From: Jan Kara <jack@suse.cz> To: Ross Zwisler <ross.zwisler@linux.intel.com> Cc: Jan Kara <jack@suse.cz>, linux-nvdimm@lists.01.org, linux-mm@kvack.org, linux-fsdevel@vger.kernel.org, Andrew Morton <akpm@linux-foundation.org>, "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com> Subject: Re: [PATCH 10/21] mm: Move handling of COW faults into DAX code Date: Thu, 17 Nov 2016 10:36:52 +0100 [thread overview] Message-ID: <20161117093652.GS21785@quack2.suse.cz> (raw) In-Reply-To: <20161116212820.GE31337@linux.intel.com> On Wed 16-11-16 14:28:20, Ross Zwisler wrote: > On Fri, Nov 04, 2016 at 05:25:06AM +0100, Jan Kara wrote: > > Move final handling of COW faults from generic code into DAX fault > > handler. That way generic code doesn't have to be aware of peculiarities > > of DAX locking so remove that knowledge and make locking functions > > private to fs/dax.c. > > > > Signed-off-by: Jan Kara <jack@suse.cz> > > --- > > > @@ -1006,13 +1007,14 @@ int dax_iomap_fault(struct vm_area_struct *vma, struct vm_fault *vmf, > > > > if (error) > > goto finish_iomap; > > - if (!radix_tree_exceptional_entry(entry)) { > > + > > + __SetPageUptodate(vmf->cow_page); > > + if (!radix_tree_exceptional_entry(entry)) > > vmf->page = entry; > > I don't think we need to set vmf->page anymore. We would clear it to NULL in > a few lines anyway, and the only call in between is finish_fault(), which > only cares about vmf->cow_page(). This allows us to remove the vmf->page = > NULL line a few lines below as well. Well, I would not like to depend too much on which fields of vm_fault finish_fault() actually uses - we should fill in as much as we have available. But the truth is we sometime have page to fill in into vmf->page and sometimes we don't so in this case I agree filling it in is pointless. Changed. > > @@ -1051,7 +1053,7 @@ int dax_iomap_fault(struct vm_area_struct *vma, struct vm_fault *vmf, > > } > > } > > unlock_entry: > > - if (!locked_status || error) > > + if (vmf_ret != VM_FAULT_LOCKED || error) > > put_locked_mapping_entry(mapping, vmf->pgoff, entry); > > I don't think this is quite right. For example, for dax_load_hole(), if we > can't get a page we put_locked_mapping_entry() and return VM_FAULT_OOM. > Previously this logic would have skipped the second call to > put_locked_mapping_entry(), but now with the strict check against > VM_FAULT_LOCKED put the entry twice. > > Maybe the right thing to do is just fix dax_load_hole() so it never calls > put_locked_mapping_entry(), and leave this check as you have it? Yeah, good catch. Actually I have follow up patches which somewhat clean up dax_iomap_fault() so that page fault is fully completed within dax_iomap_fault() even when instantiating a hole page which makes error handling simpler. But I didn't want to complicate this series with it. So for now I'll do what you suggest. Thanks. Honza -- Jan Kara <jack@suse.com> SUSE Labs, CR _______________________________________________ Linux-nvdimm mailing list Linux-nvdimm@lists.01.org https://lists.01.org/mailman/listinfo/linux-nvdimm
WARNING: multiple messages have this Message-ID (diff)
From: Jan Kara <jack@suse.cz> To: Ross Zwisler <ross.zwisler@linux.intel.com> Cc: Jan Kara <jack@suse.cz>, linux-mm@kvack.org, linux-fsdevel@vger.kernel.org, linux-nvdimm@lists.01.org, Andrew Morton <akpm@linux-foundation.org>, "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com> Subject: Re: [PATCH 10/21] mm: Move handling of COW faults into DAX code Date: Thu, 17 Nov 2016 10:36:52 +0100 [thread overview] Message-ID: <20161117093652.GS21785@quack2.suse.cz> (raw) In-Reply-To: <20161116212820.GE31337@linux.intel.com> On Wed 16-11-16 14:28:20, Ross Zwisler wrote: > On Fri, Nov 04, 2016 at 05:25:06AM +0100, Jan Kara wrote: > > Move final handling of COW faults from generic code into DAX fault > > handler. That way generic code doesn't have to be aware of peculiarities > > of DAX locking so remove that knowledge and make locking functions > > private to fs/dax.c. > > > > Signed-off-by: Jan Kara <jack@suse.cz> > > --- > > > @@ -1006,13 +1007,14 @@ int dax_iomap_fault(struct vm_area_struct *vma, struct vm_fault *vmf, > > > > if (error) > > goto finish_iomap; > > - if (!radix_tree_exceptional_entry(entry)) { > > + > > + __SetPageUptodate(vmf->cow_page); > > + if (!radix_tree_exceptional_entry(entry)) > > vmf->page = entry; > > I don't think we need to set vmf->page anymore. We would clear it to NULL in > a few lines anyway, and the only call in between is finish_fault(), which > only cares about vmf->cow_page(). This allows us to remove the vmf->page = > NULL line a few lines below as well. Well, I would not like to depend too much on which fields of vm_fault finish_fault() actually uses - we should fill in as much as we have available. But the truth is we sometime have page to fill in into vmf->page and sometimes we don't so in this case I agree filling it in is pointless. Changed. > > @@ -1051,7 +1053,7 @@ int dax_iomap_fault(struct vm_area_struct *vma, struct vm_fault *vmf, > > } > > } > > unlock_entry: > > - if (!locked_status || error) > > + if (vmf_ret != VM_FAULT_LOCKED || error) > > put_locked_mapping_entry(mapping, vmf->pgoff, entry); > > I don't think this is quite right. For example, for dax_load_hole(), if we > can't get a page we put_locked_mapping_entry() and return VM_FAULT_OOM. > Previously this logic would have skipped the second call to > put_locked_mapping_entry(), but now with the strict check against > VM_FAULT_LOCKED put the entry twice. > > Maybe the right thing to do is just fix dax_load_hole() so it never calls > put_locked_mapping_entry(), and leave this check as you have it? Yeah, good catch. Actually I have follow up patches which somewhat clean up dax_iomap_fault() so that page fault is fully completed within dax_iomap_fault() even when instantiating a hole page which makes error handling simpler. But I didn't want to complicate this series with it. So for now I'll do what you suggest. Thanks. Honza -- Jan Kara <jack@suse.com> SUSE Labs, CR -- 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:[~2016-11-17 9:36 UTC|newest] Thread overview: 124+ messages / expand[flat|nested] mbox.gz Atom feed top 2016-11-04 4:24 [PATCH 0/21 v4 RESEND] dax: Clear dirty bits after flushing caches Jan Kara 2016-11-04 4:24 ` Jan Kara 2016-11-04 4:24 ` Jan Kara 2016-11-04 4:24 ` [PATCH 01/21] mm: Join struct fault_env and vm_fault Jan Kara 2016-11-04 4:24 ` Jan Kara 2016-11-04 4:24 ` Jan Kara 2016-11-15 21:50 ` Kirill A. Shutemov 2016-11-15 21:50 ` Kirill A. Shutemov 2016-11-16 10:51 ` Peter Zijlstra 2016-11-16 10:51 ` Peter Zijlstra [not found] ` <20161116105132.GR3142-ndre7Fmf5hadTX5a5knrm8zTDFooKrT+cvkQGrU6aU0@public.gmane.org> 2016-11-16 11:01 ` Jan Kara 2016-11-16 11:01 ` Jan Kara [not found] ` <20161116110101.GE21785-4I4JzKEfoa/jFM9bn6wA6Q@public.gmane.org> 2016-11-16 17:21 ` Peter Zijlstra 2016-11-16 17:21 ` Peter Zijlstra 2016-11-17 9:07 ` Jan Kara 2016-11-16 11:13 ` Kirill A. Shutemov 2016-11-04 4:24 ` [PATCH 02/21] mm: Use vmf->address instead of of vmf->virtual_address Jan Kara 2016-11-04 4:24 ` Jan Kara 2016-11-04 4:24 ` Jan Kara 2016-11-15 21:55 ` Kirill A. Shutemov 2016-11-15 21:55 ` Kirill A. Shutemov 2016-11-16 11:05 ` Jan Kara 2016-11-16 11:05 ` Jan Kara 2016-11-16 11:32 ` Kirill A. Shutemov 2016-11-16 11:55 ` Jan Kara 2016-11-16 11:55 ` Jan Kara 2016-11-04 4:24 ` [PATCH 03/21] mm: Use pgoff in struct vm_fault instead of passing it separately Jan Kara 2016-11-04 4:24 ` Jan Kara 2016-11-04 4:24 ` Jan Kara 2016-11-15 22:01 ` Kirill A. Shutemov 2016-11-15 22:01 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 04/21] mm: Use passed vm_fault structure in __do_fault() Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:05 ` Kirill A. Shutemov 2016-11-15 22:05 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 05/21] mm: Trim __do_fault() arguments Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:10 ` Kirill A. Shutemov 2016-11-15 22:10 ` Kirill A. Shutemov 2016-11-16 13:12 ` Jan Kara 2016-11-04 4:25 ` [PATCH 06/21] mm: Use passed vm_fault structure for in wp_pfn_shared() Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:10 ` Kirill A. Shutemov 2016-11-15 22:10 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 07/21] mm: Add orig_pte field into vm_fault Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:14 ` Kirill A. Shutemov 2016-11-15 22:14 ` Kirill A. Shutemov 2016-11-16 20:00 ` Ross Zwisler 2016-11-16 20:00 ` Ross Zwisler 2016-11-04 4:25 ` [PATCH 08/21] mm: Allow full handling of COW faults in ->fault handlers Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:20 ` Kirill A. Shutemov 2016-11-15 22:20 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 09/21] mm: Factor out functionality to finish page faults Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:21 ` Kirill A. Shutemov 2016-11-15 22:21 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 10/21] mm: Move handling of COW faults into DAX code Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:22 ` Kirill A. Shutemov 2016-11-15 22:22 ` Kirill A. Shutemov 2016-11-16 21:28 ` Ross Zwisler 2016-11-17 9:36 ` Jan Kara [this message] 2016-11-17 9:36 ` Jan Kara 2016-11-04 4:25 ` [PATCH 11/21] mm: Remove unnecessary vma->vm_ops check Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:28 ` Kirill A. Shutemov 2016-11-16 13:29 ` Jan Kara 2016-11-16 14:27 ` Kirill A. Shutemov 2016-11-16 14:27 ` Kirill A. Shutemov 2016-11-16 14:43 ` Jan Kara 2016-11-04 4:25 ` [PATCH 12/21] mm: Factor out common parts of write fault handling Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:30 ` Kirill A. Shutemov 2016-11-15 22:30 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 13/21] mm: Pass vm_fault structure into do_page_mkwrite() Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:40 ` Kirill A. Shutemov 2016-11-15 22:40 ` Kirill A. Shutemov 2016-11-16 13:34 ` Jan Kara 2016-11-04 4:25 ` [PATCH 14/21] mm: Use vmf->page during WP faults Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:42 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 15/21] mm: Move part of wp_page_reuse() into the single call site Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:44 ` Kirill A. Shutemov 2016-11-15 22:44 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 16/21] mm: Provide helper for finishing mkwrite faults Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:52 ` Kirill A. Shutemov 2016-11-15 22:52 ` Kirill A. Shutemov 2016-11-16 13:39 ` Jan Kara 2016-11-04 4:25 ` [PATCH 17/21] mm: Change return values of finish_mkwrite_fault() Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-15 22:57 ` Kirill A. Shutemov 2016-11-04 4:25 ` [PATCH 18/21] mm: Export follow_pte() Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` [PATCH 19/21] dax: Make cache flushing protected by entry lock Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` [PATCH 20/21] dax: Protect PTE modification on WP fault by radix tree " Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` [PATCH 21/21] dax: Clear dirty entry tags on cache flush Jan Kara 2016-11-04 4:25 ` Jan Kara 2016-11-04 4:25 ` Jan Kara -- strict thread matches above, loose matches on Subject: below -- 2016-11-01 22:36 [PATCH 0/21 v4] dax: Clear dirty bits after flushing caches Jan Kara 2016-11-01 22:36 ` [PATCH 10/21] mm: Move handling of COW faults into DAX code Jan Kara 2016-11-01 22:36 ` Jan Kara 2016-11-01 22:36 ` Jan Kara
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=20161117093652.GS21785@quack2.suse.cz \ --to=jack@suse.cz \ --cc=akpm@linux-foundation.org \ --cc=kirill.shutemov@linux.intel.com \ --cc=linux-fsdevel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=linux-nvdimm@lists.01.org \ --cc=ross.zwisler@linux.intel.com \ /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.