On Wed, Jan 29, 2025 at 10:17:49AM -0800, Dave Hansen wrote:
> tl;dr: The VFS and several filesystems have some suspect prefaulting
> code. It is unnecessarily slow for the common case where a write's
> source buffer is resident and does not need to be faulted in.
>
> Move these "prefaulting" oper
On Thu, Jan 30, 2025 at 12:17:51AM +0100, Andreas Gruenbacher wrote:
> On Wed, Jan 29, 2025 at 10:30 PM Kent Overstreet
> wrote:
> > On Wed, Jan 29, 2025 at 10:21:31PM +0100, Andreas Gruenbacher wrote:
> > > On Wed, Jan 29, 2025 at 9:28 PM Kent Overstreet
> > > wrote:
> > > > On Wed, Jan 29, 2025
On Wed, Jan 29, 2025 at 10:30 PM Kent Overstreet
wrote:
> On Wed, Jan 29, 2025 at 10:21:31PM +0100, Andreas Gruenbacher wrote:
> > On Wed, Jan 29, 2025 at 9:28 PM Kent Overstreet
> > wrote:
> > > On Wed, Jan 29, 2025 at 07:21:49PM +0100, Andreas Gruenbacher wrote:
> > > > On Wed, Jan 29, 2025 at
On Wed, Jan 29, 2025 at 10:21:31PM +0100, Andreas Gruenbacher wrote:
> On Wed, Jan 29, 2025 at 9:28 PM Kent Overstreet
> wrote:
> > On Wed, Jan 29, 2025 at 07:21:49PM +0100, Andreas Gruenbacher wrote:
> > > On Wed, Jan 29, 2025 at 7:04 PM Kent Overstreet
> > > wrote:
> > > > On Tue, Jan 28, 2025
On Wed, Jan 29, 2025 at 9:28 PM Kent Overstreet
wrote:
> On Wed, Jan 29, 2025 at 07:21:49PM +0100, Andreas Gruenbacher wrote:
> > On Wed, Jan 29, 2025 at 7:04 PM Kent Overstreet
> > wrote:
> > > On Tue, Jan 28, 2025 at 05:38:48PM +0100, Andreas Gruenbacher wrote:
> > > > Rename eytzinger0_find_te
Add an ioctl for querying counters, the same ones provided in
/sys/fs/bcachefs//counters/, but more suitable for a 'bcachefs
top' command.
Signed-off-by: Kent Overstreet
---
fs/bcachefs/bcachefs_ioctl.h | 10 ++
fs/bcachefs/chardev.c| 32
fs/bcach
Previously, bch2_bkey_sectors_need_rebalance() called
bch2_target_accepts_data(), checking whether the target is writable.
However, this means that adding or removing devices from a target would
change the value of bch2_bkey_sectors_need_rebalance() for an existing
extent; this needs to be invaria
The discard path is supposed to issue journal flushes when there's too
many buckets empty buckets that need a journal commit before they can be
written to again, but at some point this code seems to have been lost.
Bring it back with a new optimization to make sure we don't issue too
many journal
On Wed, Jan 29, 2025 at 07:21:49PM +0100, Andreas Gruenbacher wrote:
> On Wed, Jan 29, 2025 at 7:04 PM Kent Overstreet
> wrote:
> > On Tue, Jan 28, 2025 at 05:38:48PM +0100, Andreas Gruenbacher wrote:
> > > Rename eytzinger0_find_test_val() to eytzinger0_find_test_le() and add a
> > > new eytzinge
Hi Linus, just some small stuff.
users seeing specific (reproducable) performance issues should ping me,
I have some new tools for tracking those down
The following changes since commit ff0b7ed607e779f0e109f7f24388e0ce07af2ebe:
bcachefs: Fix check_inode_hash_info_matches_root() (2025-01-15 15:
On Wed, Jan 29, 2025 at 7:04 PM Kent Overstreet
wrote:
> On Tue, Jan 28, 2025 at 05:38:48PM +0100, Andreas Gruenbacher wrote:
> > Rename eytzinger0_find_test_val() to eytzinger0_find_test_le() and add a
> > new eytzinger0_find_test_val() wrapper that calls it.
> >
> > We have already established t
From: Dave Hansen
Prefaulting the write source buffer incurs an extra userspace access
in the common fast path. Make bch2_buffered_write() consistent with
generic_perform_write(): only touch userspace an extra time when
copy_page_from_iter_atomic() has failed to make progress.
This also zaps a
tl;dr: The VFS and several filesystems have some suspect prefaulting
code. It is unnecessarily slow for the common case where a write's
source buffer is resident and does not need to be faulted in.
Move these "prefaulting" operations to slow paths where they ensure
forward progress but they do not
On Tue, Jan 28, 2025 at 05:38:48PM +0100, Andreas Gruenbacher wrote:
> Rename eytzinger0_find_test_val() to eytzinger0_find_test_le() and add a
> new eytzinger0_find_test_val() wrapper that calls it.
>
> We have already established that the array is sorted in eytzinger order,
> so we can use the e
On Tue, Jan 28, 2025 at 05:38:57PM +0100, Andreas Gruenbacher wrote:
> In this second step, transform the eytzinger indexes i, j, and k in
> eytzinger1_sort_r() from 0-based to 1-based. This step looks a bit
> messy, but the resulting code is slightly better.
I really hate the cute and paste of l
On Thu, Jan 30, 2025 at 02:17:44AM +0900, Jeongjun Park wrote:
> In the previous commit b3d82c2f2761, code was added to prevent journal
> sequence
> overflow. Among them, the code added to journal_entry_open() uses the
> bch2_fs_fatal_err_on() function to handle errors.
>
> However, __journal_res
In the previous commit b3d82c2f2761, code was added to prevent journal sequence
overflow. Among them, the code added to journal_entry_open() uses the
bch2_fs_fatal_err_on() function to handle errors.
However, __journal_res_get() , which calls journal_entry_open() , calls
journal_entry_open() while
On Wed, Jan 29, 2025 at 11:32 AM Eric Herman wrote:
> Hello Andreas,
>
> On 2025-01-28 17:38, Andreas Gruenbacher wrote:
> > This statement does nothing.
>
> I would suggest this statement does nothing only in the case of #ifndef
> CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS.
>
> In the case where CONF
Hello Andreas,
On 2025-01-28 17:38, Andreas Gruenbacher wrote:
This statement does nothing.
I would suggest this statement does nothing only in the case of #ifndef
CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS.
In the case where CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS _is_ defined,
it tells the co
Hello,
syzbot found the following issue on:
HEAD commit:ab18b8fff124 Merge tag 'auxdisplay-v6.14-1' of git://git.k..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=13a4756458
kernel config: https://syzkaller.appspot.com/x/.config?x=ac8df499d47c7efd
das
20 matches
Mail list logo
