Hi Paul
I guess you have run into the limitation of __update_cache() not supporting
highmem pages. That could be caused by an omission in my patch that forgets to
clear the PG_dcache_dirty flag in flush_icache_page. I will submit an
incremental patch for this.
- Lars
> 9 apr 2015 kl. 00:53
Hi Paul
I guess you have run into the limitation of __update_cache() not supporting
highmem pages. That could be caused by an omission in my patch that forgets to
clear the PG_dcache_dirty flag in flush_icache_page. I will submit an
incremental patch for this.
- Lars
9 apr 2015 kl. 00:53
On Thu, Feb 26, 2015 at 02:16:01PM +0100, Lars Persson wrote:
> This patch set proposes an improved fix for the race condition that
> originally was fixed in commit 2a4a8b1e5d9d ("MIPS: Remove race window
> in page fault handling").
>
> I have used the flush_icache_page API that is marked as
On Thu, Feb 26, 2015 at 02:16:01PM +0100, Lars Persson wrote:
This patch set proposes an improved fix for the race condition that
originally was fixed in commit 2a4a8b1e5d9d (MIPS: Remove race window
in page fault handling).
I have used the flush_icache_page API that is marked as deprecated
This patch set proposes an improved fix for the race condition that
originally was fixed in commit 2a4a8b1e5d9d ("MIPS: Remove race window
in page fault handling").
I have used the flush_icache_page API that is marked as deprecated in
Documentation/cachetlb.txt. There are strong reasons to keep
This patch set proposes an improved fix for the race condition that
originally was fixed in commit 2a4a8b1e5d9d (MIPS: Remove race window
in page fault handling).
I have used the flush_icache_page API that is marked as deprecated in
Documentation/cachetlb.txt. There are strong reasons to keep
6 matches
Mail list logo