From: Chao Yu <chao@kernel.org>
To: Yangtao Li <frank.li@vivo.com>, jaegeuk@kernel.org
Cc: linux-kernel@vger.kernel.org, linux-f2fs-devel@lists.sourceforge.net
Subject: Re: [f2fs-dev] [PATCH 4/4] f2fs: move ipu_policy definitions to separated file
Date: Sat, 28 Jan 2023 17:54:14 +0800 [thread overview]
Message-ID: <989046e3-b288-d966-1d21-2fccabdcb72d@kernel.org> (raw)
In-Reply-To: <20230120134029.69200-4-frank.li@vivo.com>
On 2023/1/20 21:40, Yangtao Li wrote:
> User can control f2fs ipu policy through /sys/fs/f2fs/<disk>/ipu_policy,
> export related definitions and introduce new macro to make it more
> convenient for users to use.
>
> Signed-off-by: Yangtao Li <frank.li@vivo.com>
> ---
> Documentation/ABI/testing/sysfs-fs-f2fs | 25 ++++++------
> fs/f2fs/segment.c | 2 +-
> fs/f2fs/segment.h | 37 ++---------------
> fs/f2fs/super.c | 2 +-
> fs/f2fs/sysfs.c | 2 +-
> include/uapi/linux/f2fs.h | 54 +++++++++++++++++++++++++
> 6 files changed, 73 insertions(+), 49 deletions(-)
>
> diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/testing/sysfs-fs-f2fs
> index 64b15a28fe30..d05ad2bc56d1 100644
> --- a/Documentation/ABI/testing/sysfs-fs-f2fs
> +++ b/Documentation/ABI/testing/sysfs-fs-f2fs
> @@ -49,18 +49,19 @@ Contact: "Jaegeuk Kim" <jaegeuk.kim@samsung.com>
> Description: Controls the in-place-update policy.
> updates in f2fs. User can set:
>
> - ==== =================
> - 0x01 F2FS_IPU_FORCE
> - 0x02 F2FS_IPU_SSR
> - 0x04 F2FS_IPU_UTIL
> - 0x08 F2FS_IPU_SSR_UTIL
> - 0x10 F2FS_IPU_FSYNC
> - 0x20 F2FS_IPU_ASYNC
> - 0x40 F2FS_IPU_NOCACHE
> - 0x80 F2FS_IPU_HONOR_OPU_WRITE
> - ==== =================
> -
> - Refer segment.h for details.
> + ==== ============================
> + 0x00 SET_F2FS_IPU_DISABLE
> + 0x01 SET_F2FS_IPU_FORCE
> + 0x02 SET_F2FS_IPU_SSR
> + 0x04 SET_F2FS_IPU_UTIL
> + 0x08 SET_F2FS_IPU_SSR_UTIL
> + 0x10 SET_F2FS_IPU_FSYNC
> + 0x20 SET_F2FS_IPU_ASYNC
> + 0x40 SET_F2FS_IPU_NOCACHE
> + 0x80 SET_F2FS_IPU_HONOR_OPU_WRITE
> + ==== ============================
> +
> + Refer include/uapi/linux/f2fs.h for details.
>
> What: /sys/fs/f2fs/<disk>/min_ipu_util
> Date: November 2013
> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> index a007f5084e84..c56f5a80d3a6 100644
> --- a/fs/f2fs/segment.c
> +++ b/fs/f2fs/segment.c
> @@ -5138,7 +5138,7 @@ int f2fs_build_segment_manager(struct f2fs_sb_info *sbi)
> sm_info->rec_prefree_segments = DEF_MAX_RECLAIM_PREFREE_SEGMENTS;
>
> if (!f2fs_lfs_mode(sbi))
> - sm_info->ipu_policy = BIT(F2FS_IPU_FSYNC);
> + sm_info->ipu_policy = SET_F2FS_IPU_FSYNC;
> sm_info->min_ipu_util = DEF_MIN_IPU_UTIL;
> sm_info->min_fsync_blocks = DEF_MIN_FSYNC_BLOCKS;
> sm_info->min_seq_blocks = sbi->blocks_per_seg;
> diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h
> index 50c63155ad56..c064ddffe91b 100644
> --- a/fs/f2fs/segment.h
> +++ b/fs/f2fs/segment.h
> @@ -7,6 +7,7 @@
> */
> #include <linux/blkdev.h>
> #include <linux/backing-dev.h>
> +#include <uapi/linux/f2fs.h>
>
> /* constant macro */
> #define NULL_SEGNO ((unsigned int)(~0))
> @@ -646,51 +647,19 @@ static inline int utilization(struct f2fs_sb_info *sbi)
> sbi->user_block_count);
> }
>
> -/*
> - * Sometimes f2fs may be better to drop out-of-place update policy.
> - * And, users can control the policy through sysfs entries.
> - * There are five policies with triggering conditions as follows.
> - * F2FS_IPU_FORCE - all the time,
> - * F2FS_IPU_SSR - if SSR mode is activated,
> - * F2FS_IPU_UTIL - if FS utilization is over threashold,
> - * F2FS_IPU_SSR_UTIL - if SSR mode is activated and FS utilization is over
> - * threashold,
> - * F2FS_IPU_FSYNC - activated in fsync path only for high performance flash
> - * storages. IPU will be triggered only if the # of dirty
> - * pages over min_fsync_blocks. (=default option)
> - * F2FS_IPU_ASYNC - do IPU given by asynchronous write requests.
> - * F2FS_IPU_NOCACHE - disable IPU bio cache.
> - * F2FS_IPU_HONOR_OPU_WRITE - use OPU write prior to IPU write if inode has
> - * FI_OPU_WRITE flag.
> - * F2FS_IPU_DISABLE - disable IPU. (=default option in LFS mode)
> - */
IMO, we'd better to move these comments into sysfs-fs-f2fs as ipu_policy
interface's documents.
> #define DEF_MIN_IPU_UTIL 70
> #define DEF_MIN_FSYNC_BLOCKS 8
> #define DEF_MIN_HOT_BLOCKS 16
>
> #define SMALL_VOLUME_SEGMENTS (16 * 512) /* 16GB */
>
> -#define F2FS_IPU_DISABLE 0
> -
> -/* Modification on enum should be synchronized with ipu_mode_names array */
> -enum {
> - F2FS_IPU_FORCE,
> - F2FS_IPU_SSR,
> - F2FS_IPU_UTIL,
> - F2FS_IPU_SSR_UTIL,
> - F2FS_IPU_FSYNC,
> - F2FS_IPU_ASYNC,
> - F2FS_IPU_NOCACHE,
> - F2FS_IPU_HONOR_OPU_WRITE,
> - F2FS_IPU_MAX,
> -};
> -
> #define F2FS_IPU_POLICY(name) \
> static inline int IS_##name(struct f2fs_sb_info *sbi) \
> { \
> - return SM_I(sbi)->ipu_policy & BIT(name); \
> + return TEST_##name(SM_I(sbi)->ipu_policy); \
> }
>
> +F2FS_IPU_POLICY(F2FS_IPU_DISABLE);
> F2FS_IPU_POLICY(F2FS_IPU_FORCE);
> F2FS_IPU_POLICY(F2FS_IPU_SSR);
> F2FS_IPU_POLICY(F2FS_IPU_UTIL);
> diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
> index ebc76683f05d..b19085fa335a 100644
> --- a/fs/f2fs/super.c
> +++ b/fs/f2fs/super.c
> @@ -4085,7 +4085,7 @@ static void f2fs_tuning_parameters(struct f2fs_sb_info *sbi)
> if (f2fs_block_unit_discard(sbi))
> SM_I(sbi)->dcc_info->discard_granularity = MIN_DISCARD_GRANULARITY;
> if (F2FS_OPTION(sbi).fs_mode != FS_MODE_LFS)
> - SM_I(sbi)->ipu_policy = BIT(F2FS_IPU_FORCE) | BIT(F2FS_IPU_HONOR_OPU_WRITE);
> + SM_I(sbi)->ipu_policy = SET_F2FS_IPU_FORCE | SET_F2FS_IPU_HONOR_OPU_WRITE;
> }
>
> sbi->readdir_ra = true;
> diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c
> index aaf5d8650518..375ace5f330d 100644
> --- a/fs/f2fs/sysfs.c
> +++ b/fs/f2fs/sysfs.c
> @@ -165,7 +165,7 @@ static ssize_t ipu_mode_show(struct f2fs_attr *a,
> {
> int len = 0, i = 0;
>
> - if (SM_I(sbi)->ipu_policy == F2FS_IPU_DISABLE)
> + if (IS_F2FS_IPU_DISABLE(sbi))
> return sysfs_emit(buf, "DISABLE\n");
>
> for_each_set_bit(i, &SM_I(sbi)->ipu_policy, F2FS_IPU_MAX)
> diff --git a/include/uapi/linux/f2fs.h b/include/uapi/linux/f2fs.h
> index 955d440be104..1b672fb6af91 100644
> --- a/include/uapi/linux/f2fs.h
> +++ b/include/uapi/linux/f2fs.h
> @@ -5,6 +5,60 @@
> #include <linux/types.h>
> #include <linux/ioctl.h>
>
> +/*
> + * For /sys/fs/f2fs/<disk>/ipu_policy entry
> + *
> + * Sometimes f2fs may be better to drop out-of-place update policy.
> + * And, users can control the policy through sysfs entries.
> + * There are five policies with triggering conditions as follows.
> + * F2FS_IPU_FORCE - all the time,
> + * F2FS_IPU_SSR - if SSR mode is activated,
> + * F2FS_IPU_UTIL - if FS utilization is over threashold,
> + * F2FS_IPU_SSR_UTIL - if SSR mode is activated and FS utilization is over
> + * threashold,
> + * F2FS_IPU_FSYNC - activated in fsync path only for high performance flash
> + * storages. IPU will be triggered only if the # of dirty
> + * pages over min_fsync_blocks. (=default option)
> + * F2FS_IPU_ASYNC - do IPU given by asynchronous write requests.
> + * F2FS_IPU_NOCACHE - disable IPU bio cache.
> + * F2FS_IPU_HONOR_OPU_WRITE - use OPU write prior to IPU write if inode has
> + * FI_OPU_WRITE flag.
> + * F2FS_IPU_DISABLE - disable IPU. (=default option in LFS mode)
> + */
> +#define F2FS_IPU_DISABLE 0
> +/* Modification on enum should be synchronized with ipu_mode_names array */
> +enum {
> + F2FS_IPU_FORCE,
> + F2FS_IPU_SSR,
> + F2FS_IPU_UTIL,
> + F2FS_IPU_SSR_UTIL,
> + F2FS_IPU_FSYNC,
> + F2FS_IPU_ASYNC,
> + F2FS_IPU_NOCACHE,
> + F2FS_IPU_HONOR_OPU_WRITE,
> + F2FS_IPU_MAX,
> +};
> +
> +#define SET_F2FS_IPU_DISABLE F2FS_IPU_DISABLE
> +#define SET_F2FS_IPU_FORCE BIT(F2FS_IPU_FORCE)
> +#define SET_F2FS_IPU_SSR BIT(F2FS_IPU_SSR)
> +#define SET_F2FS_IPU_UTIL BIT(F2FS_IPU_UTIL)
> +#define SET_F2FS_IPU_SSR_UTIL BIT(F2FS_IPU_SSR_UTIL)
> +#define SET_F2FS_IPU_FSYNC BIT(F2FS_IPU_FSYNC)
> +#define SET_F2FS_IPU_ASYNC BIT(F2FS_IPU_ASYNC)
> +#define SET_F2FS_IPU_NOCACHE BIT(F2FS_IPU_NOCACHE)
> +#define SET_F2FS_IPU_HONOR_OPU_WRITE BIT(F2FS_IPU_HONOR_OPU_WRITE)
> +
> +#define TEST_F2FS_IPU_DISABLE(val) ((val) == F2FS_IPU_DISABLE)
> +#define TEST_F2FS_IPU_FORCE(val) ((val) & BIT(F2FS_IPU_FORCE))
> +#define TEST_F2FS_IPU_SSR(val) ((val) & BIT(F2FS_IPU_SSR))
> +#define TEST_F2FS_IPU_UTIL(val) ((val) & BIT(F2FS_IPU_UTIL))
> +#define TEST_F2FS_IPU_SSR_UTIL(val) ((val) & BIT(F2FS_IPU_SSR_UTIL))
> +#define TEST_F2FS_IPU_FSYNC(val) ((val) & BIT(F2FS_IPU_FSYNC))
> +#define TEST_F2FS_IPU_ASYNC(val) ((val) & BIT(F2FS_IPU_ASYNC))
> +#define TEST_F2FS_IPU_NOCACHE(val) ((val) & BIT(F2FS_IPU_NOCACHE))
> +#define TEST_F2FS_IPU_HONOR_OPU_WRITE(val) ((val) & BIT(F2FS_IPU_HONOR_OPU_WRITE))
I don't see any strong reason to add these sysfs related definition to uapi,
according to other filesystems' implementation, I guess describing the usage
of ipu_policy clearly is enough.
Thanks,
> +
> /*
> * f2fs-specific ioctl commands
> */
_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
next prev parent reply other threads:[~2023-01-28 9:54 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-20 13:40 [f2fs-dev] [PATCH 1/4] f2fs: fix to set ipu policy Yangtao Li via Linux-f2fs-devel
2023-01-20 13:40 ` [f2fs-dev] [PATCH 2/4] f2fs: add sanity check for ipu_policy Yangtao Li via Linux-f2fs-devel
2023-01-28 9:27 ` Chao Yu
2023-01-20 13:40 ` [f2fs-dev] [PATCH 3/4] f2fs: introduce ipu_mode sysfs node Yangtao Li via Linux-f2fs-devel
2023-01-28 9:37 ` Chao Yu
2023-01-20 13:40 ` [f2fs-dev] [PATCH 4/4] f2fs: move ipu_policy definitions to separated file Yangtao Li via Linux-f2fs-devel
2023-01-28 9:54 ` Chao Yu [this message]
2023-01-28 9:24 ` [f2fs-dev] [PATCH 1/4] f2fs: fix to set ipu policy Chao Yu
2023-01-30 22:29 ` Jaegeuk Kim
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=989046e3-b288-d966-1d21-2fccabdcb72d@kernel.org \
--to=chao@kernel.org \
--cc=frank.li@vivo.com \
--cc=jaegeuk@kernel.org \
--cc=linux-f2fs-devel@lists.sourceforge.net \
--cc=linux-kernel@vger.kernel.org \
/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).