From: Jan Kara <jack@suse.cz>
To: Marco Stornelli <marco.stornelli@gmail.com>
Cc: Linux FS Devel <linux-fsdevel@vger.kernel.org>,
Linux Kernel <linux-kernel@vger.kernel.org>,
Al Viro <viro@ZenIV.linux.org.uk>, Jan Kara <jack@suse.cz>
Subject: Re: [PATCH 1/2] fsfreeze: add new internal __sb_start_write_wait
Date: Mon, 8 Apr 2013 19:55:17 +0200 [thread overview]
Message-ID: <20130408175517.GA25878@quack.suse.cz> (raw)
In-Reply-To: <5162EEC5.4000204@gmail.com>
On Mon 08-04-13 18:22:29, Marco Stornelli wrote:
> Added a new internal function __sb_start_write_wait. It must be called
> when we want wait for freeze events. We can wait in killable or
> uninterruptible state. The old __sb_start_write now it's used only
> when we don't want to wait. In addition, a new wrapper sb_start_write_killable
> is added.
>
> Signed-off-by: Marco Stornelli <marco.stornelli@gmail.com>
> ---
> fs/super.c | 47 ++++++++++++++++++++++++++++++++++++++++-------
> include/linux/fs.h | 23 ++++++++++++++++++-----
> 2 files changed, 58 insertions(+), 12 deletions(-)
>
> diff --git a/fs/super.c b/fs/super.c
> index 7465d43..cb0149b 100644
> --- a/fs/super.c
> +++ b/fs/super.c
> @@ -1190,15 +1190,11 @@ static void acquire_freeze_lock(struct super_block *sb, int level, bool trylock,
> * This is an internal function, please use sb_start_{write,pagefault,intwrite}
> * instead.
> */
> -int __sb_start_write(struct super_block *sb, int level, bool wait)
> +int __sb_start_write(struct super_block *sb, int level)
> {
> retry:
> - if (unlikely(sb->s_writers.frozen >= level)) {
> - if (!wait)
> - return 0;
> - wait_event(sb->s_writers.wait_unfrozen,
> - sb->s_writers.frozen < level);
> - }
> + if (unlikely(sb->s_writers.frozen >= level))
> + return 0;
>
> #ifdef CONFIG_LOCKDEP
> acquire_freeze_lock(sb, level, !wait, _RET_IP_);
> @@ -1217,6 +1213,43 @@ retry:
> }
> EXPORT_SYMBOL(__sb_start_write);
>
> +/*
> + * This is an internal function, please use sb_start_{write,pagefault,intwrite}
> + * instead. It returns zero if no error occured, the error code otherwise.
> + */
> +int __sb_start_write_wait(struct super_block *sb, int level, bool wait_killable)
> +{
> + int ret = 0;
> +
> +retry:
> + if (unlikely(sb->s_writers.frozen >= level)) {
> + if (wait_killable)
> + ret = wait_event_killable(sb->s_writers.wait_unfrozen,
> + sb->s_writers.frozen < level);
> + if (ret)
> + return -EINTR;
> + else
> + wait_event(sb->s_writers.wait_unfrozen,
> + sb->s_writers.frozen < level);
> + }
> +
> +#ifdef CONFIG_LOCKDEP
> + acquire_freeze_lock(sb, level, !wait, _RET_IP_);
> +#endif
> + percpu_counter_inc(&sb->s_writers.counter[level-1]);
> + /*
> + * Make sure counter is updated before we check for frozen.
> + * freeze_super() first sets frozen and then checks the counter.
> + */
> + smp_mb();
> + if (unlikely(sb->s_writers.frozen >= level)) {
> + __sb_end_write(sb, level);
> + goto retry;
> + }
> + return ret;
> +}
> +EXPORT_SYMBOL(__sb_start_write_wait);
Why do you duplicate this function? I'd prefer if you kept single
sb_start_write() core function which would conditionally wait (maybe wait
argument could have values NOWAIT, WAIT_KILLABLE, WAIT).
Honza
> +
> /**
> * sb_wait_write - wait until all writers to given file system finish
> * @sb: the super for which we wait
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index 2c28271..6428dbd 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -1333,7 +1333,8 @@ extern struct timespec current_fs_time(struct super_block *sb);
> */
>
> void __sb_end_write(struct super_block *sb, int level);
> -int __sb_start_write(struct super_block *sb, int level, bool wait);
> +int __sb_start_write(struct super_block *sb, int level);
> +int __sb_start_write_wait(struct super_block *sb, int level, bool wait_killable);
>
> /**
> * sb_end_write - drop write access to a superblock
> @@ -1392,12 +1393,24 @@ static inline void sb_end_intwrite(struct super_block *sb)
> */
> static inline void sb_start_write(struct super_block *sb)
> {
> - __sb_start_write(sb, SB_FREEZE_WRITE, true);
> + __sb_start_write_wait(sb, SB_FREEZE_WRITE, false);
> +}
> +
> +/**
> + * sb_start_write_killable - get write access to a superblock
> + * @sb: the super we write to
> + *
> + * Same semantic of sb_start_write, but we are going to wait in a killable
> + * state instead of waiting in uninterruptible state.
> + */
> +static inline int __must_check sb_start_write_killable(struct super_block *sb)
> +{
> + return __sb_start_write_wait(sb, SB_FREEZE_WRITE, true);
> }
>
> static inline int sb_start_write_trylock(struct super_block *sb)
> {
> - return __sb_start_write(sb, SB_FREEZE_WRITE, false);
> + return __sb_start_write(sb, SB_FREEZE_WRITE);
> }
>
> /**
> @@ -1421,7 +1434,7 @@ static inline int sb_start_write_trylock(struct super_block *sb)
> */
> static inline void sb_start_pagefault(struct super_block *sb)
> {
> - __sb_start_write(sb, SB_FREEZE_PAGEFAULT, true);
> + __sb_start_write_wait(sb, SB_FREEZE_PAGEFAULT, false);
> }
>
> /*
> @@ -1439,7 +1452,7 @@ static inline void sb_start_pagefault(struct super_block *sb)
> */
> static inline void sb_start_intwrite(struct super_block *sb)
> {
> - __sb_start_write(sb, SB_FREEZE_FS, true);
> + __sb_start_write_wait(sb, SB_FREEZE_FS, false);
> }
>
>
> --
> 1.7.3.4
--
Jan Kara <jack@suse.cz>
SUSE Labs, CR
prev parent reply other threads:[~2013-04-08 17:55 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-04-08 16:22 [PATCH 1/2] fsfreeze: add new internal __sb_start_write_wait Marco Stornelli
2013-04-08 17:55 ` Jan Kara [this message]
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=20130408175517.GA25878@quack.suse.cz \
--to=jack@suse.cz \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=marco.stornelli@gmail.com \
--cc=viro@ZenIV.linux.org.uk \
/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).