Re: [PATCH] Parallelize mtrr init between cpus

2012-08-28 Thread zhenzhong.duan
On 2012-08-08 11:53, H. Peter Anvin wrote: On 08/07/2012 07:08 PM, zhenzhong.duan wrote: 2012-08-08 00:32, H. Peter Anvin wrote: On 08/07/2012 12:29 AM, zhenzhong.duan wrote: Current code serialize mtrr init with set_atomicity_lock. Mtrr init is quite slow when we bootup on a hvm with la

Re: [PATCH] Parallelize mtrr init between cpus

2012-08-07 Thread H. Peter Anvin
On 08/07/2012 07:08 PM, zhenzhong.duan wrote: 2012-08-08 00:32, H. Peter Anvin wrote: On 08/07/2012 12:29 AM, zhenzhong.duan wrote: Current code serialize mtrr init with set_atomicity_lock. Mtrr init is quite slow when we bootup on a hvm with large mem, vcpus and pci passthroughed devices(e

Re: [PATCH] Parallelize mtrr init between cpus

2012-08-07 Thread zhenzhong.duan
2012-08-08 00:32, H. Peter Anvin wrote: On 08/07/2012 12:29 AM, zhenzhong.duan wrote: Current code serialize mtrr init with set_atomicity_lock. Mtrr init is quite slow when we bootup on a hvm with large mem, vcpus and pci passthroughed devices(eg. 24 vcpus + 90G mem). It took about ~30 mins t

Re: [PATCH] Parallelize mtrr init between cpus

2012-08-07 Thread H. Peter Anvin
On 08/07/2012 12:29 AM, zhenzhong.duan wrote: > Current code serialize mtrr init with set_atomicity_lock. > Mtrr init is quite slow when we bootup on a hvm with large mem, vcpus > and pci passthroughed devices(eg. 24 vcpus + 90G mem). > It took about ~30 mins to bootup, after patch, it took ~2 min.

Re: [PATCH] Parallelize mtrr init between cpus

2012-08-07 Thread zhenzhong.duan
I also post a question to xen-devel maillist. Although this patch fix the long time issue in hvm, but I don't know why hvm would waste such a long time at bootup. Also I'm not sure if this patch is correct in all cases. Maybe I miss something. link for reference: http://www.gossamer-threads.com/lis