From: Nikolay Borisov <nborisov@suse.com>
To: fdmanana@gmail.com
Cc: dsterba@suse.cz, linux-btrfs <linux-btrfs@vger.kernel.org>
Subject: Re: [PATCH v2] btrfs: Properly handle backref_in_log retval
Date: Thu, 26 Sep 2019 13:39:58 +0300 [thread overview]
Message-ID: <93c09683-2a67-a6e5-8853-9092912d48f7@suse.com> (raw)
In-Reply-To: <CAL3q7H6mZTNN2NuZ8dudR=F=MHVsjbyK6=3ELCOhnQJb_AFhWg@mail.gmail.com>
On 26.09.19 г. 12:31 ч., Filipe Manana wrote:
> On Thu, Sep 26, 2019 at 10:27 AM Nikolay Borisov <nborisov@suse.com> wrote:
>>
>> This function can return a negative error value if btrfs_search_slot
>> errors for whatever reason or if btrfs_alloc_path runs out of memory.
>> This is currently problemattic because backref_in_log is treated by its
>> callers as if it returns boolean.
>>
>> Fix this by adding proper error handling in callers. That also enables
>> the function to return the direct error code from btrfs_search_slot.
>>
>> Signed-off-by: Nikolay Borisov <nborisov@suse.com>
>> ---
>>
>> v2:
>> * Return 0 from backref_in_log in case btrfs_search_slot returns 1 (meaning it
>> didn't find appropriate item in the btree).
>>
>> fs/btrfs/tree-log.c | 32 +++++++++++++++++++++-----------
>> 1 file changed, 21 insertions(+), 11 deletions(-)
>>
>> diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
>> index 7cac09a6f007..7332f7a00790 100644
>> --- a/fs/btrfs/tree-log.c
>> +++ b/fs/btrfs/tree-log.c
>> @@ -952,7 +952,9 @@ static noinline int backref_in_log(struct btrfs_root *log,
>> return -ENOMEM;
>>
>> ret = btrfs_search_slot(NULL, log, key, path, 0, 0);
>> - if (ret != 0) {
>> + if (ret < 0) {
>> + goto out;
>> + } else if (ret == 1) {
>> ret = 0;
>> goto out;
>> }
>> @@ -1026,10 +1028,13 @@ static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
>> (unsigned long)(victim_ref + 1),
>> victim_name_len);
>>
>> - if (!backref_in_log(log_root, &search_key,
>> - parent_objectid,
>> - victim_name,
>> - victim_name_len)) {
>> + ret = backref_in_log(log_root, &search_key,
>> + parent_objectid, victim_name,
>> + victim_name_len);
>> + if (ret < 0) {
>> + kfree(victim_name);
>> + return ret;
>> + } else if (!ret) {
>> inc_nlink(&inode->vfs_inode);
>> btrfs_release_path(path);
>>
>> @@ -1091,10 +1096,12 @@ static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
>> search_key.offset = btrfs_extref_hash(parent_objectid,
>> victim_name,
>> victim_name_len);
>> - ret = 0;
>> - if (!backref_in_log(log_root, &search_key,
>> - parent_objectid, victim_name,
>> - victim_name_len)) {
>> + ret = backref_in_log(log_root, &search_key,
>> + parent_objectid, victim_name,
>> + victim_name_len);
>> + if (ret < 0) {
>> + return ret;
>> + } else if (!ret) {
>> ret = -ENOENT;
>> victim_parent = read_one_inode(root,
>> parent_objectid);
>> @@ -1869,16 +1876,19 @@ static bool name_in_log_ref(struct btrfs_root *log_root,
>> const u64 dirid, const u64 ino)
>> {
>> struct btrfs_key search_key;
>> + int ret;
>>
>> search_key.objectid = ino;
>> search_key.type = BTRFS_INODE_REF_KEY;
>> search_key.offset = dirid;
>> - if (backref_in_log(log_root, &search_key, dirid, name, name_len))
>> + ret = backref_in_log(log_root, &search_key, dirid, name, name_len);
>> + if (ret == 1)
>> return true;
>>
>> search_key.type = BTRFS_INODE_EXTREF_KEY;
>> search_key.offset = btrfs_extref_hash(dirid, name, name_len);
>> - if (backref_in_log(log_root, &search_key, dirid, name, name_len))
>> + ret = backref_in_log(log_root, &search_key, dirid, name, name_len);
>> + if (ret == 1)
>> return true;
>
> This function also needs to be able to return errors and its caller
> check for errors.
Yes but this is for a follow up patch. The current patch does not make
the code any more broken than it currently is.
>
> Thanks.
>
>>
>> return false;
>> --
>> 2.17.1
>>
>
>
next prev parent reply other threads:[~2019-09-26 10:40 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-30 14:44 [PATCH 0/3] Cleanup backref_in_log and its callers Nikolay Borisov
2019-08-30 14:44 ` [PATCH 1/3] btrfs: Don't opencode btrfs_find_name_in_backref in backref_in_log Nikolay Borisov
2019-09-24 17:09 ` David Sterba
2019-08-30 14:44 ` [PATCH 2/3] btrfs: Properly handle backref_in_log retval Nikolay Borisov
2019-09-24 17:09 ` David Sterba
2019-09-25 11:03 ` [PATCH v2] " Nikolay Borisov
2019-09-26 9:31 ` Filipe Manana
2019-09-26 10:39 ` Nikolay Borisov [this message]
2019-10-03 12:55 ` David Sterba
2019-10-03 13:07 ` David Sterba
2019-08-30 14:44 ` [PATCH 3/3] btrfs: Open-code name_in_log_ref in replay_one_name Nikolay Borisov
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=93c09683-2a67-a6e5-8853-9092912d48f7@suse.com \
--to=nborisov@suse.com \
--cc=dsterba@suse.cz \
--cc=fdmanana@gmail.com \
--cc=linux-btrfs@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