Re: new panic in 5.5-rc1 (parisc, kmem_cache_open)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Jan 07, 2020 at 09:52:20AM -0500, Mikulas Patocka wrote:
> 
> On Mon, 9 Dec 2019, Jeroen Roovers wrote:
> 
> > On Mon, 9 Dec 2019 20:22:17 +0200
> > Meelis Roos <mroos@xxxxxxxx> wrote:
> > 
> > > I tried 5.5-rc1 on RP3440 and got a surprise panic on boot (5.4
> > > worked):
> > 
> > Ah, I was going to reported this, too.
> > 
> > > mem auto-init: stack:off, heap alloc:off, heap free:off
> > > Memory: 12348004K/12580864K available (6438K kernel code, 2148K
> > > rwdata, 954K rodata, 1193K init, 500K bss, 232860K reserved, 0K
> > > cma-reserved)
> > 
> > I see an additional message after this:
> > 
> > random: get_random_u64 called from __kmem_cache_create+0x70/0x660 with
> > crng_init=0
> > 
> > And then the same or similar enough backtrace.
> > 
> > 
> > Regards,
> >      jer
> 
> Hi
> 
> I also have this panic. I bisected it and it is caused by the patch 
> d96885e277b5edcd1e474e8b1579005163f23dbe.

My guess would be that pgd is not populated after the change.
Can you please check if the below hack fixes the crash?


diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c
index ddca8287d43b..354cf060b67f 100644
--- a/arch/parisc/mm/init.c
+++ b/arch/parisc/mm/init.c
@@ -401,7 +401,7 @@ static void __init map_pages(unsigned long start_vaddr,
 			pmd = (pmd_t *) __pa(pmd);
 		}
 
-		pgd_populate(NULL, pg_dir, __va(pmd));
+		pud_populate(NULL, (pud_t *)pg_dir, __va(pmd));
 #endif
 		pg_dir++;
 
 
> Unfortunatelly, the patch can't be reverted because pre-patch code depends 
> on the file asm-generic/4level-fixup.h that is missing in the 5.5-rc
> branch.
> 
> Mikulas
 
-- 
Sincerely yours,
Mike.




[Index of Archives]     [Linux SoC]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux