Messages in this thread |  | | From | Kees Cook <> | | Date | Tue, 31 Oct 2017 11:28:34 -0700 | | Subject | Re: [PATCH] scsi: qla2xxx: Convert timers to use timer_setup() |
| |
On Tue, Oct 31, 2017 at 11:16 AM, Madhani, Himanshu <Himanshu.Madhani@cavium.com> wrote: > Hi Bart, > >> On Oct 31, 2017, at 11:12 AM, Bart Van Assche <bart.vanassche@wdc.com> wrote: >> >> On Tue, 2017-10-31 at 18:03 +0000, Madhani, Himanshu wrote: >>> On Oct 31, 2017, at 8:49 AM, Martin K. Petersen <martin.petersen@oracle.com> wrote: >>>>> In preparation for unconditionally passing the struct timer_list >>>>> pointer to all timer callbacks, switch to using the new timer_setup() >>>>> and from_timer() to pass the timer pointer explicitly. >>>> >>>> Cavium folks: Please verify! >>> >>> I’ve checked on my setup and i am seeing regression with this patch applied. >>> >>> Driver load locks up the system with this patch applied.
How very strange. I don't see any code change with this patch. Even the passed arguments are the same; it's only enforcing the types now. And the system is fine with only this patch reverted?
-Kees
>>> >>> NACK >> >> That feedback is not very helpful … >> > > Here’s my setup details > > 4.14.0-rc7 + scsi_misc-4.15 qla2xxx patches from Martin’s tree + current patch. > > System has 3 adapters 8G/16G/32G. > >> Anyway, what kernel source tree did you use in your testing? I may be able to >> free up some time to look into this myself. >> >> Bart. > > Thanks, > - Himanshu
-- Kees Cook Pixel Security
|  |