--- /dev/null
+From 11745ecfe8fea4b4a4c322967a7605d2ecbd5080 Mon Sep 17 00:00:00 2001
+From: Stephane Eranian <eranian@google.com>
+Date: Wed, 3 Aug 2022 09:00:31 -0700
+Subject: perf/x86/intel/uncore: Fix broken read_counter() for SNB IMC PMU
+
+From: Stephane Eranian <eranian@google.com>
+
+commit 11745ecfe8fea4b4a4c322967a7605d2ecbd5080 upstream.
+
+Existing code was generating bogus counts for the SNB IMC bandwidth counters:
+
+$ perf stat -a -I 1000 -e uncore_imc/data_reads/,uncore_imc/data_writes/
+ 1.000327813 1,024.03 MiB uncore_imc/data_reads/
+ 1.000327813 20.73 MiB uncore_imc/data_writes/
+ 2.000580153 261,120.00 MiB uncore_imc/data_reads/
+ 2.000580153 23.28 MiB uncore_imc/data_writes/
+
+The problem was introduced by commit:
+ 07ce734dd8ad ("perf/x86/intel/uncore: Clean up client IMC")
+
+Where the read_counter callback was replace to point to the generic
+uncore_mmio_read_counter() function.
+
+The SNB IMC counters are freerunnig 32-bit counters laid out contiguously in
+MMIO. But uncore_mmio_read_counter() is using a readq() call to read from
+MMIO therefore reading 64-bit from MMIO. Although this is okay for the
+uncore_perf_event_update() function because it is shifting the value based
+on the actual counter width to compute a delta, it is not okay for the
+uncore_pmu_event_start() which is simply reading the counter and therefore
+priming the event->prev_count with a bogus value which is responsible for
+causing bogus deltas in the perf stat command above.
+
+The fix is to reintroduce the custom callback for read_counter for the SNB
+IMC PMU and use readl() instead of readq(). With the change the output of
+perf stat is back to normal:
+$ perf stat -a -I 1000 -e uncore_imc/data_reads/,uncore_imc/data_writes/
+ 1.000120987 296.94 MiB uncore_imc/data_reads/
+ 1.000120987 138.42 MiB uncore_imc/data_writes/
+ 2.000403144 175.91 MiB uncore_imc/data_reads/
+ 2.000403144 68.50 MiB uncore_imc/data_writes/
+
+Fixes: 07ce734dd8ad ("perf/x86/intel/uncore: Clean up client IMC")
+Signed-off-by: Stephane Eranian <eranian@google.com>
+Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
+Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
+Link: https://lore.kernel.org/r/20220803160031.1379788-1-eranian@google.com
+Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
+---
+ arch/x86/events/intel/uncore_snb.c | 18 +++++++++++++++++-
+ 1 file changed, 17 insertions(+), 1 deletion(-)
+
+--- a/arch/x86/events/intel/uncore_snb.c
++++ b/arch/x86/events/intel/uncore_snb.c
+@@ -575,6 +575,22 @@ int snb_pci2phy_map_init(int devid)
+ return 0;
+ }
+
++static u64 snb_uncore_imc_read_counter(struct intel_uncore_box *box, struct perf_event *event)
++{
++ struct hw_perf_event *hwc = &event->hw;
++
++ /*
++ * SNB IMC counters are 32-bit and are laid out back to back
++ * in MMIO space. Therefore we must use a 32-bit accessor function
++ * using readq() from uncore_mmio_read_counter() causes problems
++ * because it is reading 64-bit at a time. This is okay for the
++ * uncore_perf_event_update() function because it drops the upper
++ * 32-bits but not okay for plain uncore_read_counter() as invoked
++ * in uncore_pmu_event_start().
++ */
++ return (u64)readl(box->io_addr + hwc->event_base);
++}
++
+ static struct pmu snb_uncore_imc_pmu = {
+ .task_ctx_nr = perf_invalid_context,
+ .event_init = snb_uncore_imc_event_init,
+@@ -594,7 +610,7 @@ static struct intel_uncore_ops snb_uncor
+ .disable_event = snb_uncore_imc_disable_event,
+ .enable_event = snb_uncore_imc_enable_event,
+ .hw_config = snb_uncore_imc_hw_config,
+- .read_counter = uncore_mmio_read_counter,
++ .read_counter = snb_uncore_imc_read_counter,
+ };
+
+ static struct intel_uncore_type snb_uncore_imc = {
--- /dev/null
+From d957e7ffb2c72410bcc1a514153a46719255a5da Mon Sep 17 00:00:00 2001
+From: Saurabh Sengar <ssengar@linux.microsoft.com>
+Date: Thu, 4 Aug 2022 08:55:34 -0700
+Subject: scsi: storvsc: Remove WQ_MEM_RECLAIM from storvsc_error_wq
+
+From: Saurabh Sengar <ssengar@linux.microsoft.com>
+
+commit d957e7ffb2c72410bcc1a514153a46719255a5da upstream.
+
+storvsc_error_wq workqueue should not be marked as WQ_MEM_RECLAIM as it
+doesn't need to make forward progress under memory pressure. Marking this
+workqueue as WQ_MEM_RECLAIM may cause deadlock while flushing a
+non-WQ_MEM_RECLAIM workqueue. In the current state it causes the following
+warning:
+
+[ 14.506347] ------------[ cut here ]------------
+[ 14.506354] workqueue: WQ_MEM_RECLAIM storvsc_error_wq_0:storvsc_remove_lun is flushing !WQ_MEM_RECLAIM events_freezable_power_:disk_events_workfn
+[ 14.506360] WARNING: CPU: 0 PID: 8 at <-snip->kernel/workqueue.c:2623 check_flush_dependency+0xb5/0x130
+[ 14.506390] CPU: 0 PID: 8 Comm: kworker/u4:0 Not tainted 5.4.0-1086-azure #91~18.04.1-Ubuntu
+[ 14.506391] Hardware name: Microsoft Corporation Virtual Machine/Virtual Machine, BIOS Hyper-V UEFI Release v4.1 05/09/2022
+[ 14.506393] Workqueue: storvsc_error_wq_0 storvsc_remove_lun
+[ 14.506395] RIP: 0010:check_flush_dependency+0xb5/0x130
+ <-snip->
+[ 14.506408] Call Trace:
+[ 14.506412] __flush_work+0xf1/0x1c0
+[ 14.506414] __cancel_work_timer+0x12f/0x1b0
+[ 14.506417] ? kernfs_put+0xf0/0x190
+[ 14.506418] cancel_delayed_work_sync+0x13/0x20
+[ 14.506420] disk_block_events+0x78/0x80
+[ 14.506421] del_gendisk+0x3d/0x2f0
+[ 14.506423] sr_remove+0x28/0x70
+[ 14.506427] device_release_driver_internal+0xef/0x1c0
+[ 14.506428] device_release_driver+0x12/0x20
+[ 14.506429] bus_remove_device+0xe1/0x150
+[ 14.506431] device_del+0x167/0x380
+[ 14.506432] __scsi_remove_device+0x11d/0x150
+[ 14.506433] scsi_remove_device+0x26/0x40
+[ 14.506434] storvsc_remove_lun+0x40/0x60
+[ 14.506436] process_one_work+0x209/0x400
+[ 14.506437] worker_thread+0x34/0x400
+[ 14.506439] kthread+0x121/0x140
+[ 14.506440] ? process_one_work+0x400/0x400
+[ 14.506441] ? kthread_park+0x90/0x90
+[ 14.506443] ret_from_fork+0x35/0x40
+[ 14.506445] ---[ end trace 2d9633159fdc6ee7 ]---
+
+Link: https://lore.kernel.org/r/1659628534-17539-1-git-send-email-ssengar@linux.microsoft.com
+Fixes: 436ad9413353 ("scsi: storvsc: Allow only one remove lun work item to be issued per lun")
+Reviewed-by: Michael Kelley <mikelley@microsoft.com>
+Signed-off-by: Saurabh Sengar <ssengar@linux.microsoft.com>
+Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
+Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
+---
+ drivers/scsi/storvsc_drv.c | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+--- a/drivers/scsi/storvsc_drv.c
++++ b/drivers/scsi/storvsc_drv.c
+@@ -1846,7 +1846,7 @@ static int storvsc_probe(struct hv_devic
+ */
+ host_dev->handle_error_wq =
+ alloc_ordered_workqueue("storvsc_error_wq_%d",
+- WQ_MEM_RECLAIM,
++ 0,
+ host->host_no);
+ if (!host_dev->handle_error_wq)
+ goto err_out2;
projects / thirdparty / kernel / stable-queue.git / commitdiff
