[Cluster-devel] [GFS2 PATCH v2] GFS2: Limit number of transaction blocks requested for truncates

Hi, On 16 December, I posted a predecessor patch for this problem. Steve Whitehouse had suggested we base the number of allowable blocks on the size of the journal rather than an arbitrary number. This version 2 patch implements that suggestion: --- GFS2: Limit number of transaction blocks

Re: [Cluster-devel] [PATCH 6/9] jbd2: mark the transaction context with the scope GFP_NOFS context

On Thu 15-12-16 15:07:12, Michal Hocko wrote: > From: Michal Hocko > > now that we have memalloc_nofs_{save,restore} api we can mark the whole > transaction context as implicitly GFP_NOFS. All allocations will > automatically inherit GFP_NOFS this way. This means that we do not

Re: [Cluster-devel] [PATCH 7/9] jbd2: make the whole kjournald2 kthread NOFS safe

On Thu 15-12-16 15:07:13, Michal Hocko wrote: > From: Michal Hocko > > kjournald2 is central to the transaction commit processing. As such any > potential allocation from this kernel thread has to be GFP_NOFS. Make > sure to mark the whole kernel thread GFP_NOFS by the

Re: [Cluster-devel] [PATCH 9/9] Revert "ext4: fix wrong gfp type under transaction"

On Thu 15-12-16 15:07:15, Michal Hocko wrote: > From: Michal Hocko > > This reverts commit 216553c4b7f3e3e2beb4981cddca9b2027523928. Now that > the transaction context uses memalloc_nofs_save and all allocations > within the this context inherit GFP_NOFS automatically, there is

Re: [Cluster-devel] [PATCH 8/9] Revert "ext4: avoid deadlocks in the writeback path by using sb_getblk_gfp"

On Thu 15-12-16 15:07:14, Michal Hocko wrote: > From: Michal Hocko > > This reverts commit c45653c341f5c8a0ce19c8f0ad4678640849cb86 because > sb_getblk_gfp is not really needed as > sb_getblk > __getblk_gfp > __getblk_slow > grow_buffers > grow_dev_page >

Re: [Cluster-devel] [PATCH 0/9 v2] scope GFP_NOFS api

On Fri 16-12-16 17:27:28, Mike Galbraith wrote: > On Fri, 2016-12-16 at 16:35 +0100, Michal Hocko wrote: > > On Fri 16-12-16 16:05:58, Mike Galbraith wrote: > > > On Thu, 2016-12-15 at 15:07 +0100, Michal Hocko wrote: > > > > Hi, > > > > I have posted the previous version here [1]. Since then I

Re: [Cluster-devel] [PATCH 2/9] xfs: introduce and use KM_NOLOCKDEP to silence reclaim lockdep false positives

On Thu, Dec 15, 2016 at 03:07:08PM +0100, Michal Hocko wrote: > From: Michal Hocko > > Now that the page allocator offers __GFP_NOLOCKDEP let's introduce > KM_NOLOCKDEP alias for the xfs allocation APIs. While we are at it > also change KM_NOFS users introduced by b17cb364dbbb

[Cluster-devel] [gfs2-utils PATCH] gfs2_edit: Add rgrepair option

Hi, This patch adds a new option, rgrepair, to gfs2_edit. Its purpose is to traverse all the resource groups in the rindex (assuming that the rindex is valid), and for each entry, verify it's really an rgrp, and that its bitmaps are bitmaps. If it finds an rgrp or bitmap block that's corrupt, it