| Message ID | 20160817101003.30055-1-bigeasy@linutronix.de (mailing list archive) |
|---|---|
| State | Deferred, archived |
| Headers | show |
On Wed, Aug 17, 2016 at 12:09:59PM +0200, Sebastian Andrzej Siewior wrote: > The driver creates its own per-CPU threads which are updated based on CPU > hotplug events. It is also possible to use kworkers and remove some of the > infrastructure get the same job done while saving a few lines of code. > > The DECLARE_PER_CPU() definition is moved into the header file where it > belongs. bnx2i_percpu_io_thread() becomes bnx2i_percpu_io_work() which is > mostly the same code. The outer loop (kthread_should_stop()) gets removed and > the remaining code is shifted to the left. > bnx2i_queue_scsi_cmd_resp() is mostly the same. The code checked ->iothread to > decide if there is an active per-CPU thread. With the kworkers this is no > longer possible nor required. > The allocation of struct bnx2i_work does not happen with ->p_work_lock held > which is not required. I am unsure about the call-stack so I can't say > if this qualifies it for the allocation with GFP_KERNEL instead of > GFP_ATOMIC (it is not _bh lock but as I said, I don't know the context). > The allocation case has been reversed so the inner if case is called on > !bnx2i_work and is just the invocation one function since the lock is not > held during allocation. The init of the new bnx2i_work struct is now > done also without the ->p_work_lock held: it is a new object, nobody > knows about it yet. It should be enough to hold the lock while adding > this item to the list. I am unsure about that atomic_inc() so I keep > things as they were. > > The remaining part is the removal CPU hotplug notifier since it is taken > care by the kworker code. > > This patch was only compile-tested due to -ENODEV. > > Cc: QLogic-Storage-Upstream@qlogic.com > Cc: Christoph Hellwig <hch@lst.de> > Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> I'm sorry it took so long but for the whole series: Tested-by: Johannes Thumshirn <jth@kernel.org> Acked-by: Johannes Thumshirn <jth@kernel.org> Testing was conducted by booting from a QLogic BCM57840 CNA. Johannes
>>>>> "Johannes" == Johannes Thumshirn <jthumshirn@suse.de> writes: >> On Wed, Aug 17, 2016 at 12:09:59PM +0200, Sebastian Andrzej Siewior wrote: >> The driver creates its own per-CPU threads which are updated based on >> CPU hotplug events. It is also possible to use kworkers and remove >> some of the infrastructure get the same job done while saving a few >> lines of code. Johannes> I'm sorry it took so long but for the whole series: Tested-by: Johannes> Johannes Thumshirn <jth@kernel.org> Acked-by: Johannes Johannes> Thumshirn <jth@kernel.org> Johannes> Testing was conducted by booting from a QLogic BCM57840 CNA. Chad?
On Fri, 9 Sep 2016, 11:28am -0000, Martin K. Petersen wrote: > >>>>> "Johannes" == Johannes Thumshirn <jthumshirn@suse.de> writes: > > >> On Wed, Aug 17, 2016 at 12:09:59PM +0200, Sebastian Andrzej Siewior > wrote: > >> The driver creates its own per-CPU threads which are updated based on > >> CPU hotplug events. It is also possible to use kworkers and remove > >> some of the infrastructure get the same job done while saving a few > >> lines of code. > > Johannes> I'm sorry it took so long but for the whole series: Tested-by: > Johannes> Johannes Thumshirn <jth@kernel.org> Acked-by: Johannes > Johannes> Thumshirn <jth@kernel.org> > > Johannes> Testing was conducted by booting from a QLogic BCM57840 CNA. > > Chad? > We're regression testing the patches now. Please hold off on applying them. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
>>>>> "Chad" == Chad Dupuis <chad.dupuis@cavium.com> writes:
Johannes> Testing was conducted by booting from a QLogic BCM57840 CNA.
Chad> We're regression testing the patches now. Please hold off on
Chad> applying them.
OK. I will wait.
On 2016-09-14 13:25:12 [-0400], Martin K. Petersen wrote: > >>>>> "Chad" == Chad Dupuis <chad.dupuis@cavium.com> writes: > > Chad> We're regression testing the patches now. Please hold off on > Chad> applying them. > > OK. I will wait. Chad, any updates on this tests? Sebastian -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
It seems like the whole damn driver should just use threaded interrupts. Of course it's a giant beast and not just the iSCSI one. But even if we don't go all the way I'd much prefer workqueues. kthread work is simply the worst API ever and I'd prefer to not have it proliferate. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, 7 Nov 2016, Christoph Hellwig wrote: > It seems like the whole damn driver should just use threaded interrupts. > Of course it's a giant beast and not just the iSCSI one. But even > if we don't go all the way I'd much prefer workqueues. kthread work > is simply the worst API ever and I'd prefer to not have it proliferate. That's what the patch is doing. It uses INIT_WORK() and schedule_work[_on](). I can't find any reference to kthread work. Thanks, tglx -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Nov 07, 2016 at 05:31:03PM +0100, Thomas Gleixner wrote: > On Mon, 7 Nov 2016, Christoph Hellwig wrote: > > > It seems like the whole damn driver should just use threaded interrupts. > > Of course it's a giant beast and not just the iSCSI one. But even > > if we don't go all the way I'd much prefer workqueues. kthread work > > is simply the worst API ever and I'd prefer to not have it proliferate. > > That's what the patch is doing. It uses INIT_WORK() and > schedule_work[_on](). I can't find any reference to kthread work. You're right it does - between the incorrect subject and the fact that it still keeps the linked list of items arounds instead of fully using the workqueue infrastructure I was a bit confused before my first coffee this morning. Same applies to patch 2. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 2016-11-07 17:38:48 [+0100], Christoph Hellwig wrote: > You're right it does - between the incorrect subject and the fact > that it still keeps the linked list of items arounds instead of fully > using the workqueue infrastructure I was a bit confused before my > first coffee this morning. Same applies to patch 2. sorry for the confusion in the subject. If I remember correctly you said that we may not have enough room for a larger / work_struct struct and I should keep the list for now. Sebastian -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Nov 07, 2016 at 05:46:29PM +0100, Sebastian Andrzej Siewior wrote: > sorry for the confusion in the subject. If I remember correctly you said > that we may not have enough room for a larger / work_struct struct and I > should keep the list for now. But that was for libfc where it only has a tiny bit of private data i nthe skb - this is for cnix which has plenty of space, and in fact probably should do the workqueue offload in the core before invoking the protocol drivers. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On 2016-11-07 17:48:46 [+0100], Christoph Hellwig wrote: > On Mon, Nov 07, 2016 at 05:46:29PM +0100, Sebastian Andrzej Siewior wrote: > > sorry for the confusion in the subject. If I remember correctly you said > > that we may not have enough room for a larger / work_struct struct and I > > should keep the list for now. > > But that was for libfc where it only has a tiny bit of private data > i nthe skb - this is for cnix which has plenty of space, and in fact Right. now I remember. > probably should do the workqueue offload in the core before invoking > the protocol drivers. So we keep things as they are right now or are we getting also rid of the internal list? This was tested by Johannes and Chad (claimed to do testing) and the removal the internal list afterwards shouldn't be that big deal. Not sure what the last point is about. Sebastian -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Nov 07, 2016 at 06:04:21PM +0100, Sebastian Andrzej Siewior wrote: > So we keep things as they are right now or are we getting also rid of > the internal list? This was tested by Johannes and Chad (claimed to do > testing) IFF the patches actually are tested as-is let's get them in. I don't think they are how we want the code to like in the long run, though. > > Not sure what the last point is about. It's about bnx2i and bnx2fc being clients of the main cnix driver, which does the actual interrupt handling. It seems we should actually offloading things to a workqueue or threaded interrupts in that core module instead of duplicating it in the clients. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Nov 07, 2016 at 08:10:25PM +0100, Christoph Hellwig wrote: > On Mon, Nov 07, 2016 at 06:04:21PM +0100, Sebastian Andrzej Siewior wrote: > > So we keep things as they are right now or are we getting also rid of > > the internal list? This was tested by Johannes and Chad (claimed to do > > testing) > > IFF the patches actually are tested as-is let's get them in. I don't > think they are how we want the code to like in the long run, though. I'm definitively sure I did test the bnx2fc patches. I'm not 100% sure I did test the bnx2i patches as well (althoug I wrote "for the whole series"). Testing was IIRC booting from the FCoE LUN and the doing a series of fio tests for about an hour. Byte, Johannes
diff --git a/drivers/scsi/bnx2i/bnx2i.h b/drivers/scsi/bnx2i/bnx2i.h index ed7f3228e234..78cdc493bab5 100644 --- a/drivers/scsi/bnx2i/bnx2i.h +++ b/drivers/scsi/bnx2i/bnx2i.h @@ -31,7 +31,6 @@ #include <linux/netdevice.h> #include <linux/completion.h> #include <linux/kthread.h> -#include <linux/cpu.h> #include <scsi/scsi_cmnd.h> #include <scsi/scsi_device.h> @@ -775,12 +774,11 @@ struct bnx2i_work { }; struct bnx2i_percpu_s { - struct task_struct *iothread; + struct work_struct work; struct list_head work_list; spinlock_t p_work_lock; }; - /* Global variables */ extern unsigned int error_mask1, error_mask2; extern u64 iscsi_error_mask; @@ -797,7 +795,7 @@ extern unsigned int rq_size; extern struct device_attribute *bnx2i_dev_attributes[]; - +DECLARE_PER_CPU(struct bnx2i_percpu_s, bnx2i_percpu); /* * Function Prototypes @@ -875,8 +873,5 @@ extern void bnx2i_print_active_cmd_queue(struct bnx2i_conn *conn); extern void bnx2i_print_xmit_pdu_queue(struct bnx2i_conn *conn); extern void bnx2i_print_recv_state(struct bnx2i_conn *conn); -extern int bnx2i_percpu_io_thread(void *arg); -extern int bnx2i_process_scsi_cmd_resp(struct iscsi_session *session, - struct bnx2i_conn *bnx2i_conn, - struct cqe *cqe); +extern void bnx2i_percpu_io_work(struct work_struct *work); #endif diff --git a/drivers/scsi/bnx2i/bnx2i_hwi.c b/drivers/scsi/bnx2i/bnx2i_hwi.c index 42921dbba927..9be58f6523b3 100644 --- a/drivers/scsi/bnx2i/bnx2i_hwi.c +++ b/drivers/scsi/bnx2i/bnx2i_hwi.c @@ -19,8 +19,6 @@ #include <scsi/libiscsi.h> #include "bnx2i.h" -DECLARE_PER_CPU(struct bnx2i_percpu_s, bnx2i_percpu); - /** * bnx2i_get_cid_num - get cid from ep * @ep: endpoint pointer @@ -1350,9 +1348,9 @@ int bnx2i_send_fw_iscsi_init_msg(struct bnx2i_hba *hba) * * process SCSI CMD Response CQE & complete the request to SCSI-ML */ -int bnx2i_process_scsi_cmd_resp(struct iscsi_session *session, - struct bnx2i_conn *bnx2i_conn, - struct cqe *cqe) +static int bnx2i_process_scsi_cmd_resp(struct iscsi_session *session, + struct bnx2i_conn *bnx2i_conn, + struct cqe *cqe) { struct iscsi_conn *conn = bnx2i_conn->cls_conn->dd_data; struct bnx2i_hba *hba = bnx2i_conn->hba; @@ -1862,45 +1860,37 @@ static void bnx2i_process_cmd_cleanup_resp(struct iscsi_session *session, /** - * bnx2i_percpu_io_thread - thread per cpu for ios + * bnx2i_percpu_io_work - thread per cpu for ios * - * @arg: ptr to bnx2i_percpu_info structure + * @work_s: The work struct */ -int bnx2i_percpu_io_thread(void *arg) +void bnx2i_percpu_io_work(struct work_struct *work_s) { - struct bnx2i_percpu_s *p = arg; + struct bnx2i_percpu_s *p; struct bnx2i_work *work, *tmp; LIST_HEAD(work_list); - set_user_nice(current, MIN_NICE); + p = container_of(work_s, struct bnx2i_percpu_s, work); - while (!kthread_should_stop()) { - spin_lock_bh(&p->p_work_lock); - while (!list_empty(&p->work_list)) { - list_splice_init(&p->work_list, &work_list); - spin_unlock_bh(&p->p_work_lock); - - list_for_each_entry_safe(work, tmp, &work_list, list) { - list_del_init(&work->list); - /* work allocated in the bh, freed here */ - bnx2i_process_scsi_cmd_resp(work->session, - work->bnx2i_conn, - &work->cqe); - atomic_dec(&work->bnx2i_conn->work_cnt); - kfree(work); - } - spin_lock_bh(&p->p_work_lock); - } - set_current_state(TASK_INTERRUPTIBLE); + spin_lock_bh(&p->p_work_lock); + while (!list_empty(&p->work_list)) { + list_splice_init(&p->work_list, &work_list); spin_unlock_bh(&p->p_work_lock); - schedule(); + + list_for_each_entry_safe(work, tmp, &work_list, list) { + list_del_init(&work->list); + /* work allocated in the bh, freed here */ + bnx2i_process_scsi_cmd_resp(work->session, + work->bnx2i_conn, + &work->cqe); + atomic_dec(&work->bnx2i_conn->work_cnt); + kfree(work); + } + spin_lock_bh(&p->p_work_lock); } - __set_current_state(TASK_RUNNING); - - return 0; + spin_unlock_bh(&p->p_work_lock); } - /** * bnx2i_queue_scsi_cmd_resp - queue cmd completion to the percpu thread * @bnx2i_conn: bnx2i connection @@ -1920,7 +1910,6 @@ static int bnx2i_queue_scsi_cmd_resp(struct iscsi_session *session, struct bnx2i_percpu_s *p = NULL; struct iscsi_task *task; struct scsi_cmnd *sc; - int rc = 0; int cpu; spin_lock(&session->back_lock); @@ -1939,33 +1928,29 @@ static int bnx2i_queue_scsi_cmd_resp(struct iscsi_session *session, spin_unlock(&session->back_lock); - p = &per_cpu(bnx2i_percpu, cpu); - spin_lock(&p->p_work_lock); - if (unlikely(!p->iothread)) { - rc = -EINVAL; - goto err; - } /* Alloc and copy to the cqe */ bnx2i_work = kzalloc(sizeof(struct bnx2i_work), GFP_ATOMIC); - if (bnx2i_work) { - INIT_LIST_HEAD(&bnx2i_work->list); - bnx2i_work->session = session; - bnx2i_work->bnx2i_conn = bnx2i_conn; - memcpy(&bnx2i_work->cqe, cqe, sizeof(struct cqe)); - list_add_tail(&bnx2i_work->list, &p->work_list); - atomic_inc(&bnx2i_conn->work_cnt); - wake_up_process(p->iothread); - spin_unlock(&p->p_work_lock); - goto done; - } else - rc = -ENOMEM; -err: - spin_unlock(&p->p_work_lock); - bnx2i_process_scsi_cmd_resp(session, bnx2i_conn, (struct cqe *)cqe); -done: - return rc; -} + if (!bnx2i_work) { + bnx2i_process_scsi_cmd_resp(session, bnx2i_conn, + (struct cqe *)cqe); + return -ENOMEM; + } + p = per_cpu_ptr(&bnx2i_percpu, cpu); + + INIT_LIST_HEAD(&bnx2i_work->list); + bnx2i_work->session = session; + bnx2i_work->bnx2i_conn = bnx2i_conn; + memcpy(&bnx2i_work->cqe, cqe, sizeof(struct cqe)); + + spin_lock(&p->p_work_lock); + list_add_tail(&bnx2i_work->list, &p->work_list); + atomic_inc(&bnx2i_conn->work_cnt); + spin_unlock(&p->p_work_lock); + + schedule_work_on(cpu, &p->work); + return 0; +} /** * bnx2i_process_new_cqes - process newly DMA'ed CQE's diff --git a/drivers/scsi/bnx2i/bnx2i_init.c b/drivers/scsi/bnx2i/bnx2i_init.c index c8b410c24cf0..976a6bc86d39 100644 --- a/drivers/scsi/bnx2i/bnx2i_init.c +++ b/drivers/scsi/bnx2i/bnx2i_init.c @@ -70,14 +70,6 @@ u64 iscsi_error_mask = 0x00; DEFINE_PER_CPU(struct bnx2i_percpu_s, bnx2i_percpu); -static int bnx2i_cpu_callback(struct notifier_block *nfb, - unsigned long action, void *hcpu); -/* notification function for CPU hotplug events */ -static struct notifier_block bnx2i_cpu_notifier = { - .notifier_call = bnx2i_cpu_callback, -}; - - /** * bnx2i_identify_device - identifies NetXtreme II device type * @hba: Adapter structure pointer @@ -410,93 +402,6 @@ int bnx2i_get_stats(void *handle) return 0; } - -/** - * bnx2i_percpu_thread_create - Create a receive thread for an - * online CPU - * - * @cpu: cpu index for the online cpu - */ -static void bnx2i_percpu_thread_create(unsigned int cpu) -{ - struct bnx2i_percpu_s *p; - struct task_struct *thread; - - p = &per_cpu(bnx2i_percpu, cpu); - - thread = kthread_create_on_node(bnx2i_percpu_io_thread, (void *)p, - cpu_to_node(cpu), - "bnx2i_thread/%d", cpu); - /* bind thread to the cpu */ - if (likely(!IS_ERR(thread))) { - kthread_bind(thread, cpu); - p->iothread = thread; - wake_up_process(thread); - } -} - - -static void bnx2i_percpu_thread_destroy(unsigned int cpu) -{ - struct bnx2i_percpu_s *p; - struct task_struct *thread; - struct bnx2i_work *work, *tmp; - - /* Prevent any new work from being queued for this CPU */ - p = &per_cpu(bnx2i_percpu, cpu); - spin_lock_bh(&p->p_work_lock); - thread = p->iothread; - p->iothread = NULL; - - /* Free all work in the list */ - list_for_each_entry_safe(work, tmp, &p->work_list, list) { - list_del_init(&work->list); - bnx2i_process_scsi_cmd_resp(work->session, - work->bnx2i_conn, &work->cqe); - kfree(work); - } - - spin_unlock_bh(&p->p_work_lock); - if (thread) - kthread_stop(thread); -} - - -/** - * bnx2i_cpu_callback - Handler for CPU hotplug events - * - * @nfb: The callback data block - * @action: The event triggering the callback - * @hcpu: The index of the CPU that the event is for - * - * This creates or destroys per-CPU data for iSCSI - * - * Returns NOTIFY_OK always. - */ -static int bnx2i_cpu_callback(struct notifier_block *nfb, - unsigned long action, void *hcpu) -{ - unsigned cpu = (unsigned long)hcpu; - - switch (action) { - case CPU_ONLINE: - case CPU_ONLINE_FROZEN: - printk(KERN_INFO "bnx2i: CPU %x online: Create Rx thread\n", - cpu); - bnx2i_percpu_thread_create(cpu); - break; - case CPU_DEAD: - case CPU_DEAD_FROZEN: - printk(KERN_INFO "CPU %x offline: Remove Rx thread\n", cpu); - bnx2i_percpu_thread_destroy(cpu); - break; - default: - break; - } - return NOTIFY_OK; -} - - /** * bnx2i_mod_init - module init entry point * @@ -533,22 +438,12 @@ static int __init bnx2i_mod_init(void) /* Create percpu kernel threads to handle iSCSI I/O completions */ for_each_possible_cpu(cpu) { - p = &per_cpu(bnx2i_percpu, cpu); + p = per_cpu_ptr(&bnx2i_percpu, cpu); INIT_LIST_HEAD(&p->work_list); spin_lock_init(&p->p_work_lock); - p->iothread = NULL; + INIT_WORK(&p->work, bnx2i_percpu_io_work); } - cpu_notifier_register_begin(); - - for_each_online_cpu(cpu) - bnx2i_percpu_thread_create(cpu); - - /* Initialize per CPU interrupt thread */ - __register_hotcpu_notifier(&bnx2i_cpu_notifier); - - cpu_notifier_register_done(); - return 0; unreg_xport: @@ -587,14 +482,12 @@ static void __exit bnx2i_mod_exit(void) } mutex_unlock(&bnx2i_dev_lock); - cpu_notifier_register_begin(); + for_each_possible_cpu(cpu) { + struct bnx2i_percpu_s *p; - for_each_online_cpu(cpu) - bnx2i_percpu_thread_destroy(cpu); - - __unregister_hotcpu_notifier(&bnx2i_cpu_notifier); - - cpu_notifier_register_done(); + p = per_cpu_ptr(&bnx2i_percpu, cpu); + flush_work(&p->work); + } iscsi_unregister_transport(&bnx2i_iscsi_transport); cnic_unregister_driver(CNIC_ULP_ISCSI);
The driver creates its own per-CPU threads which are updated based on CPU hotplug events. It is also possible to use kworkers and remove some of the infrastructure get the same job done while saving a few lines of code. The DECLARE_PER_CPU() definition is moved into the header file where it belongs. bnx2i_percpu_io_thread() becomes bnx2i_percpu_io_work() which is mostly the same code. The outer loop (kthread_should_stop()) gets removed and the remaining code is shifted to the left. bnx2i_queue_scsi_cmd_resp() is mostly the same. The code checked ->iothread to decide if there is an active per-CPU thread. With the kworkers this is no longer possible nor required. The allocation of struct bnx2i_work does not happen with ->p_work_lock held which is not required. I am unsure about the call-stack so I can't say if this qualifies it for the allocation with GFP_KERNEL instead of GFP_ATOMIC (it is not _bh lock but as I said, I don't know the context). The allocation case has been reversed so the inner if case is called on !bnx2i_work and is just the invocation one function since the lock is not held during allocation. The init of the new bnx2i_work struct is now done also without the ->p_work_lock held: it is a new object, nobody knows about it yet. It should be enough to hold the lock while adding this item to the list. I am unsure about that atomic_inc() so I keep things as they were. The remaining part is the removal CPU hotplug notifier since it is taken care by the kworker code. This patch was only compile-tested due to -ENODEV. Cc: QLogic-Storage-Upstream@qlogic.com Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de> --- drivers/scsi/bnx2i/bnx2i.h | 11 +--- drivers/scsi/bnx2i/bnx2i_hwi.c | 101 ++++++++++++++------------------- drivers/scsi/bnx2i/bnx2i_init.c | 121 +++------------------------------------- 3 files changed, 53 insertions(+), 180 deletions(-)