| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <47c7d469-c807-760d-7e28-ae8108a46af5@huawei.com>
Date: Fri, 31 Aug 2018 11:29:03 +0800
From: Chao Yu <yuchao0@...wei.com>
To: Gao Xiang <hsiangkao@....com>,
Pavel Zemlyanoy <zemlyanoy@...ras.ru>
CC: <ldv-project@...uxtesting.org>, <devel@...verdev.osuosl.org>,
<linux-erofs@...ts.ozlabs.org>, <linux-kernel@...r.kernel.org>,
"Greg Kroah-Hartman" <gregkh@...uxfoundation.org>
Subject: Re: [PATCH 2/6] staging: erofs: formatting fix to NULL comparison
On 2018/8/31 0:32, Gao Xiang via Linux-erofs wrote:
>
>
> On 2018/8/31 0:09, Gao Xiang via Linux-erofs wrote:
>> Hi Pavel,
>>
>> On 2018/8/30 23:13, Pavel Zemlyanoy wrote:
>>> This patch does not change the logic, it only
>>> corrects the formatting and checkpatch checks by
>>> to NULL comparison.
>>>
>>> The patch fixes 5 checks of type:
>>> "Comparison to NULL could be written".
>>>
>>> Signed-off-by: Pavel Zemlyanoy <zemlyanoy@...ras.ru>
>>
>> Sorry about the late reply. I am on vacation now.
>>
>> Personally, I use "== NULL" or "!= NULL" on purpose, since it is more
>> emphasized than the checkpatch.pl suggested way, and I tend to use the nullptr explicitly.
>>
>> BTW, It seems other filesystems still use "== NULL" or "!= NULL" explicitly, too, eg:
>> xfs, ext4, ext2, ocfs2, etc... You could 'grep' in the fs directory...
>>
>> Other commits look good for me at glance.
>>
>
> p.s. I personally tend to drop this patch. :(
>
> Here are bunch of 'NULL comparison' usages in xfs, eg.
>
> ...
> ./xfs_qm_syscalls.c:218: if (ino == NULLFSINO) ./xfs_qm_syscalls.c:747: ASSERT(ip->i_udquot == NULL); ./xfs_qm_syscalls.c:748: ASSERT(ip->i_gdquot == NULL); ./xfs_qm_syscalls.c:749: ASSERT(ip->i_pdquot == NULL); ./xfs_quota.h:27: ((XFS_IS_UQUOTA_ON(mp) && (ip)->i_udquot == NULL) || \ ./xfs_quota.h:28: (XFS_IS_GQUOTA_ON(mp) && (ip)->i_gdquot == NULL) || \ ./xfs_quota.h:29: (XFS_IS_PQUOTA_ON(mp) && (ip)->i_pdquot == NULL)) ./xfs_quotaops.c:31: if (!ip && ino == NULLFSINO) ./xfs_reflink.c:213: if (fbno == NULLAGBLOCK) { ./xfs_reflink.c:652: if (count == NULLFILEOFF) ./xfs_reflink.c:1462: if (rbno == NULLAGBLOCK) ./xfs_rtalloc.c:836: if (bp == NULL) { ./xfs_rtalloc.c:899: if (mp->m_rtdev_targp == NULL || mp->m_rbmip == NULL || ./xfs_rtalloc.c:1165: if (mp->m_rtdev_targp == NULL) { ./xfs_rtalloc.c:1209: if (sbp->sb_rbmino == NULLFSINO) ./xfs_trans.c:185: ASSERT(tp->t_ticket == NULL); ./xfs_trans_ail.c:59: (prev_lsn == NULLCOMMITLSN || XFS_LSN_CMP(prev_lsn, lsn) <= 0) && ./xfs_trans_ail.c:60: (next_lsn == NULLCOMMITLSN || XFS_LSN_CMP(next_lsn, lsn) >= 0)) ./xfs_trans_ail.c:65: ASSERT(prev_lsn == NULLCOMMITLSN || XFS_LSN_CMP(prev_lsn, lsn) <= 0); ./xfs_trans_ail.c:66: ASSERT(next_lsn == NULLCOMMITLSN || XFS_LSN_CMP(next_lsn, lsn) >= 0); ./xfs_trans_ail.c:475: if (lip == NULL) ./xfs_trans_buf.c:70: ASSERT(bp->b_transp == NULL); ./xfs_trans_buf.c:155: if (bp == NULL) { ./xfs_trans_buf.c:187: if (tp == NULL) ./xfs_trans_buf.c:207: if (bp == NULL) ./xfs_trans_buf.c:350: if (tp == NULL) { ./xfs_trans_buf.c:351: ASSERT(bp->b_transp == NULL); ./xfs_trans_buf.c:495: ASSERT(bp->b_iodone == NULL || ./xfs_trans_dquot.c:103: if (oqa[i].qt_dquot == NULL) ./xfs_trans_dquot.c:149: if (tp->t_dqinfo == NULL) ./xfs_trans_dquot.c:178: if (qa[i].qt_dquot == NULL || ./xfs_trans_dquot.c:205: if (tp->t_dqinfo == NULL) ./xfs_trans_dquot.c:213: if (qtrx->qt_dquot == NULL) ./xfs_trans_dquot.c:295: if (q[1].qt_dquot == NULL) { ./xfs_trans_dquot.c:332: if (qa[0].qt_dquot == NULL) ./xfs_trans_dquot.c:346: if ((dqp = qtrx->qt_dquot) == NULL) ./xfs_trans_dquot.c:519: if ((dqp = qtrx->qt_dquot) == NULL) ./xfs_trans_dquot.c:757: if (tp && tp->t_dqinfo == NULL) ./xfs_trans_inode.c:36: if (ip->i_itemp == NULL)
> ...
>
> And in ext4, eg.
> ...
> ./mballoc.c:2892: if (ext4_free_data_cachep == NULL) {
> ./mballoc.c:3046: BUG_ON(lg == NULL);
> ./mballoc.c:3286: if (pa == NULL) {
> ./mballoc.c:3377: if (cpa == NULL) {
> ./mballoc.c:3447: if (lg == NULL)
> ./mballoc.c:3635: if (pa == NULL)
> ./mballoc.c:3727: BUG_ON(ext4_pspace_cachep == NULL);
> ./mballoc.c:3729: if (pa == NULL)
> ./mballoc.c:3756: BUG_ON(lg == NULL);
> ./mballoc.c:4637: BUG_ON(e4b->bd_bitmap_page == NULL);
> ./mballoc.c:4638: BUG_ON(e4b->bd_buddy_page == NULL);
> ./move_extent.c:34: if (path[ext_depth(inode)].p_ext == NULL) {
> ./namei.c:874: if (frames[0].bh == NULL)
> ./namei.c:879: if (frames[i].bh == NULL)
> ./namei.c:1432: if ((bh = bh_use[ra_ptr++]) == NULL)
> ./page-io.c:38: if (io_end_cachep == NULL)
> ./page-io.c:398: if (io->io_bio == NULL) {
> ./readpage.c:242: if (bio == NULL) {
> ./resize.c:200: if (flex_gd == NULL)
> ./resize.c:210: if (flex_gd->groups == NULL)
> ./resize.c:215: if (flex_gd->bg_flags == NULL)
> ./resize.c:265: BUG_ON(flex_gd->count == 0 || group_data == NULL);
> ./resize.c:1345: BUG_ON(flex_gd->count == 0 || group_data == NULL);
> ./resize.c:2012: if (flex_gd == NULL) {
> ./super.c:365: return bdi->dev == NULL;
> ./super.c:596: if (errno == -EROFS && journal_current_handle() == NULL && sb_rdonly(sb))
> ./super.c:1086: if (ext4_inode_cachep == NULL)
> ./super.c:4050: if (sbi->s_group_desc == NULL) {
> ./super.c:4617: if (bdev == NULL)
> ./super.c:5288: if (sbi->s_journal == NULL && !(old_sb_flags & SB_RDONLY)) {
> ./xattr.c:286: if (name == NULL)
> ./xattr.c:1905: if (s->base == NULL)
> ./xattr.c:1948: if (s->base == NULL)
> ./xattr.c:2665: if (entry == NULL) {
> ./xattr.c:2666: if (small_entry == NULL)
> ./xattr.c:2804: if (*ea_inode_array == NULL) {
> ./xattr.c:2813: if (*ea_inode_array == NULL)
> ./xattr.c:2826: if (new_array == NULL)
> ./xattr.c:2947: if (ea_inode_array == NULL)
> ...
>
> Anyway, I'd like to listen the Greg's and Chao's ideas.
Hi Xiang,
I'm not against this change which follows checkpatch's rule, since I think this
can help to unify coding style in different modules of Linux. Maybe cleanup in
other filesystem is needed as well.
Also, personally speaking, I'm used to judge point/variable is valid or not by
using 'if {,!}{value,pointer}', it will be easy for me to know which case next
branch belongs to, just my habit being trained during f2fs development... :P
Thanks,
>
> Thanks,
>
>> Thanks,
>> Gao Xiang
>>
Powered by blists - more mailing lists