search for: atomic_xchg

Now with the below kernel patch, the excl operations like dev add/replace/resize and balance returns the btrfs error code defined in btrfs.h, this patch will help btrfs-progs (and thus user) to know the error string on the terminal (instead of /var/log/messages as previously kernel did). This patch depends on the btrfs kernel patch: btrfs: return btrfs error code for dev excl ops err

..._prog(struct nvkm_clk *clk, int pstateid) } static void -nvkm_clk_update_work(struct work_struct *work) +nvkm_clk_update_impl(struct nvkm_clk *clk) { - struct nvkm_clk *clk = container_of(work, typeof(*clk), work); struct nvkm_subdev *subdev = &clk->subdev; int pstate, ret; - if (!atomic_xchg(&clk->waiting, 0)) - return; clk->pwrsrc = power_supply_is_system_supplied(); if (clk->pstate) @@ -350,6 +347,17 @@ nvkm_clk_update_work(struct work_struct *work) nvkm_error(subdev, "error setting pstate %d: %d\n", pstate, ret); } +} + +static void +nvkm_cl...

...iterations and then updates its value. In the case where other threads manipulate the same &reg->hr_steady_iterations concurrently, race condition might be possible. I think it would be better to guarantee consecutive executions of read and write by special purposed atomic operations (e.g. atomic_xchg) Please examine the issue and let me know your opinion. Thank you. Sincerely Shin Hong

...l.c | 8 -------- 1 file changed, 8 deletions(-) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index b908960..40631cf 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -2245,13 +2245,6 @@ static int btrfs_ioctl_defrag(struct file *file, void __user *argp) if (ret) return ret; - if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running, - 1)) { - pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n"); - mnt_drop_write_file(file); - return -EINVAL; - } - if (btrfs_root_readonly(root)) { ret = -EROFS; goto out; @@ -2306,7...

...y usage of this arg, perhaps it would be better to simply remove it and shrink the caller's code a bit? It is also used in 3/8, but we can read the "fresh" value of ->qlcode (trylock does this anyway), and perhaps it can actually help if it is already unlocked. > + prev_qcode = atomic_xchg(&lock->qlcode, my_qcode); > + /* > + * It is possible that we may accidentally steal the lock. If this is > + * the case, we need to either release it if not the head of the queue > + * or get the lock and be done with it. > + */ > + if (unlikely(!(prev_qcode & _QSP...

...y usage of this arg, perhaps it would be better to simply remove it and shrink the caller's code a bit? It is also used in 3/8, but we can read the "fresh" value of ->qlcode (trylock does this anyway), and perhaps it can actually help if it is already unlocked. > + prev_qcode = atomic_xchg(&lock->qlcode, my_qcode); > + /* > + * It is possible that we may accidentally steal the lock. If this is > + * the case, we need to either release it if not the head of the queue > + * or get the lock and be done with it. > + */ > + if (unlikely(!(prev_qcode & _QSP...

...y(pstate, &clk->states, head) { if (idx++ == pstatei) break; @@ -292,7 +295,7 @@ nvkm_pstate_work(struct work_struct *work) { struct nvkm_clk *clk = container_of(work, typeof(*clk), work); struct nvkm_subdev *subdev = &clk->subdev; - int pstate; + int pstate, ret; if (!atomic_xchg(&clk->waiting, 0)) return; @@ -312,12 +315,10 @@ nvkm_pstate_work(struct work_struct *work) } nvkm_trace(subdev, "-> %d\n", pstate); - if (pstate != clk->pstate) { - int ret = nvkm_pstate_prog(clk, pstate); - if (ret) { - nvkm_error(subdev, "error setting ps...

...++ b/drm/nouveau/nvkm/subdev/clk/base.c @@ -320,6 +320,7 @@ nvkm_clk_update_work(struct work_struct *work) { struct nvkm_clk *clk = container_of(work, typeof(*clk), work); struct nvkm_subdev *subdev = &clk->subdev; + struct device *dev = subdev->device->dev; int pstate; if (!atomic_xchg(&clk->waiting, 0)) @@ -345,7 +346,14 @@ nvkm_clk_update_work(struct work_struct *work) pstate = NVKM_CLK_PSTATE_DEFAULT; } - clk->func->update(clk, pstate); + // only call into the code if the GPU is powered on + if (!pm_runtime_suspended(dev)) { + // it would be a shame if the...

...&clk->subdev; int pstate; @@ -349,7 +349,7 @@ nvkm_clk_update_impl(struct nvkm_clk *clk) pstate = -1; } - clk->func->update(clk, pstate); + clk->func->update(clk, pstate, force); } static void @@ -360,7 +360,7 @@ nvkm_clk_update_work(struct work_struct *work) if (!atomic_xchg(&clk->waiting, 0)) return; - nvkm_clk_update_impl(clk); + nvkm_clk_update_impl(clk, false); wake_up_all(&clk->wait); nvkm_notify_get(&clk->pwrsrc_ntfy); @@ -613,11 +613,7 @@ nvkm_clk_init(struct nvkm_subdev *subdev) if (clk->func->init) return clk->func...

...nvkm_clk *clk, int pstatei) } static void -nvkm_pstate_work(struct work_struct *work) +nvkm_clk_update_work(struct work_struct *work) { struct nvkm_clk *clk = container_of(work, typeof(*clk), work); struct nvkm_subdev *subdev = &clk->subdev; - int pstate; + int pstate, ret; if (!atomic_xchg(&clk->waiting, 0)) return; @@ -327,21 +331,25 @@ nvkm_pstate_work(struct work_struct *work) } nvkm_trace(subdev, "-> %d\n", pstate); - if (pstate != clk->pstate) { - int ret = nvkm_pstate_prog(clk, pstate); - if (ret) { - nvkm_error(subdev, "error setting ps...

This series addresses various issues inside the reclocking code: 1. after resume the set clocks are reset 2. reclocking not possible while GPU is suspended 3. nouveau always does full reclocks even if only a change of the voltage is required Some of the patches were part of the bigger reclocking series I sent months ago, some things have changed though. This is also preparation work of

We are slowly getting there! v4 of the series with some realy good improvements, so I am sure this is like 95% done and only needs some proper polishing and proper Reviews! I also added the NvVoltOffsetmV module parameter, so that a user is able to over and !under!-volt the GPU. Overvolting makes sense, when there are still some reclocking issues left, which might be solved by a higher voltage.

...UT] + * @ncode: New queue code to be exchanged + * Return: An enum exitval value + */ +static inline enum exitval +queue_code_xchg(struct qspinlock *lock, u32 *ocode, u32 ncode) +{ + ncode |= _QLOCK_LOCKED; /* Set lock bit */ + + /* + * Exchange current copy of the queue node code + */ + *ocode = atomic_xchg(&lock->qlcode, ncode); + + if (likely(*ocode & _QLOCK_LOCKED)) { + *ocode &= ~_QLOCK_LOCKED; /* Clear the lock bit */ + return NORMAL_EXIT; + } + /* + * It is possible that we may accidentally steal the lock during + * the unlock-lock transition. If this is the case, we need to e...

...L; + + /* + * The lock may be available at this point, try again if no task was + * waiting in the queue. + */ + if (!(qsval >> _QCODE_OFFSET) && queue_spin_trylock(lock)) { + put_qnode(); + return; + } + + /* + * Exchange current copy of the queue node code + */ + prev_qcode = atomic_xchg(&lock->qlcode, my_qcode); + /* + * It is possible that we may accidentally steal the lock. If this is + * the case, we need to either release it if not the head of the queue + * or get the lock and be done with it. + */ + if (unlikely(!(prev_qcode & _QSPINLOCK_LOCKED))) { + if (prev...

...L; + + /* + * The lock may be available at this point, try again if no task was + * waiting in the queue. + */ + if (!(qsval >> _QCODE_OFFSET) && queue_spin_trylock(lock)) { + put_qnode(); + return; + } + + /* + * Exchange current copy of the queue node code + */ + prev_qcode = atomic_xchg(&lock->qlcode, my_qcode); + /* + * It is possible that we may accidentally steal the lock. If this is + * the case, we need to either release it if not the head of the queue + * or get the lock and be done with it. + */ + if (unlikely(!(prev_qcode & _QSPINLOCK_LOCKED))) { + if (prev...

This series addresses various issues inside the reclocking code: 1. after resume the set clocks are reset 2. reclocking not possible while GPU is suspended Some of the patches were part of the bigger reclocking series I sent months ago, some things have changed though. This is also preparation work of changing the clock state due to temperature changes and dynamic reclocking. v2: remove commits

The goal is to detect when drives start to get an increased error rate, when drives should be replaced soon. Therefore statistic counters are added that count IO errors (read, write and flush). Additionally, the software detected errors like checksum errors and corrupted blocks are counted. An ioctl interface is added to get the device statistic counters. A second ioctl is added to atomically get

v4->v5: - Move the optimized 2-task contending code to the generic file to enable more architectures to use it without code duplication. - Address some of the style-related comments by PeterZ. - Allow the use of unfair queue spinlock in a real para-virtualized execution environment. - Add para-virtualization support to the qspinlock code by ensuring that the lock holder and queue

v4->v5: - Move the optimized 2-task contending code to the generic file to enable more architectures to use it without code duplication. - Address some of the style-related comments by PeterZ. - Allow the use of unfair queue spinlock in a real para-virtualized execution environment. - Add para-virtualization support to the qspinlock code by ensuring that the lock holder and queue

v2 is updated to account for feedback from Will, Peter, and Waiman (thank you), and trims off a couple of RFC and unrelated patches. Thanks, Nick Nicholas Piggin (6): powerpc/powernv: must include hvcall.h to get PAPR defines powerpc/pseries: move some PAPR paravirt functions to their own file powerpc: move spinlock implementation to simple_spinlock powerpc/64s: implement queued