blktests failures with v6.17-rc1 kernel
Hi all,
I ran the latest blktests (git hash: a08678c0cf2d) with the v6.17-rc1 kernel. I
observed 6 failures listed below. Comparing with the previous report with the
v6.16 kernel [1], 2 failures are new.
[1] https://lore.kernel.org/linux-block/g4svtamuk3jhhhzb52reoj3nj2agi4ws7fwyc45vca5uykjkb4@glfr4emapv7n/
List of failures
================
#1: block/005
#2: nvme/041 (fc transport)
#3: nvme/050 (new)
#4: nvme/061 (fc transport)
#5: nvme/063 (tcp transport)
#6: scsi/006 (IDE/PATA device)(new)
Failure description
===================
#1: block/005
When the test case is run for a NVME device as TEST_DEV, the kernel reports
a lockdep WARN related to the three locks q->q_usage_counter, fs_reclaim and
cpu_hotplug_lock. Refer to the report for the v6.16-rc1 kernel [2]. Nilay
posted the fix patch candiate [3].
[2] https://lore.kernel.org/linux-block/4fdm37so3o4xricdgfosgmohn63aa7wj3ua4e5vpihoamwg3ui@fq42f5q5t5ic/
[3] https://lore.kernel.org/linux-block/20250805171749.3448694-1-nilay@linux.ibm.com/
#2: nvme/041 (fc transport)
The test case nvme/041 fails for fc transport. Refer to the report for the
v6.12 kernel [4].
[4] https://lore.kernel.org/linux-nvme/6crydkodszx5vq4ieox3jjpwkxtu7mhbohypy24awlo5w7f4k6@to3dcng24rd4/
#3: nvme/050 (new)
The test case fails with the message below:
nvme/050 => nvme0n1 (test nvme-pci timeout with fio jobs) [failed]
runtime 90.974s ... 90.912s
--- tests/nvme/050.out 2024-09-20 11:20:26.422380826 +0900
+++ /home/shin/Blktests/blktests/results/nvme0n1/nvme/050.out.bad 2025-08-13 11:12:54.655610693 +0900
@@ -1,2 +1,3 @@
Running nvme/050
Test complete
+Failed to restore /dev/nvme0n1
This needs further debug.
#4: nvme/061 (fc transport)
The test case nvme/061 sometimes fails for fc transport due to a WARN and
refcount message "refcount_t: underflow; use-after-free." Refer to the
report for the v6.15 kernel [5].
[5] https://lore.kernel.org/linux-block/2xsfqvnntjx5iiir7wghhebmnugmpfluv6ef22mghojgk6gilr@mvjscqxroqqk/
#5: nvme/063 (tcp transport)
The test case nvme/063 fails for tcp transport due to the lockdep WARN
related to the three locks q->q_usage_counter, q->elevator_lock and
set->srcu. Refer to the report for the v6.16-rc1 kernel [2].
#6: scsi/006 (IDE/PATA device)(new)
When the test case scsi/006 is run for QEMU IDE/PATA devices, it disables
the devices and causes I/O errors. On the worst case, it makes the test
system hang. Damien posted the fix candidate patch [6].
[6] https://lore.kernel.org/linux-ide/20250813092707.447479-1-dlemoal@kernel.org/T/#u
Reply to: