On Tue, Jul 05, 2016 at 06:08:46PM +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. > > bnx2fc_percpu_io_thread() becomes bnx2fc_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. > In bnx2fc_process_new_cqes() the code checked for ->iothread to > decide if there is an active per-CPU thread. With the kworkers this > is no longer possible nor required. The allocation of a new work item > (via bnx2fc_alloc_work()) does no longer happen with the ->fp_work_lock > lock held. It performs only a memory allocation + initialization which > does not require any kind of serialization. The lock is held while > adding the new member to fps->work_list list. > > 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@xxxxxxxxxx > Cc: Christoph Hellwig <hch@xxxxxx> > Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx> Hi Sebastian, I've tried to test your series and unfortunately I encountered the following oops. [ 113.969654] Modules linked in: af_packet(E) 8021q(E) garp(E) stp(E) llc(E) mrp(E) mgag200(E) i2c_algo_bit(E) bnx2x(E) drm_kms_helper(E) syscopyarea(E) mdio(E) sysfillrect(E) libcrc32c(E) sysimgblt(E) dm_service_time(E) bnx2fc(E) fb_sys_fops(E) geneve(E) xhci_pci(E) uhci_hcd(E) ttm(E) sd_mod(E) ehci_hcd(E) xhci_hcd(E) cnic(E) vxlan(E) uio(E) ip6_udp_tunnel(E) udp_tunnel(E) crc32c_intel(E) hpsa(E) ptp(E) usbcore(E) drm(E) scsi_transport_sas(E) usb_common(E) pps_core(E) fjes(E) scsi_dh_rdac(E) scsi_dh_emc(E) scsi_dh_alua(E) dm_mirror(E) dm_region_hash(E) dm_log(E) fcoe(E) libfcoe(E) libfc(E) scsi_transport_fc(E) sg(E) dm_multipath(E) dm_mod(E) scsi_mod(E) efivarfs(E) autofs4(E) [ 113.969669] Supported: Yes [ 113.969671] CPU: 11 PID: 398 Comm: kworker/11:1 Tainted: G EL 4.4.15-0.g6f92308-default #1 [ 113.969672] Hardware name: HP ProLiant BL460c Gen9, BIOS I36 12/24/2014 [ 113.969674] Workqueue: events bnx2fc_percpu_io_work [bnx2fc] [ 113.969675] task: ffff880667a38600 ti: ffff880667a3c000 task.ti: ffff880667a3c000 [ 113.969676] RIP: 0010:[<ffffffff810c4c2f>] [<ffffffff810c4c2f>] native_queued_spin_lock_slowpath+0x16f/0x190 [ 113.969677] RSP: 0018:ffff880667a3fde8 EFLAGS: 00000202 [ 113.969678] RAX: 0000000000000101 RBX: ffff88065553cd20 RCX: 0000000000000001 [ 113.969678] RDX: 0000000000000101 RSI: 0000000000000001 RDI: ffff8806676d80d0 [ 113.969679] RBP: ffff880667a3fdf8 R08: 0000000000000101 R09: 0000000000000001 [ 113.969679] R10: ffffffff81d46fc0 R11: 0000000000000000 R12: ffff880667a3fdf8 [ 113.969680] R13: ffff8806676d80a0 R14: ffff8806676d80c0 R15: ffff8806676d80d0 [ 113.969681] FS: 0000000000000000(0000) GS:ffff8806676c0000(0000) knlGS:0000000000000000 [ 113.969681] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 113.969682] CR2: 00007fee621724a0 CR3: 0000000001c0a000 CR4: 00000000001406e0 [ 113.969682] Stack: [ 113.969683] ffffffff81181a17 ffffffffa026a1dc ffff880667a3fdf8 ffff880667a3fdf8 [ 113.969684] ffff8806676d80a0 ffff88017c303b80 ffff8806676d5400 ffffe8fa068c1200 [ 113.969685] 0000000000000000 00000000000002c0 ffffffff8109394e 0000000067a38600 [ 113.969685] Call Trace: [ 113.969688] [<ffffffff81181a17>] queued_spin_lock_slowpath+0x7/0xa [ 113.969690] [<ffffffffa026a1dc>] bnx2fc_percpu_io_work+0xac/0xd0 [bnx2fc] [ 113.969693] [<ffffffff8109394e>] process_one_work+0x14e/0x410 [ 113.969695] [<ffffffff810941a6>] worker_thread+0x116/0x490 [ 113.969697] [<ffffffff8109964d>] kthread+0xbd/0xe0 [ 113.969699] [<ffffffff815debbf>] ret_from_fork+0x3f/0x70 [ 113.970696] DWARF2 unwinder stuck at ret_from_fork+0x3f/0x70 [ 113.970696] [ 113.970696] Leftover inexact backtrace: [ 113.970696] [ 113.970698] [<ffffffff81099590>] ? kthread_park+0x50/0x50 [ 113.970705] Code: 89 d0 66 31 c0 39 c1 74 e7 4d 85 c9 c6 07 01 74 2d 41 c7 41 08 01 00 00 00 e9 52 ff ff ff 83 fa 01 74 17 8b 07 84 c0 74 08 f3 90 <8b> 07 84 c0 75 f8 b8 01 00 00 00 66 89 07 c3 f3 c3 f3 90 4d 8b I'll try to narrow it down to the specific patch, but currently I'm a bit busy, sorry. Johannes -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
