Re: [PATCH][RFC] parisc: Use per-pagetable spinlock (v2)

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

 




On Thu, 2 May 2019, Helge Deller wrote:

> On 02.05.19 16:34, John David Anglin wrote:
> > On 2019-05-02 9:43 a.m., Mikulas Patocka wrote:
> >> My obeservation is:
> >>
> >> CONFIG_FLATMEM_MANUAL=y - doesn't compile. With the Helge's patch, it
> >> compiles and works - but it only sees the first 1GiB of memory.
> > I didn't test FLATMEM.
> 
> It should compile now if you check out the for-next branch again.
> And it only sees 1GiB of memory, which is correct for FLATMEM.
> Instead it tells you to turn on CONFIG_SPARSEMEM:
> 
> [0.000000] Large gap in memory detected (786432 pages). Consider turning on CONFIG_SPARSEMEM
> [0.000000] Memory Ranges:
> [0.000000]  0) Start 0x0000000000000000 End 0x000000003fffffff Size   1024 MB
> [0.000000] Total Memory: 1024 MB
> ..
> 
> >> CONFIG_SPARSEMEM_MANUAL=y, CONFIG_SPARSEMEM_VMEMMAP=n - works.
> >> CONFIG_SPARSEMEM_MANUAL=y, CONFIG_SPARSEMEM_VMEMMAP=y - hangs on boot.
> > I thought I selected CONFIG_SPARSEMEM_VMEMMAP but will check.  Have multiple
> > builds with original SPARSEMEM patch that were okay.
> 
> It sometimes hung for me too.
> I think my VMEMMAP patch overwrites other memory and thus only sometimes crashes the machine...
> 
> By the way, I've rebased my for-next tree, fixed a few small issues and dropped the VMEMMAP patch for now.
> Please give it a new try:
> https://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux.git/log/?h=for-next

OK. I confirm that this branch works.

Mikulas

> In addition I addded a for-next-testing branch for further testing of the remaining patches (VMEMMAP, JUMP_LABEL, ...).
> 
> Helge
> 

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

  Powered by Linux