All of lore.kernel.org
 help / color / mirror / Atom feed
From: Eric Sandeen <sandeen@redhat.com>
To: kreijack@inwind.it
Cc: Goffredo Baroncelli <kreijack@gmail.com>, linux-btrfs@vger.kernel.org
Subject: Re: [PATCH 16/17] btrfs-progs: fix fd leak in cmd_subvol_set_default
Date: Tue, 26 Feb 2013 14:10:04 -0600	[thread overview]
Message-ID: <512D169C.3020100@redhat.com> (raw)
In-Reply-To: <512D031D.20408@gmail.com>

On 2/26/13 12:46 PM, Goffredo Baroncelli wrote:
> Hi Eric,
> 
> On 02/25/2013 11:54 PM, Eric Sandeen wrote:
>> Rearrange cmd_subvol_set_default() slightly so we
>> don't have to close the fd on an error return.
>>
>> While we're at it, fix whitespace & remove magic
>> return values.
>>
>> Signed-off-by: Eric Sandeen <sandeen@redhat.com>
>> ---
>>  cmds-subvolume.c |   17 +++++++++--------
>>  1 files changed, 9 insertions(+), 8 deletions(-)
>>
>> diff --git a/cmds-subvolume.c b/cmds-subvolume.c
>> index 0dfaefe..461eed9 100644
>> --- a/cmds-subvolume.c
>> +++ b/cmds-subvolume.c
>> @@ -712,24 +712,25 @@ static int cmd_subvol_set_default(int argc, char **argv)
>>  	subvolid = argv[1];
>>  	path = argv[2];
>>  
>> +	objectid = (unsigned long long)strtoll(subvolid, NULL, 0);
> 
> Could you replace strtoll() with strtoull() ? Note that:
> 
> strtoull("0xffffffffffffffff",0,0)  == 0xffffffffffffffff
> strtoull("-1",0,0)  == 0xffffffffffffffff
> strtoll("-1",0,0)  == 0xffffffffffffffff
> strtoll("0xffffffffffffffff",0,0)  -> ERANGE

Probably a good idea, I think I had noticed that earlier and
then spaced it.  :(

But I figure one functional change per patch is the way to go;
making this other change would probably be best under its own commit;
one to fix the fd leak, and one to fix this issue?

>> +	if (errno == ERANGE) {
> 
> Pay attention that if strtoull() doesn't encounter a problem errno *is
> not* touched: this check could catch a previous error. I don't know if
> it is an hole in the standard or a bug in the gnu-libc; however I think
> that before strtoXll() we should put 'errno = 0;'.

yeah, ugh.  But this problem existed before, correct?  So I think a
separate fix makes sense, do you agree?  Or have I made something
worse here with this change?

Thanks,
-Eric



>> +		fprintf(stderr, "ERROR: invalid tree id (%s)\n", subvolid);
>> +		return 1;
>> +	}
>> +
>>  	fd = open_file_or_dir(path);
>>  	if (fd < 0) {
>>  		fprintf(stderr, "ERROR: can't access to '%s'\n", path);
>> -		return 12;
>> +		return 1;
>>  	}
>>  
>> -	objectid = (unsigned long long)strtoll(subvolid, NULL, 0);
>> -	if (errno == ERANGE) {
>> -		fprintf(stderr, "ERROR: invalid tree id (%s)\n",subvolid);
>> -		return 30;
>> -	}
>>  	ret = ioctl(fd, BTRFS_IOC_DEFAULT_SUBVOL, &objectid);
>>  	e = errno;
>>  	close(fd);
>> -	if( ret < 0 ){
>> +	if (ret < 0) {
>>  		fprintf(stderr, "ERROR: unable to set a new default subvolume - %s\n",
>>  			strerror(e));
>> -		return 30;
>> +		return 1;
>>  	}
>>  	return 0;
>>  }
> 
> 


  reply	other threads:[~2013-02-26 20:23 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-02-25 22:54 [PATCH 00/17] btrfs-progs: More misc fixes & cleanups Eric Sandeen
2013-02-25 22:54 ` [PATCH 01/17] btrfs-progs: Unify size-parsing Eric Sandeen
2013-02-25 23:26   ` Zach Brown
2013-02-25 23:37     ` Eric Sandeen
2013-02-26  0:26       ` Zach Brown
2013-02-26 18:50   ` Goffredo Baroncelli
2013-02-26 20:17     ` Eric Sandeen
2013-02-26 21:15       ` Goffredo Baroncelli
2013-02-25 22:54 ` [PATCH 02/17] btrfs-progs: fix btrfs_get_subvol cut/paste error Eric Sandeen
2013-02-25 22:54 ` [PATCH 03/17] btrfs-progs: Remove write-only var fdres in cmd_dev_stats() Eric Sandeen
2013-02-25 22:54 ` [PATCH 04/17] btrfs-progs: btrfs_list_get_path_rootid error handling Eric Sandeen
2013-02-25 22:54 ` [PATCH 05/17] btrfs-progs: avoid double-free in __btrfs_map_block Eric Sandeen
2013-02-25 22:54 ` [PATCH 06/17] btrfs-progs: fix open error test in cmd_start_replace Eric Sandeen
2013-02-25 22:54 ` [PATCH 07/17] btrfs-progs: fix close of error fd in scrub cancel Eric Sandeen
2013-02-25 22:54 ` [PATCH 08/17] btrfs-progs: more scrub cancel error handling Eric Sandeen
2013-02-25 22:54 ` [PATCH 09/17] btrfs-progs: free memory before error exit in read_whole_eb Eric Sandeen
2013-02-25 22:54 ` [PATCH 10/17] btrfs-progs: don't call close on error fd Eric Sandeen
2013-02-25 22:54 ` [PATCH 11/17] btrfs-progs: provide positive errno to strerror in cmd_restore Eric Sandeen
2013-02-25 22:54 ` [PATCH 12/17] btrfs-progs: free allocated di_args in cmd_start_replace Eric Sandeen
2013-02-25 22:54 ` [PATCH 13/17] btrfs-progs: close fd on cmd_subvol_get_default return Eric Sandeen
2013-02-25 22:54 ` [PATCH 14/17] btrfs-progs: fix mem leak in resolve_root Eric Sandeen
2013-02-26  0:36   ` Shilong Wang
2013-02-26  4:36     ` Eric Sandeen
2013-02-27 13:03       ` David Sterba
2013-02-27 13:12         ` Shilong Wang
2013-02-25 22:54 ` [PATCH 15/17] btrfs-progs: Tidy up resolve_root Eric Sandeen
2013-02-25 22:54 ` [PATCH 16/17] btrfs-progs: fix fd leak in cmd_subvol_set_default Eric Sandeen
2013-02-26 18:46   ` Goffredo Baroncelli
2013-02-26 20:10     ` Eric Sandeen [this message]
2013-02-26 21:04       ` Goffredo Baroncelli
2013-02-27 12:38         ` David Sterba
2013-02-25 22:54 ` [PATCH 17/17] btrfs-progs: replace strtok_r with strsep Eric Sandeen
2013-02-26 18:47   ` Goffredo Baroncelli
2013-02-26 20:13     ` Eric Sandeen
2013-02-26 20:20     ` [PATCH 17/17 V2] " Eric Sandeen
2013-02-26 20:40       ` Ilya Dryomov
2013-02-26 20:46         ` Eric Sandeen
2013-02-26 21:07           ` Ilya Dryomov
2013-02-26 21:50             ` [PATCH 17/17 V3] btrfs-progs: initialize save_ptr prior to strtok_r Eric Sandeen
2013-02-27 13:54 ` [PATCH 00/17] btrfs-progs: More misc fixes & cleanups David Sterba

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=512D169C.3020100@redhat.com \
    --to=sandeen@redhat.com \
    --cc=kreijack@gmail.com \
    --cc=kreijack@inwind.it \
    --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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.