On 02/11/17 06:37 PM, Christian König wrote:
Am 02.11.2017 um 18:34 schrieb Michel Dänzer:
From: Michel Dänzer michel.daenzer@amd.com
Memory allocation failure should generally be handled gracefully by callers. In particular, with transparent hugepage support, attempts to allocate huge pages can fail under memory pressure, but the callers fall back to allocating individual pages instead. In that case, there would be spurious
[TTM] Unable to get page %u
error messages in dmesg.
Signed-off-by: Michel Dänzer michel.daenzer@amd.com
Reminds me that I wanted to better test this code path, so how did you triggered that?
1. Enable zswap by writing 1 to /sys/module/zswap/parameters/enabled
2. Create enough memory pressure that the system starts hitting swap. E.g. in my case I think a full LLVM build does the trick. /sys/kernel/debug/zswap/pool_total_size should contain a non-0 value now.
3. Run the piglit gpu profile.
Anyway patch is Reviewed-by: Christian König christian.koenig@amd.com.
Thanks!