On 2018/10/12 6:20, Sotirios-Efstathios Maneas wrote:
> The following patch adds a few missing statements related to error checking.
>
> Signed-off-by: Sotirios-Efstathios Maneas
> ---
> mkfs/f2fs_format.c | 10 --
> mkfs/f2fs_format_utils.c | 5 +
> 2 files changed, 13
On 10/16/18 1:23 PM, Jaegeuk Kim wrote:
> Thanks Jens,
>
> On top of the patch killing the dead code, I wrote another one to detect
> the idle time by the internal account logic like below. IMHO, it'd be
> better to decouple f2fs with other layers, if possible.
I agree, that's what got us into
Thanks Jens,
On top of the patch killing the dead code, I wrote another one to detect
the idle time by the internal account logic like below. IMHO, it'd be
better to decouple f2fs with other layers, if possible.
Thanks,
>From 85daf5190671b3d98ef779bdea77b4a046658708 Mon Sep 17 00:00:00 2001
On 10/16, Jens Axboe wrote:
> On 10/16/18 1:23 PM, Jaegeuk Kim wrote:
> > Thanks Jens,
> >
> > On top of the patch killing the dead code, I wrote another one to detect
> > the idle time by the internal account logic like below. IMHO, it'd be
> > better to decouple f2fs with other layers, if
On 10/16/18 1:31 PM, Jaegeuk Kim wrote:
> On 10/16, Jens Axboe wrote:
>> On 10/16/18 1:23 PM, Jaegeuk Kim wrote:
>>> Thanks Jens,
>>>
>>> On top of the patch killing the dead code, I wrote another one to detect
>>> the idle time by the internal account logic like below. IMHO, it'd be
>>> better to
On 10/16/18 10:06 AM, Chao Yu wrote:
> Hi Jens,
>
> On 2018-10-16 22:34, Jens Axboe wrote:
>> This doesn't work on stacked devices, and it doesn't work on
>> blk-mq devices. The request_list is only used on legacy, which
>> we don't have much of anymore, and soon won't have any of.
>>
>> Kill the
Jaegeuk,
Sorry for the long delay, I'm busy on other thing.
I'm trying your fixing code on both fsck and kernel with 'run.sh
por_fsstress' case.
And got below output, is that normal in updated fsck? I didn't have time to
look into this.
Info: checkpoint state = 8c6 : quota_need_fsck nat_bits
On 2018/10/17 0:20, Jens Axboe wrote:
> On 10/16/18 10:06 AM, Chao Yu wrote:
>> Hi Jens,
>>
>> On 2018-10-16 22:34, Jens Axboe wrote:
>>> This doesn't work on stacked devices, and it doesn't work on
>>> blk-mq devices. The request_list is only used on legacy, which
>>> we don't have much of
Hi Jaegeuk, Jens,
On 2018/10/17 3:23, Jaegeuk Kim wrote:
> Thanks Jens,
>
> On top of the patch killing the dead code, I wrote another one to detect
> the idle time by the internal account logic like below. IMHO, it'd be
> better to decouple f2fs with other layers, if possible.
>
> Thanks,
>
In 32-bits platform, {f,}stat on a large size file during mkfs, it will
cause EOVERFLOW error, this patch fixes to add macro definition
_FILE_OFFSET_BITS to avoid that error.
Signed-off-by: Chao Yu
---
lib/libf2fs.c| 1 +
mkfs/f2fs_format_utils.c | 4
2 files changed, 5
On 2018/10/17 10:19, Jaegeuk Kim wrote:
> On 10/17, Chao Yu wrote:
>> Hi Jaegeuk, Jens,
>>
>> On 2018/10/17 3:23, Jaegeuk Kim wrote:
>>> Thanks Jens,
>>>
>>> On top of the patch killing the dead code, I wrote another one to detect
>>> the idle time by the internal account logic like below. IMHO,
On 2018/10/17 10:34, Jaegeuk Kim wrote:
> This reverts commit 66110abc4c931f879d70e83e1281f891699364bf.
>
> If we clear the cold data flag out of the writeback flow, we can miscount
> -1 by end_io.
>
> Balancing F2FS Async:
> - IO (CP:1, Data: -1, Flush: ( 001), Discard: ( ...
On 2018/10/16 11:10, Jaegeuk Kim wrote:
> On 10/16, Chao Yu wrote:
>> On 2018/10/16 7:08, Jaegeuk Kim wrote:
>>> On 10/15, Chao Yu wrote:
On 2018/10/11 5:22, Jaegeuk Kim wrote:
> If we clear the cold data flag out of the writeback flow, we can miscount
> -1 by end_io.
I
On 10/17, Chao Yu wrote:
> On 2018/10/17 10:19, Jaegeuk Kim wrote:
> > On 10/17, Chao Yu wrote:
> >> Hi Jaegeuk, Jens,
> >>
> >> On 2018/10/17 3:23, Jaegeuk Kim wrote:
> >>> Thanks Jens,
> >>>
> >>> On top of the patch killing the dead code, I wrote another one to detect
> >>> the idle time by the
On 10/17, Chao Yu wrote:
> Hi Jaegeuk, Jens,
>
> On 2018/10/17 3:23, Jaegeuk Kim wrote:
> > Thanks Jens,
> >
> > On top of the patch killing the dead code, I wrote another one to detect
> > the idle time by the internal account logic like below. IMHO, it'd be
> > better to decouple f2fs with
This reverts commit 66110abc4c931f879d70e83e1281f891699364bf.
If we clear the cold data flag out of the writeback flow, we can miscount
-1 by end_io.
Balancing F2FS Async:
- IO (CP:1, Data: -1, Flush: ( 001), Discard: ( ...
GC thread: IRQ
-
On 10/17, Chao Yu wrote:
> On 2018/10/16 11:10, Jaegeuk Kim wrote:
> > On 10/16, Chao Yu wrote:
> >> On 2018/10/16 7:08, Jaegeuk Kim wrote:
> >>> On 10/15, Chao Yu wrote:
> On 2018/10/11 5:22, Jaegeuk Kim wrote:
> > If we clear the cold data flag out of the writeback flow, we can
> >
On 2018/10/12 18:49, Sheng Yong wrote:
> During recover, we will try to create new dentries for inodes with
> dentry_mark. But if the parent is missing (e.g. killed by fsck),
> recover will break. But those recovered dirty pages are not cleanup.
> This will hit f2fs_bug_on:
>
> [ 53.519566]
This doesn't work on stacked devices, and it doesn't work on
blk-mq devices. The request_list is only used on legacy, which
we don't have much of anymore, and soon won't have any of.
Kill the check.
Cc: Jaegeuk Kim
Cc: linux-f2fs-devel@lists.sourceforge.net
Signed-off-by: Jens Axboe
---
Hi Jens,
On 2018-10-16 22:34, Jens Axboe wrote:
> This doesn't work on stacked devices, and it doesn't work on
> blk-mq devices. The request_list is only used on legacy, which
> we don't have much of anymore, and soon won't have any of.
>
> Kill the check.
In order to avoid conflicting with
20 matches
Mail list logo