On 2018/4/10 10:26, Bart Van Assche wrote:
On Fri, 2018-03-30 at 17:16 +0000, Bart Van Assche wrote:
The test I ran is as follows:
* Clone the linux-next repository
(https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git)
* Use label next-20180329: git checkout next-20180329 -b linux-next-20180329
* Configure the kernel as explained on https://github.com/bvanassche/srp-test.
* Build and install the linux-next kernel in a virtual machine.
* Boot the linux-next kernel.
* Run the srp-test software as follows: srp-test/run_tests -c -d -r 10
* Shut down the virtual machine by running the poweroff command.
During shutdown the rxe driver crashes. That crash does not occur with kernel
v4.16-rc7. Additionally, that crash does not occur if I run the srp-test
software against the mlx4_ib driver. So I think this crash indicates a recently
introduced regression in the rdma_rxe driver. Can you have a look?
An update: with kernel commit 38c23685b273 ("Merge tag 'armsoc-drivers' of
git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc") from Linus' tree
and patch "[PATCH v4] blk-mq: Fix race conditions in request timeout handling"
(https://www.mail-archive.com/linux-block@xxxxxxxxxxxxxxx/msg20165.html)
applied on top I don't see that crash anymore. The version of Linus' tree I
tested includes Jason's initial pull request for v4.17. I assume this means
that the crash I ran into was triggered by the blk-mq timeout handling code
instead of the rxe driver.
Thanks a lot. I will try to reproduce this problem.
If any update, I will let you know as soon as possible.
Zhu Yanjun
Bart.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
