linux-ext4.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andreas Dilger <adilger@dilger.ca>
To: harshad shirwadkar <harshadshirwadkar@gmail.com>
Cc: Ext4 Developers List <linux-ext4@vger.kernel.org>,
	"Theodore Y. Ts'o" <tytso@mit.edu>
Subject: Re: [PATCH v6 7/7] ext4: make prefetch_block_bitmaps default
Date: Fri, 2 Apr 2021 11:39:51 -0600	[thread overview]
Message-ID: <4CA7B7B0-BC40-43D3-A34F-38C92EDB38A4@dilger.ca> (raw)
In-Reply-To: <CAD+ocbwhQva2d7H2E=67_aSSzr0VkR+xiBjEmBTb4ENggvU6Hw@mail.gmail.com>

On Apr 2, 2021, at 10:46, harshad shirwadkar <harshadshirwadkar@gmail.com> wrote:
> 
> Thanks for taking a look!
> 
>> On Thu, Apr 1, 2021 at 10:16 PM Andreas Dilger <adilger@dilger.ca> wrote:
>> 
>>> On Apr 1, 2021, at 11:45, Harshad Shirwadkar <harshadshirwadkar@gmail.com> wrote:
>>> 
>>> Block bitmap prefetching is needed for these allocator optimization
>>> data structures to get populated and provide better group scanning
>>> order. So, turn it on bu default. prefetch_block_bitmaps mount option
>>> is now marked as removed and a new option no_prefetch_block_bitmaps is
>>> added to disable block bitmap prefetching.
>> 
>> This makes it more difficult to change between an old kernel and a new one
>> using this option. It would be better to keep prefetch_block_bitmaps to turn
>> the option on (not harmful if it is already on), and no_* turn it off.
> How so? This patch doesn't get rid of the prefetch_block_bitmaps mount
> option, it just marks it as "removed". So, you can still pass
> "prefetch_block_bitmaps" mount option and get the prefetching
> behavior, the only difference is that you'll get an additional kernel
> message saying that "ignoring removed mount option
> prefetch_block_bitmaps", which I thought is good since looking at that
> message, users will eventually remove that mount option from their
> mount arguments.

Sorry, I guess I was reading too quickly, and wasn't familiar with the
Opt_removed mechanism. It seems OK as-is, you can add:

Reviewed-by: Andreas Dilger <adilger@dilger.ca>

Cheers, Andreas

>>> Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
>>> ---
>>> fs/ext4/ext4.h  |  2 +-
>>> fs/ext4/super.c | 15 ++++++++-------
>>> 2 files changed, 9 insertions(+), 8 deletions(-)
>>> 
>>> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
>>> index 9a5afe9d2310..20c757f711e7 100644
>>> --- a/fs/ext4/ext4.h
>>> +++ b/fs/ext4/ext4.h
>>> @@ -1227,7 +1227,7 @@ struct ext4_inode_info {
>>> #define EXT4_MOUNT_JOURNAL_CHECKSUM    0x800000 /* Journal checksums */
>>> #define EXT4_MOUNT_JOURNAL_ASYNC_COMMIT    0x1000000 /* Journal Async Commit */
>>> #define EXT4_MOUNT_WARN_ON_ERROR    0x2000000 /* Trigger WARN_ON on error */
>>> -#define EXT4_MOUNT_PREFETCH_BLOCK_BITMAPS 0x4000000
>>> +#define EXT4_MOUNT_NO_PREFETCH_BLOCK_BITMAPS 0x4000000
>>> #define EXT4_MOUNT_DELALLOC        0x8000000 /* Delalloc support */
>>> #define EXT4_MOUNT_DATA_ERR_ABORT    0x10000000 /* Abort on file data write */
>>> #define EXT4_MOUNT_BLOCK_VALIDITY    0x20000000 /* Block validity checking */
>>> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
>>> index 6116640081c0..cec0fb07916b 100644
>>> --- a/fs/ext4/super.c
>>> +++ b/fs/ext4/super.c
>>> @@ -1687,7 +1687,7 @@ enum {
>>>   Opt_dioread_nolock, Opt_dioread_lock,
>>>   Opt_discard, Opt_nodiscard, Opt_init_itable, Opt_noinit_itable,
>>>   Opt_max_dir_size_kb, Opt_nojournal_checksum, Opt_nombcache,
>>> -    Opt_prefetch_block_bitmaps, Opt_mb_optimize_scan,
>>> +    Opt_no_prefetch_block_bitmaps, Opt_mb_optimize_scan,
>>> #ifdef CONFIG_EXT4_DEBUG
>>>   Opt_fc_debug_max_replay, Opt_fc_debug_force
>>> #endif
>>> @@ -1787,7 +1787,8 @@ static const match_table_t tokens = {
>>>   {Opt_inlinecrypt, "inlinecrypt"},
>>>   {Opt_nombcache, "nombcache"},
>>>   {Opt_nombcache, "no_mbcache"},    /* for backward compatibility */
>>> -    {Opt_prefetch_block_bitmaps, "prefetch_block_bitmaps"},
>>> +    {Opt_removed, "prefetch_block_bitmaps"},
>>> +    {Opt_no_prefetch_block_bitmaps, "no_prefetch_block_bitmaps"},
>>>   {Opt_mb_optimize_scan, "mb_optimize_scan=%d"},
>>>   {Opt_removed, "check=none"},    /* mount option from ext2/3 */
>>>   {Opt_removed, "nocheck"},    /* mount option from ext2/3 */
>>> @@ -2009,7 +2010,7 @@ static const struct mount_opts {
>>>   {Opt_max_dir_size_kb, 0, MOPT_GTE0},
>>>   {Opt_test_dummy_encryption, 0, MOPT_STRING},
>>>   {Opt_nombcache, EXT4_MOUNT_NO_MBCACHE, MOPT_SET},
>>> -    {Opt_prefetch_block_bitmaps, EXT4_MOUNT_PREFETCH_BLOCK_BITMAPS,
>>> +    {Opt_no_prefetch_block_bitmaps, EXT4_MOUNT_NO_PREFETCH_BLOCK_BITMAPS,
>>>    MOPT_SET},
>>>   {Opt_mb_optimize_scan, EXT4_MOUNT2_MB_OPTIMIZE_SCAN, MOPT_GTE0},
>>> #ifdef CONFIG_EXT4_DEBUG
>>> @@ -3706,11 +3707,11 @@ static struct ext4_li_request *ext4_li_request_new(struct super_block *sb,
>>> 
>>>   elr->lr_super = sb;
>>>   elr->lr_first_not_zeroed = start;
>>> -    if (test_opt(sb, PREFETCH_BLOCK_BITMAPS))
>>> -        elr->lr_mode = EXT4_LI_MODE_PREFETCH_BBITMAP;
>>> -    else {
>>> +    if (test_opt(sb, NO_PREFETCH_BLOCK_BITMAPS)) {
>>>       elr->lr_mode = EXT4_LI_MODE_ITABLE;
>>>       elr->lr_next_group = start;
>>> +    } else {
>>> +        elr->lr_mode = EXT4_LI_MODE_PREFETCH_BBITMAP;
>>>   }
>>> 
>>>   /*
>>> @@ -3741,7 +3742,7 @@ int ext4_register_li_request(struct super_block *sb,
>>>       goto out;
>>>   }
>>> 
>>> -    if (!test_opt(sb, PREFETCH_BLOCK_BITMAPS) &&
>>> +    if (test_opt(sb, NO_PREFETCH_BLOCK_BITMAPS) &&
>>>       (first_not_zeroed == ngroups || sb_rdonly(sb) ||
>>>        !test_opt(sb, INIT_INODE_TABLE)))
>>>       goto out;
>>> --
>>> 2.31.0.291.g576ba9dcdaf-goog
>>> 

  reply	other threads:[~2021-04-02 17:39 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-01 17:21 [PATCH v6 0/7] Block Allocator Improvements Harshad Shirwadkar
2021-04-01 17:21 ` [PATCH v6 1/7] ext4: drop s_mb_bal_lock and convert protected fields to atomic Harshad Shirwadkar
2021-04-01 17:21 ` [PATCH v6 2/7] ext4: add ability to return parsed options from parse_options Harshad Shirwadkar
2021-04-01 17:21 ` [PATCH v6 3/7] ext4: add mballoc stats proc file Harshad Shirwadkar
2021-04-01 17:21 ` [PATCH v6 4/7] ext4: add MB_NUM_ORDERS macro Harshad Shirwadkar
2021-04-01 17:21 ` [PATCH v6 5/7] ext4: improve cr 0 / cr 1 group scanning Harshad Shirwadkar
2021-04-01 17:21 ` [PATCH v6 6/7] ext4: add proc files to monitor new structures Harshad Shirwadkar
2021-04-01 17:21 ` [PATCH v6 7/7] ext4: make prefetch_block_bitmaps default Harshad Shirwadkar
2021-04-02  5:16   ` Andreas Dilger
2021-04-02 16:46     ` harshad shirwadkar
2021-04-02 17:39       ` Andreas Dilger [this message]
2021-04-09 16:36 ` [PATCH v6 0/7] Block Allocator Improvements Theodore Ts'o

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=4CA7B7B0-BC40-43D3-A34F-38C92EDB38A4@dilger.ca \
    --to=adilger@dilger.ca \
    --cc=harshadshirwadkar@gmail.com \
    --cc=linux-ext4@vger.kernel.org \
    --cc=tytso@mit.edu \
    /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: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).