting data from
Author: gallatin
Date: Fri Sep 27 19:17:40 2019
New Revision: 352814
URL: https://svnweb.freebsd.org/changeset/base/352814
Log:
kTLS support for TLS 1.3
TLS 1.3 requires a few changes because 1.3 pretends to be 1.2
with a record type of application data. The "real" record type is
then
Author: gallatin
Date: Fri Sep 20 09:36:07 2019
New Revision: 352552
URL: https://svnweb.freebsd.org/changeset/base/352552
Log:
remove redundant "ktls" in KTLS thr name
This reducesthe string width of the ktls thread name
and improves "ps" output.
Glanced at by: jhb
Event:
On 2019-09-11 17:16, Conrad Meyer wrote:
Small nitpick:
On Wed, Sep 11, 2019 at 11:48 AM Andrew Gallatin wrote:
Note that on a system under a syn flood attack, arc4random()
becomes quite expensive, and the chacha_poly crypto that it calls
arc4random uses chacha20 — there is no "
Author: gallatin
Date: Wed Sep 11 18:48:26 2019
New Revision: 352228
URL: https://svnweb.freebsd.org/changeset/base/352228
Log:
Avoid unneeded call to arc4random() in syncache_add()
Don't call arc4random() unconditionally to initialize sc_iss, and
then when syncookies are enabled, just
Author: gallatin
Date: Mon Apr 22 19:24:21 2019
New Revision: 346579
URL: https://svnweb.freebsd.org/changeset/base/346579
Log:
Track device's NUMA domain in ifnet & alloc ifnet from NUMA local memory
This commit adds new if_alloc_domain() and if_alloc_dev() methods to
allocate ifnets.
Author: gallatin
Date: Wed Apr 24 13:32:04 2019
New Revision: 346632
URL: https://svnweb.freebsd.org/changeset/base/346632
Log:
iflib: Add pfil hooks
As with mlx5en, the idea is to drop unwanted traffic as early
in receive as possible, before mbufs are allocated and anything
is passed
Author: gallatin
Date: Tue Apr 16 16:49:34 2019
New Revision: 346281
URL: https://svnweb.freebsd.org/changeset/base/346281
Log:
Replace cosqos with numa_domain in mbuf pkthdr
The cosqos field was added nearly 6 years ago in r254804, and it is
still unused by any in-tree consumers. I
Author: gallatin
Date: Mon Apr 15 17:14:50 2019
New Revision: 346247
URL: https://svnweb.freebsd.org/changeset/base/346247
Log:
mlx5en: Enable new pfil(9) KPI ethernet filtering hooks
This allows efficient filtering at packet ingress on mlx5en.
Note that the packets are filtered (and
On 2019-08-18 19:44, Jeff Roberson wrote:
Author: jeff
Date: Sun Aug 18 23:44:23 2019
New Revision: 351200
<..>
Log:
Allocate all per-cpu datastructures in domain correct memory.
Reviewed by: kib, gallatin (some objections)
No objection to what you actually committed. The only
Author: gallatin
Date: Fri Aug 2 00:13:11 2019
New Revision: 350527
URL: https://svnweb.freebsd.org/changeset/base/350527
Log:
MFC r350245
pciconf: report PCI Gen4 speeds
PCIe gen4 runs at 16GT/s. Report this as
the speed of Gen4 links.
Reviewed by:imp
Sponsored by:
Author: gallatin
Date: Tue Jul 23 16:28:17 2019
New Revision: 350245
URL: https://svnweb.freebsd.org/changeset/base/350245
Log:
pciconf: report PCI Gen4 speeds
PCIe gen4 runs at 16GT/s. Report this as
the speed of Gen4 links.
Reviewed by: imp
MFC after:7 days
Sponsored by:
On 2019-06-15 11:59, Marius Strobl wrote:
On Sat, Jun 15, 2019 at 09:08:05AM -0400, Andrew Gallatin wrote:
On 2019-06-15 07:07, Marius Strobl wrote:
Author: marius
Date: Sat Jun 15 11:07:41 2019
New Revision: 349055
Log:
- Replace unused and only ever written to members of public iflib
On 2019-06-15 07:07, Marius Strobl wrote:
Author: marius
Date: Sat Jun 15 11:07:41 2019
New Revision: 349055
Log:
- Replace unused and only ever written to members of public iflib(9)
structs with placeholders (in the latter case, IFLIB_MAX_TX_BYTES
etc. are also only ever used
On 2019-05-24 11:45, Mark Johnston wrote:
Modernize the MAKE_JUST_KERNELS hint in the top-level makefile.
It doesn't make sense to limit to -j12 anymore, build scalability
is better than it used to be. Fold the hint into the description
of the universe target.
Reviewed
On 2019-05-22 13:09, Andriy Gapon wrote:
On 22/05/2019 16:44, Andrew Gallatin wrote:
This is needed for AMD SMCA processors, as SMCA uses different
MSR address for access MCA banks.
Just curious, what is SMCA?
" Scalable Machine Check Architecture "
Author: gallatin
Date: Wed May 22 13:44:15 2019
New Revision: 348109
URL: https://svnweb.freebsd.org/changeset/base/348109
Log:
x86 MCA: introduce MCA hooks for different vendor implementations
This is needed for AMD SMCA processors, as SMCA uses different
MSR address for access MCA
On 2019-05-10 11:50, Kristof Provost wrote:
On 10 May 2019, at 8:31, Andrew Gallatin wrote:
On 2019-05-10 08:44, Slawa Olhovchenkov wrote:
pf have ifdef for IPSEC, but don't have support IPSEC_SUPPORT
(netpfil/pf/if_pfsync.c).
Thanks for pointing this out. It seems
On 2019-05-10 08:44, Slawa Olhovchenkov wrote:
pf have ifdef for IPSEC, but don't have support IPSEC_SUPPORT
(netpfil/pf/if_pfsync.c).
Thanks for pointing this out. It seems like IPSEC_SUPPORT would work
for this. I've made a patch, and it compiles and the pf module loads.
However, I
Author: gallatin
Date: Fri May 10 13:41:19 2019
New Revision: 347430
URL: https://svnweb.freebsd.org/changeset/base/347430
Log:
Bind TCP HPTS (pacer) threads to NUMA domains
Bind the TCP pacer threads to NUMA domains and build per-domain
pacer-thread lookup tables. These tables allow us
Author: gallatin
Date: Thu May 9 22:38:15 2019
New Revision: 347410
URL: https://svnweb.freebsd.org/changeset/base/347410
Log:
Remove IPSEC from GENERIC due to performance issues
Having IPSEC compiled into the kernel imposes a non-trivial
performance penalty on multi-threaded workloads
Author: gallatin
Date: Fri May 3 14:43:21 2019
New Revision: 347055
URL: https://svnweb.freebsd.org/changeset/base/347055
Log:
Select lacp egress ports based on NUMA domain
This change creates an array of port maps indexed by numa domain
for lacp port selection. If we have lacp
On 2019-04-29 10:54, Emmanuel Vadot wrote:
On Mon, 29 Apr 2019 10:49:01 -0400
Andrew Gallatin wrote:
On 2019-04-29 10:21, Rodney W. Grimes wrote:
On Tue, 23 Apr 2019 at 13:26, Rodney W. Grimes
wrote:
Very cool, now how do I get a PCIe slot into a RPI3!!! lol :-)
I know you're joking
On 2019-04-29 10:21, Rodney W. Grimes wrote:
On Tue, 23 Apr 2019 at 13:26, Rodney W. Grimes
wrote:
Very cool, now how do I get a PCIe slot into a RPI3!!! lol :-)
I know you're joking but the comment does highlight an issue in the
AArch64 world - there's a lack of good mid-range developer
Author: gallatin
Date: Thu Apr 25 15:37:28 2019
New Revision: 346677
URL: https://svnweb.freebsd.org/changeset/base/346677
Log:
Track TCP connection's NUMA domain in the inpcb
Drivers can now pass up numa domain information via the
mbuf numa domain field. This information is then used
Author: gallatin
Date: Wed Apr 24 13:32:04 2019
New Revision: 346632
URL: https://svnweb.freebsd.org/changeset/base/346632
Log:
iflib: Add pfil hooks
As with mlx5en, the idea is to drop unwanted traffic as early
in receive as possible, before mbufs are allocated and anything
is passed
Author: gallatin
Date: Mon Apr 22 19:24:21 2019
New Revision: 346579
URL: https://svnweb.freebsd.org/changeset/base/346579
Log:
Track device's NUMA domain in ifnet & alloc ifnet from NUMA local memory
This commit adds new if_alloc_domain() and if_alloc_dev() methods to
allocate ifnets.
Author: gallatin
Date: Tue Apr 16 16:49:34 2019
New Revision: 346281
URL: https://svnweb.freebsd.org/changeset/base/346281
Log:
Replace cosqos with numa_domain in mbuf pkthdr
The cosqos field was added nearly 6 years ago in r254804, and it is
still unused by any in-tree consumers. I
Author: gallatin
Date: Mon Apr 15 17:14:50 2019
New Revision: 346247
URL: https://svnweb.freebsd.org/changeset/base/346247
Log:
mlx5en: Enable new pfil(9) KPI ethernet filtering hooks
This allows efficient filtering at packet ingress on mlx5en.
Note that the packets are filtered (and
Author: gallatin
Date: Mon Mar 18 12:41:42 2019
New Revision: 345273
URL: https://svnweb.freebsd.org/changeset/base/345273
Log:
Fix a typo introduced in r344133
The line was misedited to change tt to st instead of
changing ut to st.
The use of st as the denominator in
On 3/14/19 11:36 PM, Rodney W. Grimes wrote:
[ Charset UTF-8 unsupported, converting... ]
On Thu, 14 Mar 2019 at 22:39, Rodney W. Grimes
wrote:
4. There is no easy way to show
"changed byte at offset 0x432 from 0xef to 0xfe"
How do we represent Copyright and License in such objects?
On 3/5/19 4:06 PM, Matthew Macy wrote:
This represents a misunderstanding of how defines are used. This left
the option open to the user to enable the use of larger than page size
buffers as it does enable better performance. Over the course of a
long uptime memory can get too fragmented.
I think the misunderstanding here is that I think he's not getting the
ifp from the route.
My recollection is that he is holding the ifps when he enables HW pacing
in BBR. Due to limitations in different NIC hardware, you can only have
N different rates, etc. So he goes ahead and allocates
Author: gallatin
Date: Mon Jan 28 14:34:59 2019
New Revision: 343526
URL: https://svnweb.freebsd.org/changeset/base/343526
Log:
MFC r343430
Fix an iflib driver unload panic introduced in r343085
The new loop to sync and unload descriptors was indexed
by "i", rather than "j".
Author: gallatin
Date: Fri Jan 25 15:02:18 2019
New Revision: 343430
URL: https://svnweb.freebsd.org/changeset/base/343430
Log:
Fix an iflib driver unload panic introduced in r343085
The new loop to sync and unload descriptors was indexed
by "i", rather than "j". The panic was caused
Author: gallatin
Date: Tue Jan 22 17:34:53 2019
New Revision: 343304
URL: https://svnweb.freebsd.org/changeset/base/343304
Log:
MFC r341095:
Use busdma unconditionally in iflib
- Remove the complex mechanism to choose between using busdma
and raw pmap_kextract at runtime.
On 1/21/19 1:42 PM, Navdeep Parhar wrote:
Log:
cxgbe(4): Allow negative values in hw.cxgbe.fw_install and take them to
Thank you!
Drew
___
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To
On 1/14/19 8:02 PM, Gleb Smirnoff wrote:
Log:
Allocate pager bufs from UMA instead of 80-ish mutex protected linked list.
<...>
Together with: gallatin
Thank you so much for carrying this over the finish line!
Drew
___
Author: gallatin
Date: Fri Jan 4 18:38:27 2019
New Revision: 342774
URL: https://svnweb.freebsd.org/changeset/base/342774
Log:
Limit git history searches in newvers.sh
newvers.sh takes upwards of 4-5 seconds to complete on trees checked
out from github, due to searching the entire
On 12/17/18 2:08 PM, Bruce Evans wrote:
On Mon, 17 Dec 2018, Andrew Gallatin wrote:
On 12/5/18 9:20 AM, Slava Shwartsman wrote:
Author: slavash
Date: Wed Dec 5 14:20:57 2018
New Revision: 341578
URL:
https://urldefense.proofpoint.com/v2/url?u=https
On 12/5/18 9:20 AM, Slava Shwartsman wrote:
Author: slavash
Date: Wed Dec 5 14:20:57 2018
New Revision: 341578
URL:
Author: gallatin
Date: Tue Nov 27 20:01:05 2018
New Revision: 341095
URL: https://svnweb.freebsd.org/changeset/base/341095
Log:
Use busdma unconditionally in iflib
- Remove the complex mechanism to choose between using busdma
and raw pmap_kextract at runtime. The reduced complexity
On 11/2/18 11:43 PM, Matt Macy wrote:
Author: mmacy
Date: Sat Nov 3 03:43:32 2018
New Revision: 340097
URL:
Author: gallatin
Date: Mon Oct 1 14:14:21 2018
New Revision: 339043
URL: https://svnweb.freebsd.org/changeset/base/339043
Log:
Allow empty NUMA memory domains to support Threadripper2
The AMD Threadripper 2990WX is basically a slightly crippled Epyc.
Rather than having 4 memory
On 10/1/18 10:14 AM, Andrew Gallatin wrote:
Author: gallatin
Date: Mon Oct 1 14:14:21 2018
New Revision: 339043
URL:
https://urldefense.proofpoint.com/v2/url?u=https-3A__svnweb.freebsd.org_changeset_base_339043=DwIDaQ=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc=Ed
Author: gallatin
Date: Mon Aug 27 18:13:20 2018
New Revision: 338341
URL: https://svnweb.freebsd.org/changeset/base/338341
Log:
Reject IPv4 SO_REUSEPORT_LB groups when looking up an IPv6 listening socket
Similar to how the IPv4 code will reject an IPv6 LB group,
we must ignore IPv4 LB
It could probably be MFCed if somebody could verify that it causes no
harm in 11.
I have no way to test lagg/lacp on 11, so I did not mark it for MFC.
Drew
On 8/13/18 9:58 PM, Kubilay Kocak wrote:
On 14/08/2018 12:13 am, Andrew Gallatin wrote:
Author: gallatin
Date: Mon Aug 13 14:13:25 2018
Author: gallatin
Date: Mon Aug 13 14:13:25 2018
New Revision: 337709
URL: https://svnweb.freebsd.org/changeset/base/337709
Log:
lagg: allow lacp to manage the link state
Lacp needs to manage the link state itself. Unlike other
lagg protocols, the ability of lacp to pass traffic
depends
On 07/05/18 19:59, John Baldwin wrote:
You misunderstand. /usr/local/sys/modules would hold module sources so that
they can be recompiled when building a kernel without having to rebuild the
package or reinstall the package. Binary modules would continue to be
installed in /boot/modules.
On 07/06/18 15:33, Navdeep Parhar wrote:
Log:
cxgbe(4): Assume that any unknown flash on the card is 4MB and has 64KB
sectors, instead of refusing to attach to the card.
Thank you!
Drew
___
svn-src-all@freebsd.org mailing list
On 07/05/18 17:14, Rick Macklem wrote:
Andrew Gallatin wrote:
On 7/4/18 9:20 PM, Rodney W. Grimes wrote:
[stuff snipped]
It is using a magic constant twice, where one has a
derived value that is dependent on the value of the other.
That is bad and error prone and does not document that
one
Author: gallatin
Date: Thu Jul 5 02:43:10 2018
New Revision: 335973
URL: https://svnweb.freebsd.org/changeset/base/335973
Log:
mxge: replace 65536 with IP_MAXPACKET in tso settings.
Modified:
head/sys/dev/mxge/if_mxge.c
Modified: head/sys/dev/mxge/if_mxge.c
On 7/4/18 9:20 PM, Rodney W. Grimes wrote:
On 07/04/18 15:46, Rodney W. Grimes wrote:
Author: gallatin
Date: Wed Jul 4 19:29:06 2018
New Revision: 335967
URL:
On 07/04/18 15:46, Rodney W. Grimes wrote:
Author: gallatin
Date: Wed Jul 4 19:29:06 2018
New Revision: 335967
URL:
Author: gallatin
Date: Wed Jul 4 19:29:06 2018
New Revision: 335967
URL: https://svnweb.freebsd.org/changeset/base/335967
Log:
mxge: choose appropriate values for hw tso
Modified:
head/sys/dev/mxge/if_mxge.c
Modified: head/sys/dev/mxge/if_mxge.c
On 07/04/18 14:54, Andrew Gallatin wrote:
mxge: Add SIOCGI2C support for devices with SFP/XFP cages
Note that I do not have any XFP devices to test with, only SFP
and CX4. If this causes problems for XFP devices, I can
restrict SIOCGI2C support to just SFP if needed.
Drew
Author: gallatin
Date: Wed Jul 4 18:54:44 2018
New Revision: 335966
URL: https://svnweb.freebsd.org/changeset/base/335966
Log:
mxge: Add SIOCGI2C support for devices with SFP/XFP cages
Modified:
head/sys/dev/mxge/if_mxge.c
Modified: head/sys/dev/mxge/if_mxge.c
Author: gallatin
Date: Wed Jul 4 14:25:38 2018
New Revision: 335957
URL: https://svnweb.freebsd.org/changeset/base/335957
Log:
mxge: fix panic at module unload
r333175 (multicast changes) exposed a bug where
mxge was not checking to see if the driver was being
unloaded while handing
On 05/24/18 06:18, Navdeep Parhar wrote:
Log:
cxgbe(4): Data path for rate-limited tx.
This is hardware support for the SO_MAX_PACING_RATE sockopt (see
setsockopt(2)), which is available in kernels built with "options
RATELIMIT".
Relnotes: Yes
Sponsored by:
Author: gallatin
Date: Fri May 18 14:14:04 2018
New Revision: 333793
URL: https://svnweb.freebsd.org/changeset/base/333793
Log:
Teach pmcannotate about $TMPDIR and _PATH_TMP
Convert pmcannotate to using $TMPDIR and _PATH_TMP rather than hard
coding /tmp for temporary files. Pmcannotate
On 05/17/18 14:07, Mark Johnston wrote:
On Thu, May 17, 2018 at 10:07:34AM -0700, Conrad Meyer wrote:
On Wed, May 16, 2018 at 9:27 PM, Mark Johnston wrote:
Author: markj
Date: Thu May 17 04:27:08 2018
New Revision: 333703
URL:
Author: gallatin
Date: Tue May 15 23:55:38 2018
New Revision: 333655
URL: https://svnweb.freebsd.org/changeset/base/333655
Log:
Unhook DEBUG_BUFRING from INVARIANTS
Some of the DEBUG_BUFRING checks are racy, and can lead to
spurious assertions when run under high load. Unhook these
On 05/10/18 20:11, Ed Maste wrote:
On 10 May 2018 at 20:00, Andrew Gallatin <galla...@cs.duke.edu> wrote:
Unfortunately, it looks like this method will get blown away by an
installworld:
Ah. You can set WITH_LLD_IS_LD in /etc/src.conf and installworld will
install ld as a symlink to
On 05/10/18 19:14, Ed Maste wrote:
On 10 May 2018 at 18:24, Andrew Gallatin <galla...@cs.duke.edu> wrote:
Rather than erroring out, we please just use the appropriate linker?
That's my goal, but it's a bit of an involved change and will take
some time to make sure we don't introdu
On 05/10/18 16:10, Ed Maste wrote:
Author: emaste
Date: Thu May 10 20:10:02 2018
New Revision: 333470
URL:
Author: gallatin
Date: Thu May 10 16:19:41 2018
New Revision: 333462
URL: https://svnweb.freebsd.org/changeset/base/333462
Log:
Fix a panic in the IPv6 multicast code.
Use LIST_FOREACH_SAFE in in6m_disconnect() since we're
deleting and freeing item from the membership list
while
On 05/10/18 07:36, Ed Maste wrote:
Author: emaste
Date: Thu May 10 11:36:16 2018
New Revision: 333457
URL:
Author: gallatin
Date: Thu May 10 13:19:42 2018
New Revision: 333459
URL: https://svnweb.freebsd.org/changeset/base/333459
Log:
Fix the build after r333457
In r333457, the arguments to kern_pwritev() were accidentally
re-ordered as part of ANSIfication, breaking the build.
Modified:
Author: gallatin
Date: Mon May 7 18:11:22 2018
New Revision: 29
URL: https://svnweb.freebsd.org/changeset/base/29
Log:
Fix an off-by-one error when deciding to request a tx interrupt
The canonical check for whether or not a ring is drainable is
TXQ_AVAIL() > MAX_TX_DESC() + 2.
1 - 100 of 220 matches
Mail list logo