On Thu, Oct 15, 2020 at 01:17:14PM -0700, si-wei liu wrote:You didn't, sorry. I was looking into a way to speed up the boot time for large memory guest by multi-threading the page pinning process, and it turns out I need more time on that. Will post the fix I have now soon, hopefully tomorrow.
On 10/15/2020 6:11 AM, Michael S. Tsirkin wrote:did I miss the patch?
On Thu, Oct 15, 2020 at 02:15:32PM +0800, Jason Wang wrote:Thanks, Michael & Jason. I will post the fix shortly. Stay tuned.
On 2020/10/14 上午7:42, si-wei liu wrote:Fine by me too.
Fine.So what I suggest is to fix the pinning leakage first and do theOK. Unfortunately, this was picked and got merged in upstream. So I will
possible optimization on top (which is still questionable to me).
post a follow up patch set to 1) revert the commit to the original
__get_free_page() implementation, and 2) fix the accounting and leakage
on top. Will it be fine?
Thanks
-Siwei