Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1641050
| From | Anshuman Khandual <khandual@linux.vnet.ibm.com> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | Re: [PATCH] mm/madvise: Dont poison entire HugeTLB page for single page errors |
| Date | 2017-05-14 04:40 +0200 |
| Message-ID | <tGRBT-400-3@gated-at.bofh.it> (permalink) |
| References | <tycvT-4JA-1@gated-at.bofh.it> <tyi8i-8cx-21@gated-at.bofh.it> <tGdXP-1DJ-3@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
On 05/12/2017 01:40 PM, Naoya Horiguchi wrote: > On Thu, Apr 20, 2017 at 04:36:27PM +0530, Anshuman Khandual wrote: >> Currently soft_offline_page() migrates the entire HugeTLB page, then >> dequeues it from the active list by making it a dangling HugeTLB page >> which ofcourse can not be used further and marks the entire HugeTLB >> page as poisoned. This might be a costly waste of memory if the error >> involved affects only small section of the entire page. >> >> This changes the behaviour so that only the affected page is marked >> poisoned and then the HugeTLB page is released back to buddy system. > Hi Anshuman, > > This is a good catch, and we can solve this issue now because freeing > hwpoisoned page (previously forbidden) is available now. > > And I'm thinking that the same issue for hard/soft-offline on free > hugepages can be solved, so I'll submit a patchset which includes > updated version of your patch. Sure, thanks Naoya.
Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread
Re: [PATCH] mm/madvise: Dont poison entire HugeTLB page for single page errors Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> - 2017-05-12 10:20 +0200 Re: [PATCH] mm/madvise: Dont poison entire HugeTLB page for single page errors Anshuman Khandual <khandual@linux.vnet.ibm.com> - 2017-05-14 04:40 +0200
csiph-web