netif_set_real_num_rx_queues() can decrement and increment
the number of rx queues. For example ixgbe does this as
features and offloads are toggled. Presumably this could
also happen across down/up on most devices if the available
resources changed (cpu offlined).
The kobject needs to be zero'd in this case so that the
state is not preserved across kobject_put()/kobject_init_and_add().
This resolves the following error report.
ixgbe 0000:03:00.0: eth2: NIC Link is Up 10 Gbps, Flow Control: RX/TX
kobject (ffff880324b83210): tried to init an initialized object, something is seriously wrong.
Pid: 1972, comm: lldpad Not tainted 2.6.37-rc18021qaz+ #169
Call Trace:
[<ffffffff8121c940>] kobject_init+0x3a/0x83
[<ffffffff8121cf77>] kobject_init_and_add+0x23/0x57
[<ffffffff8107b800>] ? mark_lock+0x21/0x267
[<ffffffff813c6d11>] net_rx_queue_update_kobjects+0x63/0xc6
[<ffffffff813b5e0e>] netif_set_real_num_rx_queues+0x5f/0x78
[<ffffffffa0261d49>] ixgbe_set_num_queues+0x1c6/0x1ca [ixgbe]
[<ffffffffa0262509>] ixgbe_init_interrupt_scheme+0x1e/0x79c [ixgbe]
[<ffffffffa0274596>] ixgbe_dcbnl_set_state+0x167/0x189 [ixgbe]
Signed-off-by: John Fastabend <[email protected]>
Acked-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
jiffies is defined as "volatile".
extern unsigned long volatile __jiffy_data jiffies;
ACCESS_ONCE() uses "volatile".
As a result, some compilers warn duplicate `volatile' for ACCESS_ONCE(jiffies).
Signed-off-by: Tetsuo Handa <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
Add STMMAC to the list of supported Ethernet drivers
and myself as maintainer.
Signed-off-by: Giuseppe Cavallaro <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
RCU conversion in IGMP code done in net-next-2.6 raised a race in
__bond_resend_igmp_join_requests().
It iterates in_dev->mc_list without appropriate protection (RTNL, or
read_lock on in_dev->mc_list_lock).
Another cpu might delete an entry while we use it and trigger a fault.
Signed-off-by: Eric Dumazet <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
One of the readdir filldir_t callers was passing the raw ceph 64-bit ino
instead of the hashed 32-bit one, producing an EOVERFLOW in the filler
callback. Fix this by calling the ceph_vino_to_ino() helper to do the
conversion.
Reported-by: Jan Smets <[email protected]>
Tested-by: Jan Smets <[email protected]>
Signed-off-by: Sage Weil <[email protected]>
This follows wireless-testing 9236d838c9
("cfg80211: fix extension channel checks to initiate communication") and
fixes accidental case fall-through. Without this fix, HT40 is entirely
blocked.
Signed-off-by: Mark Mentovai <[email protected]>
Cc: [email protected]
Acked-by: Luis R. Rodriguez <[email protected]
Signed-off-by: John W. Linville <[email protected]>
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/linux-2.6-kgdb:
kgdb,ppc: Fix regression in evr register handling
kgdb,x86: fix regression in detach handling
kdb: fix crash when KDB_BASE_CMD_MAX is exceeded
kdb: fix memory leak in kdb_main.c
* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
powerpc: Update a BKL related comment
powerpc/mm: Fix module instruction tlb fault handling on Book-E 64
powerpc: Fix call to subpage_protection()
powerpc: Set CONFIG_32BIT on ppc32
powerpc/mm: Fix build error in setup_initial_memory_limit
powerpc/pseries: Don't override CONFIG_PPC_PSERIES_DEBUG
powerpc: Fix div64 in bootloader
Formerly sched_group_set_shares would force a rebalance by overflowing domain
share sums. Now that per-cpu averages are maintained we can set the true value
by issuing an update_cfs_shares() following a tg->shares update.
Also initialize tg se->load to 0 for consistency since we'll now set correct
weights on enqueue.
Signed-off-by: Paul Turner <[email protected]?>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Refactor the global load updates from update_shares_cpu() so that
update_cfs_load() can update global load when it is more than ~10%
out of sync.
The new global_load parameter allows us to force an update, regardless of
the error factor so that we can synchronize w/ update_shares().
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
When the system is busy, dilation of rq->next_balance makes lb->update_shares()
insufficiently frequent for threads which don't sleep (no dequeue/enqueue
updates). Adjust for this by making demand based updates based on the
accumulation of execution time sufficient to wrap our averaging window.
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Since shares updates are no longer expensive and effectively local, update them
at idle_balance(). This allows us to more quickly redistribute shares to
another cpu when our load becomes idle.
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Introduce a new sysctl for the shares window and disambiguate it from
sched_time_avg.
A 10ms window appears to be a good compromise between accuracy and performance.
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Avoid duplicate shares update calls by ensuring children always appear before
parents in rq->leaf_cfs_rq_list.
This allows us to do a single in-order traversal for update_shares().
Since we always enqueue in bottom-up order this reduces to 2 cases:
1) Our parent is already in the list, e.g.
root
\
b
/\
c d* (root->b->c already enqueued)
Since d's parent is enqueued we push it to the head of the list, implicitly ahead of b.
2) Our parent does not appear in the list (or we have no parent)
In this case we enqueue to the tail of the list, if our parent is subsequently enqueued
(bottom-up) it will appear to our right by the same rule.
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Using cfs_rq->nr_running is not sufficient to synchronize update_cfs_load with
the put path since nr_running accounting occurs at deactivation.
It's also not safe to make the removal decision based on load_avg as this fails
with both high periods and low shares. Resolve this by clipping history after
4 periods without activity.
Note: the above will always occur from update_shares() since in the
last-task-sleep-case that task will still be cfs_rq->curr when update_cfs_load
is called.
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
As part of enqueue_entity both a new entity weight and its contribution to the
queuing cfs_rq / rq are updated. Since update_cfs_shares will only update the
queueing weights when the entity is on_rq (which in this case it is not yet),
there's a dependency loop here:
update_cfs_shares needs account_entity_enqueue to update cfs_rq->load.weight
account_entity_enqueue needs the updated weight for the queuing cfs_rq load[*]
Fix this and avoid spurious dequeue/enqueues by issuing update_cfs_shares as
if we had accounted the enqueue already.
This was also resulting in rq->load corruption previously.
[*]: this dependency also exists when using the group cfs_rq w/
update_cfs_shares as the weight of the enqueued entity changes
without the load being updated.
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Make tg_shares_up() use the active cgroup list, this means we cannot
do a strict bottom-up walk of the hierarchy, but assuming its a very
wide tree with a small number of active groups it should be a win.
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
Make certain load-balance actions scale per number of active cgroups
instead of the number of existing cgroups.
This makes wakeup/sleep paths more expensive, but is a win for systems
where the vast majority of existing cgroups are idle.
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
By tracking a per-cpu load-avg for each cfs_rq and folding it into a
global task_group load on each tick we can rework tg_shares_up to be
strictly per-cpu.
This should improve cpu-cgroup performance for smp systems
significantly.
[ Paul: changed to use queueing cfs_rq + bug fixes ]
Signed-off-by: Paul Turner <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>
While discussing the need for sched_idle_next(), Oleg remarked that
since try_to_wake_up() ensures sleeping tasks will end up running on a
sane cpu, we can do away with migrate_live_tasks().
If we then extend the existing hack of migrating current from
CPU_DYING to migrating the full rq worth of tasks from CPU_DYING, the
need for the sched_idle_next() abomination disappears as well, since
idle will be the only possible thread left after the migration thread
stops.
This greatly simplifies the hot-unplug task migration path, as can be
seen from the resulting code reduction (and about half the new lines
are comments).
Suggested-by: Oleg Nesterov <[email protected]>
Signed-off-by: Peter Zijlstra <[email protected]>
LKML-Reference: <1289851597.2109.547.camel@laptop>
Signed-off-by: Ingo Molnar <[email protected]>
A process suspended waiting for a higher sequence or no sequence to unreserve,
a bo may be beaten to the reservation by a process with a lower sequence.
In that case the first process should give up trying to reserve and
return -EAGAIN. In order for that to happen, we must wake waiting processes
when we change sequence, so that they have a chance to detect the new
sequence.
Signed-off-by: Thomas Hellstrom <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
* 'nouveau/for-airlied' of /ssd/git/drm-nouveau-next: (25 commits)
nouveau: Acknowledge HPD irq in handler, not bottom half
drm/nouveau: Fix a few confusions between "chipset" and "card_type".
drm/nouveau: don't expose backlight control when available through ACPI
drm/nouveau/pm: improve memtiming mappings
drm/nouveau: Make PCIE GART size depend on the available RAMIN space.
drm/nouveau: Return error from nouveau_gpuobj_new if we're out of RAMIN.
drm/nouveau: Fix compilation issues in nouveau_pm when CONFIG_HWMON is not set
drm/nouveau: Don't use load detection for connector polling.
drm/nv10-nv20: Fix instability after MPLL changes.
drm/nv50: implement possible workaround for NV86 PGRAPH TLB flush hang
drm/nouveau: Don't poll LVDS outputs.
drm/nouveau: Use "force" to decide if analog load detection is ok or not.
drm/nv04: Fix scanout over the 16MB mark.
drm/nouveau: fix nv40 pcie gart size
drm/nva3: fix overflow in fixed point math used for pll calculation
drm/nv10: Balance RTs expected to be accessed simultaneously by the 3d engine.
drm/nouveau: Expose some BO usage flags to userspace.
drm/nouveau: Reduce severity of the unknown getparam error.
drm/nouveau: Avoid lock dependency between ramht and ramin spinlocks.
drm/nouveau: Some random cleanups.
...
We aren't currently using tiling in userspace on evergreen,
but the info we currently return for the tiling info query
(gb_addr_config) is no adequate for userspace tiling alignment
calculations. It does not contain the bank info. Create a custom
tiling info dword with all the necessary info (num channels,
num banks, group size, row size).
Signed-off-by: Alex Deucher <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
We were previously dropping alignment requests on the floor
when allocating buffers so we always ended up page aligned.
Certain tiling modes on 6xx+ require larger alignment which
wasn't happening before.
Signed-off-by: Alex Deucher <[email protected]>
Cc: Jerome Glisse <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
These are external encoder chips connected via DVO or DP.
The actual external encoder programming is handled by the
kms encoder functions for primary encoder.
Signed-off-by: Alex Deucher <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Handle all the various asic family specific things for DVO.
Signed-off-by: Alex Deucher <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
The eDP panel must be powered up for aux transactions, so power it
up for detect and mode probe functions, otherwise power it up or
down based on dpms.
v2:
- only mess with eDP panel on DCE4+
- only mess with eDP panel on eDP connectors, not all DP connectors
v3:
- be extra careful to only mess with eDP panels on eDP connectors
v4:
- avoid possible null derefernce if a connector has not been
assigned to the encoder
Signed-off-by: Alex Deucher <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
If there was no connector mapped to the encoder, atombios_get_encoder_mode()
returned 0 which is the id for DP. Return something sane instead based on
the encoder id. This avoids hitting the DP paths on non-DP encoders.
Signed-off-by: Alex Deucher <[email protected]>
Cc: [email protected]
Signed-off-by: Dave Airlie <[email protected]>
Looks like a typo in:
drm/radeon/r600: fix tiling issues in CS checker.
(f30df2fad0)
Signed-off-by: Alex Deucher <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
enable_vblank implementations should use negative result to indicate error.
radeon_enable_vblank() returns EINVAL in this case. Change this to -EINVAL.
Signed-off-by: Vasiliy Kulikov <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Tiled buffers have the same alignment requirements regardless of
whether the surface is for db, cb, or textures. Previously, the
calculations where inconsistent for each buffer type.
- Unify the alignment calculations in a common function
- Standardize the alignment units (pixels for pitch/height/depth,
bytes for base)
- properly check the buffer base alignments
Signed-off-by: Alex Deucher <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
The old code generated an interrupt storm bad enough to completely
take down my system.
Signed-off-by: Andy Lutomirski <[email protected]>
Signed-off-by: Ben Skeggs <[email protected]>
Avoid confusing userspace by not publishing backlight controls if ACPI
equivalents are available.
Reported-by: Aaron Sowry <[email protected]>
Signed-off-by: Ben Skeggs <[email protected]>
Improvements:
- Fix bug in switch statement
- Add parts of 0x10022c, 0x10023c
- Clean up 0x100234
- Comment out assumption in 0x100228 until verified
Signed-off-by: Roy Spliet <[email protected]>
Signed-off-by: Ben Skeggs <[email protected]>
Analog output polling makes GL programs jerky when pageflip is being
used because it's carried out with the mode_config mutex held.
Signed-off-by: Francisco Jerez <[email protected]>
Signed-off-by: Ben Skeggs <[email protected]>
