From: Andrea Arcangeli <aarcange@redhat.com> To: Nadav Amit <nadav.amit@gmail.com> Cc: "Jerome Glisse" <jglisse@redhat.com>, "Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>, "open list:MEMORY MANAGEMENT" <linux-mm@kvack.org>, "Dan Williams" <dan.j.williams@intel.com>, "Ross Zwisler" <ross.zwisler@linux.intel.com>, "Linus Torvalds" <torvalds@linux-foundation.org>, "Bernhard Held" <berny156@gmx.de>, "Adam Borowski" <kilobyte@angband.pl>, "Radim Krčmář" <rkrcmar@redhat.com>, "Wanpeng Li" <kernellwp@gmail.com>, "Paolo Bonzini" <pbonzini@redhat.com>, "Takashi Iwai" <tiwai@suse.de>, "Mike Galbraith" <efault@gmx.de>, "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>, axie <axie@amd.com>, "Andrew Morton" <akpm@linux-foundation.org> Subject: Re: [PATCH 02/13] mm/rmap: update to new mmu_notifier semantic Date: Wed, 30 Aug 2017 22:55:38 +0200 [thread overview] Message-ID: <20170830205538.GH13559@redhat.com> (raw) In-Reply-To: <180A2625-E3AB-44BF-A3B7-E687299B9DA9@gmail.com> On Wed, Aug 30, 2017 at 11:40:08AM -0700, Nadav Amit wrote: > The mmu_notifier users would have to be aware that invalidations may be > deferred. If they perform their ``invalidations’’ unconditionally, it may be > ok. If the notifier users avoid invalidations based on the PTE in the > secondary page-table, it can be a problem. invalidate_page was always deferred post PT lock release. This ->invalidate_range post PT lock release, is not a new thing, we're still back to squre one to find out if invalidate_page callout after PT lock release has always been broken here or not. > On another note, you may want to consider combining the secondary page-table > mechanisms with the existing TLB-flush mechanisms. Right now, it is > partially done: tlb_flush_mmu_tlbonly(), for example, calls > mmu_notifier_invalidate_range(). However, tlb_gather_mmu() does not call > mmu_notifier_invalidate_range_start(). If you implement ->invalidate_range_start you don't care about tlb gather at all and you must not implement ->invalidate_range. > This can also prevent all kind of inconsistencies, and potential bugs. For > instance, clear_refs_write() calls mmu_notifier_invalidate_range_start/end() > but in between there is no call for mmu_notifier_invalidate_range(). It's done in mmu_notifier_invalidate_range_end which is again fully equivalent except run after PT lock release.
WARNING: multiple messages have this Message-ID (diff)
From: Andrea Arcangeli <aarcange@redhat.com> To: Nadav Amit <nadav.amit@gmail.com> Cc: "Jerome Glisse" <jglisse@redhat.com>, "Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>, "open list:MEMORY MANAGEMENT" <linux-mm@kvack.org>, "Dan Williams" <dan.j.williams@intel.com>, "Ross Zwisler" <ross.zwisler@linux.intel.com>, "Linus Torvalds" <torvalds@linux-foundation.org>, "Bernhard Held" <berny156@gmx.de>, "Adam Borowski" <kilobyte@angband.pl>, "Radim Krčmář" <rkrcmar@redhat.com>, "Wanpeng Li" <kernellwp@gmail.com>, "Paolo Bonzini" <pbonzini@redhat.com>, "Takashi Iwai" <tiwai@suse.de>, "Mike Galbraith" <efault@gmx.de>, "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>, axie <axie@amd.com>, "Andrew Morton" <akpm@linux-foundation.org> Subject: Re: [PATCH 02/13] mm/rmap: update to new mmu_notifier semantic Date: Wed, 30 Aug 2017 22:55:38 +0200 [thread overview] Message-ID: <20170830205538.GH13559@redhat.com> (raw) In-Reply-To: <180A2625-E3AB-44BF-A3B7-E687299B9DA9@gmail.com> On Wed, Aug 30, 2017 at 11:40:08AM -0700, Nadav Amit wrote: > The mmu_notifier users would have to be aware that invalidations may be > deferred. If they perform their ``invalidationsa??a?? unconditionally, it may be > ok. If the notifier users avoid invalidations based on the PTE in the > secondary page-table, it can be a problem. invalidate_page was always deferred post PT lock release. This ->invalidate_range post PT lock release, is not a new thing, we're still back to squre one to find out if invalidate_page callout after PT lock release has always been broken here or not. > On another note, you may want to consider combining the secondary page-table > mechanisms with the existing TLB-flush mechanisms. Right now, it is > partially done: tlb_flush_mmu_tlbonly(), for example, calls > mmu_notifier_invalidate_range(). However, tlb_gather_mmu() does not call > mmu_notifier_invalidate_range_start(). If you implement ->invalidate_range_start you don't care about tlb gather at all and you must not implement ->invalidate_range. > This can also prevent all kind of inconsistencies, and potential bugs. For > instance, clear_refs_write() calls mmu_notifier_invalidate_range_start/end() > but in between there is no call for mmu_notifier_invalidate_range(). It's done in mmu_notifier_invalidate_range_end which is again fully equivalent except run after PT lock release. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2017-08-30 20:55 UTC|newest] Thread overview: 160+ messages / expand[flat|nested] mbox.gz Atom feed top 2017-08-29 23:54 [PATCH 00/13] mmu_notifier kill invalidate_page callback Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 01/13] dax: update to new mmu_notifier semantic Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 02/13] mm/rmap: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-30 2:46 ` Nadav Amit 2017-08-30 2:46 ` Nadav Amit 2017-08-30 2:59 ` Jerome Glisse 2017-08-30 2:59 ` Jerome Glisse 2017-08-30 3:16 ` Nadav Amit 2017-08-30 3:16 ` Nadav Amit 2017-08-30 3:18 ` Nadav Amit 2017-08-30 3:18 ` Nadav Amit 2017-08-30 17:27 ` Andrea Arcangeli 2017-08-30 17:27 ` Andrea Arcangeli 2017-08-30 18:00 ` Nadav Amit 2017-08-30 18:00 ` Nadav Amit 2017-08-30 21:25 ` Andrea Arcangeli 2017-08-30 21:25 ` Andrea Arcangeli 2017-08-30 23:25 ` Nadav Amit 2017-08-30 23:25 ` Nadav Amit 2017-08-31 0:47 ` Jerome Glisse 2017-08-31 0:47 ` Jerome Glisse 2017-08-31 0:47 ` Jerome Glisse 2017-08-31 17:12 ` Andrea Arcangeli 2017-08-31 17:12 ` Andrea Arcangeli 2017-08-31 17:12 ` Andrea Arcangeli 2017-08-31 19:15 ` Nadav Amit 2017-08-31 19:15 ` Nadav Amit 2017-08-30 18:20 ` Jerome Glisse 2017-08-30 18:20 ` Jerome Glisse 2017-08-30 18:40 ` Nadav Amit 2017-08-30 18:40 ` Nadav Amit 2017-08-30 20:45 ` Jerome Glisse 2017-08-30 20:45 ` Jerome Glisse 2017-08-30 22:17 ` Andrea Arcangeli 2017-08-30 22:17 ` Andrea Arcangeli 2017-08-30 20:55 ` Andrea Arcangeli [this message] 2017-08-30 20:55 ` Andrea Arcangeli 2017-08-30 16:52 ` Andrea Arcangeli 2017-08-30 16:52 ` Andrea Arcangeli 2017-08-30 17:48 ` Jerome Glisse 2017-08-30 17:48 ` Jerome Glisse 2017-08-30 21:53 ` Linus Torvalds 2017-08-30 21:53 ` Linus Torvalds 2017-08-30 23:01 ` Andrea Arcangeli 2017-08-30 23:01 ` Andrea Arcangeli 2017-08-31 18:25 ` Jerome Glisse 2017-08-31 18:25 ` Jerome Glisse 2017-08-31 19:40 ` Linus Torvalds 2017-08-31 19:40 ` Linus Torvalds 2017-08-29 23:54 ` [PATCH 03/13] powerpc/powernv: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 04/13] drm/amdgpu: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-30 6:18 ` Christian König 2017-08-30 6:18 ` Christian König 2017-08-29 23:54 ` [PATCH 05/13] IB/umem: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-30 6:13 ` Leon Romanovsky 2017-08-29 23:54 ` [PATCH 06/13] IB/hfi1: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-09-06 14:08 ` Arumugam, Kamenee 2017-08-29 23:54 ` [PATCH 07/13] iommu/amd: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 08/13] iommu/intel: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 09/13] misc/mic/scif: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 10/13] sgi-gru: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 11/13] xen/gntdev: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-30 19:32 ` Boris Ostrovsky 2017-08-30 19:32 ` Boris Ostrovsky 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 12/13] KVM: " Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse 2017-08-29 23:54 ` [PATCH 13/13] mm/mmu_notifier: kill invalidate_page Jérôme Glisse 2017-08-29 23:54 ` Jérôme Glisse [not found] ` <20170829235447.10050-1-jglisse-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> 2017-08-30 0:11 ` [PATCH 00/13] mmu_notifier kill invalidate_page callback Linus Torvalds 2017-08-30 0:11 ` Linus Torvalds 2017-08-30 0:11 ` Linus Torvalds 2017-08-30 0:11 ` Linus Torvalds 2017-08-30 0:56 ` Jerome Glisse 2017-08-30 0:56 ` Jerome Glisse 2017-08-30 0:56 ` Jerome Glisse 2017-08-30 0:56 ` Jerome Glisse 2017-08-30 8:40 ` Mike Galbraith 2017-08-30 8:40 ` Mike Galbraith 2017-08-30 8:40 ` Mike Galbraith 2017-08-30 8:40 ` Mike Galbraith 2017-08-30 8:40 ` Mike Galbraith 2017-08-30 14:57 ` Adam Borowski 2017-08-30 14:57 ` Adam Borowski 2017-08-30 14:57 ` Adam Borowski 2017-08-30 14:57 ` Adam Borowski 2017-09-01 14:47 ` Jeff Cook 2017-09-01 14:47 ` Jeff Cook 2017-09-01 14:47 ` Jeff Cook 2017-09-01 14:47 ` Jeff Cook 2017-09-01 14:47 ` Jeff Cook 2017-09-01 14:50 ` taskboxtester 2017-09-01 14:50 ` taskboxtester 2017-08-30 21:51 ` Felix Kuehling 2017-08-31 13:59 ` Jerome Glisse [not found] ` <20170831135953.GA9227-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> 2017-08-31 14:14 ` Christian König 2017-08-31 18:39 ` Felix Kuehling 2017-08-31 19:00 ` Jerome Glisse [not found] ` <20170831190021.GG9227-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> 2017-08-31 23:19 ` Felix Kuehling 2017-08-31 23:29 ` Jerome Glisse 2017-08-30 0:11 ` Linus Torvalds 2017-11-30 9:33 ` BSOD with " Fabian Grünbichler 2017-11-30 9:33 ` Fabian Grünbichler 2017-11-30 11:20 ` Paolo Bonzini 2017-11-30 11:20 ` Paolo Bonzini 2017-11-30 11:20 ` Paolo Bonzini 2017-11-30 16:19 ` Radim Krčmář 2017-11-30 16:19 ` Radim Krčmář 2017-11-30 18:05 ` [PATCH 1/2] KVM: x86: fix APIC page invalidation Radim Krčmář 2017-11-30 18:05 ` Radim Krčmář 2017-11-30 18:05 ` Radim Krčmář 2017-11-30 18:05 ` [PATCH 2/2] TESTING! KVM: x86: add invalidate_range mmu notifier Radim Krčmář 2017-11-30 18:05 ` Radim Krčmář 2017-12-01 15:15 ` Paolo Bonzini 2017-12-01 15:15 ` Paolo Bonzini 2017-12-01 15:15 ` Paolo Bonzini 2017-12-03 17:24 ` Andrea Arcangeli 2017-12-03 17:24 ` Andrea Arcangeli 2017-12-03 17:24 ` Andrea Arcangeli 2017-12-01 12:21 ` [PATCH 1/2] KVM: x86: fix APIC page invalidation Fabian Grünbichler 2017-12-01 12:21 ` Fabian Grünbichler 2017-12-01 15:27 ` Paolo Bonzini 2017-12-01 15:27 ` Paolo Bonzini 2017-12-03 17:28 ` Andrea Arcangeli 2017-12-03 17:28 ` Andrea Arcangeli 2017-12-03 17:28 ` Andrea Arcangeli 2017-12-06 2:32 ` Wanpeng Li 2017-12-06 2:32 ` Wanpeng Li 2017-12-06 9:50 ` 王金浦 2017-12-06 9:50 ` 王金浦 2017-12-06 10:00 ` Paolo Bonzini 2017-12-06 10:00 ` Paolo Bonzini 2017-12-06 10:00 ` Paolo Bonzini 2017-12-06 8:15 ` Fabian Grünbichler 2017-12-06 8:15 ` Fabian Grünbichler 2017-12-06 8:15 ` Fabian Grünbichler 2017-12-13 12:54 ` Richard Purdie 2017-12-13 12:54 ` Richard Purdie 2017-12-13 12:54 ` Richard Purdie 2017-11-30 16:19 ` BSOD with [PATCH 00/13] mmu_notifier kill invalidate_page callback Radim Krčmář 2017-11-30 11:20 ` Paolo Bonzini 2017-11-30 9:33 ` Fabian Grünbichler
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20170830205538.GH13559@redhat.com \ --to=aarcange@redhat.com \ --cc=akpm@linux-foundation.org \ --cc=axie@amd.com \ --cc=berny156@gmx.de \ --cc=dan.j.williams@intel.com \ --cc=efault@gmx.de \ --cc=jglisse@redhat.com \ --cc=kernellwp@gmail.com \ --cc=kilobyte@angband.pl \ --cc=kirill.shutemov@linux.intel.com \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=nadav.amit@gmail.com \ --cc=pbonzini@redhat.com \ --cc=rkrcmar@redhat.com \ --cc=ross.zwisler@linux.intel.com \ --cc=tiwai@suse.de \ --cc=torvalds@linux-foundation.org \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.