Sometimes the software loses the race when appending more descriptors to
the tx ring and the tx queue stops.
This commit detects this condition and restart the tx queue whenever it stall.
Tested by: sobomax@, Keith White <kwhite@site.uottawa.ca>,
Paul Mather <paul@gromit.dlib.vt.edu>
Sponsored by: Rubicon Communications (Netgate)
Approved by: re (kib)
of CPUs present. On amd64 this unbreaks the boot for systems with 92 or
more CPUs; the limit will vary on other systems depending on the size of
their uma_zone and uma_cache structures.
The major consumer of pages during UMA startup is the 19 zone structures
which are set up before UMA has bootstrapped itself sufficiently to use
the rest of the available memory: UMA Slabs, UMA Hash, 4 / 6 / 8 / 12 /
16 / 32 / 64 / 128 / 256 Bucket, vmem btag, VM OBJECT, RADIX NODE, MAP,
KMAP ENTRY, MAP ENTRY, VMSPACE, and fakepg. If the zone structures occupy
more than one page, they will not share pages and the number of pages
currently needed for startup is 19 * pages_per_zone + N, where N is the
number of pages used for allocating other structures; on amd64 N = 3 at
present (2 pages are allocated for UMA Kegs, and one page for UMA Hash).
This patch adds a new definition UMA_BOOT_PAGES_ZONES, currently set to 32,
and if a zone structure does not fit into a single page sets boot_pages to
UMA_BOOT_PAGES_ZONES * pages_per_zone instead of UMA_BOOT_PAGES (which
remains at 64). Consequently this patch has no effect on systems where the
zone structure fits into 2 or fewer pages (on amd64, 59 or fewer CPUs), but
increases boot_pages sufficiently on systems where the large number of CPUs
makes this structure larger. It seems safe to assume that systems with 60+
CPUs can afford to set aside an additional 128kB of memory per 32 CPUs.
The vm.boot_pages tunable continues to override this computation, but is
unlikely to be necessary in the future.
Tested on: EC2 x1.32xlarge
Relnotes: FreeBSD can now boot on 92+ CPU systems without requiring
vm.boot_pages to be manually adjusted.
Reviewed by: jeff, alc, adrian
Approved by: re (kib)
any open vnodes before proceeding. Make autounmound(8) use this flag.
Without it, even an unsuccessfull unmount causes filesystem flush,
which interferes with normal operation.
Reviewed by: kib@
Approved by: re (gjb@)
MFC after: 1 month
Differential Revision: https://reviews.freebsd.org/D7047
Due to ARC initial configuration not being done and kmem information
not being available we need to blindly set zfs_arc_max and zfs_arc_min
when configured via the tunable.
This fixes vfs.zfs.arc_(min|max) configuration via loader.conf broken by
r302265.
Approved by: re(gjb)
MFC after: 1 week
to INTRNG in r301565 with the old code no longer being built by default with
no reports of issues on any supported hardware.
Approved by: re (gjb)
Obtained from: ABT Systems Ltd
Sponsored by: The FreeBSD Foundation
debugging TCP connections. This commit provides a mechanism to free those
mbufs when the system is under memory pressure.
Because this will result in lost debugging information, the behavior is
controllable by a sysctl. The default setting is to free the mbufs.
Reviewed by: gnn
Approved by: re (gjb)
Differential Revision: https://reviews.freebsd.org/D6931
Input from: novice_techie.com
mp_maxid or CPU_FOREACH() as appropriate. This fixes a number of places in
the kernel that assumed CPU IDs are dense in [0, mp_ncpus) and would try,
for example, to run tasks on CPUs that did not exist or to allocate too
few buffers on systems with sparse CPU IDs in which there are holes in the
range and mp_maxid > mp_ncpus. Such circumstances generally occur on
systems with SMT, but on which SMT is disabled. This patch restores system
operation at least on POWER8 systems configured in this way.
There are a number of other places in the kernel with potential problems
in these situations, but where sparse CPU IDs are not currently known
to occur, mostly in the ARM machine-dependent code. These will be fixed
in a follow-up commit after the stable/11 branch.
PR: kern/210106
Reviewed by: jhb
Approved by: re (glebius)
doing the teardown. ipf_destroy_all() may free ipfmain in case
of ipf_dynamic_softc being true, thus we are avoiding a possible
memory modified after free as well.
Reported by: Coverity
Coverity CID: 1357320
Approved by: re (hrs)
MFC after: 10 days
Add `WRAPPED_CTASSERT` macro by annotating CTASSERTs with __unused
to deal with -Wunused-local-typedefs warnings from gcc 4.8+.
All other compilers (clang, etc) use CTASSERT as-is. A more generic
solution for this issue will be proposed after ^/stable/11 is forked.
Consolidate all CTASSERTs under one block instead of inlining them in
functions.
Approved by: re (gjb)
Differential Revision: https://reviews.freebsd.org/D7119
MFC after: 1 week
Reported by: Jenkins
Reviewed by: grehan (maintainer)
Sponsored by: EMC / Isilon Storage Division
Put cfl/prdt under AHCI_DEBUG #defines as they are only used in
those cases.
Approved by: re (gjb)
Differential Revision: https://reviews.freebsd.org/D7119
MFC after: 1 week
Reported by: Jenkins
Reviewed by: grehan (maintainer)
Sponsored by: EMC / Isilon Storage Division
Both __alloc_align and __alloc_size can't be used when the function
returns a pointer to memory. This fixes breakage when building with
clang 3.4:
In file included from /usr/src/svn/usr.sbin/bhyve/atkbdc.c:40:
/usr/include/stdlib.h:176:6: error: '__alloc_size__' attribute only
applies to functions that return a pointer [-Werror,-Wignored-attributes]
Pointed out by: ngie, cem
Approved by: re (gjb)
Fix the race between ioat_reset_hw and ioat_process_events.
HW reset isn't protected by a lock because it can sleep for a long time
(40.1 ms). This resulted in a race where we would process bogus parts
of the descriptor ring as if it had completed. This looked like
duplicate completions on old events, if your ring had looped at least
once.
Block callout and interrupt work while reset runs so the completion end
of things does not observe indeterminate state and process invalid parts
of the ring.
Start the channel with a manually implemented ioat_null() to keep other
submitters quiesced while we wait for the channel to start (100 us).
r295605 may have made the race between ioat_reset_hw and
ioat_process_events wider, but I believe it already existed before that
revision. ioat_process_events can be invoked by two asynchronous
sources: callout (softclock) and device interrupt. Those could race
each other, to the same effect.
Reviewed by: markj
Approved by: re
Sponsored by: EMC / Isilon Storage Division
Differential Revision: https://reviews.freebsd.org/D7097
not scheduled -> scheduled -> running -> not scheduled. The API and the
manual page assume that, some comments in the code assume that, and looks
like some contributors to the code also did. The problem is that this
paradigm isn't true. A callout can be scheduled and running at the same
time, which makes API description ambigouous. In such case callout_stop()
family of functions/macros should return 1 and 0 at the same time, since it
successfully unscheduled future callout but the current one is running.
Before this change we returned 1 in such a case, with an exception that
if running callout was migrating we returned 0, unless CS_MIGRBLOCK was
specified.
With this change, we now return 0 in case if future callout was unscheduled,
but another one is still in action, indicating to API users that resources
are not yet safe to be freed.
However, the sleepqueue code relies on getting 1 return code in that case,
and there already was CS_MIGRBLOCK flag, that covered one of the edge cases.
In the new return path we will also use this flag, to keep sleepqueue safe.
Since the flag CS_MIGRBLOCK doesn't block migration and now isn't limited to
migration edge case, rename it to CS_EXECUTING.
This change fixes panics on a high loaded TCP server.
Reviewed by: jch, hselasky, rrs, kib
Approved by: re (gjb)
Differential Revision: https://reviews.freebsd.org/D7042
option, not INVARIANTS. The function is required if we want
to load in a module that is compiled with INVARIANTS.
Reviewed by: jhb
Approved by: re (gjb)
vpanic() uses spinlock_enter() to disable interrupts before dumping core.
However, when the scheduler is stopped and INVARIANTS is not configured,
thread_lock() does not acquire a spinlock section, while thread_unlock()
releases one. This can result in interrupts staying enabled while the
kernel dumps core, complicating post-mortem analysis of the crash.
Approved by: re (gjb)
MFC after: 1 week
Sponsored by: EMC / Isilon Storage Division
method implementations: fstat(2), close(2), and poll(2). This change
synchronises auditing here with similar auditing for VFS-specific system
calls such as stat(2) that audit more complete vnode information.
Sponsored by: DARPA, AFRL
Approved by: re (kib)
MFC after: 1 week
objects.
Assert that there is no new waiters for the already terminated objects.
Old waiters should have been notified by the termination calling
vnode_pager_dealloc() (old/new are with regard of the lock acquisition
interval).
Only clear the vp->v_object for the case of already terminated object,
since other branches call vnode_pager_dealloc(), which should clear
the pointer. Assert this.
Tested by: pho
Reviewed by: alc
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Approved by: re (gjb)
or several hops we have segmentation fault because we overwrite the same
structure to store information for host and gateway.
Submitted by: Maryse Levavasseur <maryse.levavasseur@stormshield.eu>
Reworked by: hrs
Approved by: re (hrs)
Differential Revision: https://reviews.freebsd.org/D6980
Remove the use of fdt_data_to_res(), and instead construct the resources
manually. Additionally, avoid the 32-bit size limitation of fdt_data_get(), by
building physical addresses manually from the lbc ranges property.
Approved by: re@(gjb)
related to "shared" CPLs.
a) Combine t4_set_tcb_field and t4_set_tcb_field_rpl into a single
function. Allow callers to direct the response to any iq. Tidy up
set_ulp_mode_iscsi while there to use names from t4_tcb.h instead of
magic constants.
b) Remove all CPL handler tables from struct adapter. This reduces its
size by around 2KB. All handlers are now registered at MOD_LOAD instead
of attach or some kind of initialization/activation. The registration
functions do not need an adapter parameter any more.
c) Add per-iq handlers to deal with CPLs whose destination cannot be
determined solely from the opcode. There are 2 such CPLs in use right
now: SET_TCB_RPL and L2T_WRITE_RPL. The base driver continues to send
filter and L2T_WRITEs over the mgmtq and solicits the reply on fwq.
t4_tom (including the DDP code) now uses the port's ctrlq to send
L2T_WRITEs and SET_TCB_FIELDs and solicits the reply on an ofld_rxq.
fwq and ofld_rxq have different handlers that know what kind of tid to
expect in the reply. Update t4_write_l2e and callers to to support any
wrq/iq combination.
Approved by: re@ (kib@)
Sponsored by: Chelsio Communications
callout thread that can cause a kernel panic. Always do the final cleanup
in the callout thread by passing a separate callout function for that task
to callout_reset_sbt().
Protect the ref_count decrement in the callout with DN_BH_WLOCK(). All
other ref_count manipulation is protected with this lock.
There is still a tiny window between ref_count reaching zero and the end
of the callout function where it is unsafe to unload the module. Fixing
this would require the use of callout_drain(), but this can't be done
because dummynet holds a mutex and callout_drain() might sleep.
Remove the callout_pending(), callout_active(), and callout_deactivate()
calls from calculate_drop_prob(). They are not needed because this callout
uses callout_init_mtx().
Submitted by: Rasool Al-Saadi <ralsaadi@swin.edu.au>
Approved by: re (gjb)
MFC after: 3 days
Differential Revision: https://reviews.freebsd.org/D6928
