All of lore.kernel.org
 help / color / mirror / Atom feed
* [bvanassche:block-small-segments 10/13] block/blk.h:343 bio_may_exceed_limits() error: we previously assumed 'bio->bi_io_vec' could be null (see line 340)
@ 2023-01-30 11:36 ` Dan Carpenter
  0 siblings, 0 replies; 3+ messages in thread
From: kernel test robot @ 2023-01-29 12:21 UTC (permalink / raw)
  To: oe-kbuild; +Cc: lkp, Dan Carpenter

BCC: lkp@intel.com
CC: oe-kbuild-all@lists.linux.dev
TO: Bart Van Assche <bvanassche@acm.org>

tree:   https://github.com/bvanassche/linux block-small-segments
head:   c776ad1c2e1e0afaefb3ef5f1710dc3886f5348f
commit: b91c4a42d14ad8bfd74387c10c8c21db9fb2b998 [10/13] block: Add support for filesystem requests and small segments
:::::: branch date: 35 hours ago
:::::: commit date: 35 hours ago
config: m68k-randconfig-m041-20230129 (https://download.01.org/0day-ci/archive/20230129/202301292052.ofpUPNlC-lkp@intel.com/config)
compiler: m68k-linux-gcc (GCC) 12.1.0

If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@intel.com>
| Reported-by: Dan Carpenter <error27@gmail.com>

smatch warnings:
block/blk.h:343 bio_may_exceed_limits() error: we previously assumed 'bio->bi_io_vec' could be null (see line 340)

vim +343 block/blk.h

83d016ac86428d Bart Van Assche   2018-01-17  308  
3ad5cee5cd000d Christoph Hellwig 2020-03-24  309  ssize_t part_size_show(struct device *dev, struct device_attribute *attr,
3ad5cee5cd000d Christoph Hellwig 2020-03-24  310  		char *buf);
3ad5cee5cd000d Christoph Hellwig 2020-03-24  311  ssize_t part_stat_show(struct device *dev, struct device_attribute *attr,
3ad5cee5cd000d Christoph Hellwig 2020-03-24  312  		char *buf);
3ad5cee5cd000d Christoph Hellwig 2020-03-24  313  ssize_t part_inflight_show(struct device *dev, struct device_attribute *attr,
3ad5cee5cd000d Christoph Hellwig 2020-03-24  314  		char *buf);
3ad5cee5cd000d Christoph Hellwig 2020-03-24  315  ssize_t part_fail_show(struct device *dev, struct device_attribute *attr,
3ad5cee5cd000d Christoph Hellwig 2020-03-24  316  		char *buf);
3ad5cee5cd000d Christoph Hellwig 2020-03-24  317  ssize_t part_fail_store(struct device *dev, struct device_attribute *attr,
3ad5cee5cd000d Christoph Hellwig 2020-03-24  318  		const char *buf, size_t count);
581d4e28d9195a Jens Axboe        2008-09-14  319  ssize_t part_timeout_show(struct device *, struct device_attribute *, char *);
581d4e28d9195a Jens Axboe        2008-09-14  320  ssize_t part_timeout_store(struct device *, struct device_attribute *,
581d4e28d9195a Jens Axboe        2008-09-14  321  				const char *, size_t);
581d4e28d9195a Jens Axboe        2008-09-14  322  
c55ddd9082f757 Christoph Hellwig 2022-07-27  323  static inline bool bio_may_exceed_limits(struct bio *bio,
aa261f20589d89 Bart Van Assche   2022-10-25  324  					 const struct queue_limits *lim)
abd45c159df5fc Jens Axboe        2021-10-13  325  {
abd45c159df5fc Jens Axboe        2021-10-13  326  	switch (bio_op(bio)) {
abd45c159df5fc Jens Axboe        2021-10-13  327  	case REQ_OP_DISCARD:
abd45c159df5fc Jens Axboe        2021-10-13  328  	case REQ_OP_SECURE_ERASE:
abd45c159df5fc Jens Axboe        2021-10-13  329  	case REQ_OP_WRITE_ZEROES:
abd45c159df5fc Jens Axboe        2021-10-13  330  		return true; /* non-trivial splitting decisions */
abd45c159df5fc Jens Axboe        2021-10-13  331  	default:
abd45c159df5fc Jens Axboe        2021-10-13  332  		break;
abd45c159df5fc Jens Axboe        2021-10-13  333  	}
abd45c159df5fc Jens Axboe        2021-10-13  334  
abd45c159df5fc Jens Axboe        2021-10-13  335  	/*
b91c4a42d14ad8 Bart Van Assche   2022-09-27  336  	 * Check whether bio splitting should be performed. This check may
b91c4a42d14ad8 Bart Van Assche   2022-09-27  337  	 * trigger the bio splitting code even if splitting is not necessary.
abd45c159df5fc Jens Axboe        2021-10-13  338  	 */
b91c4a42d14ad8 Bart Van Assche   2022-09-27  339  	if (blk_queue_sub_page_limits(lim) &&
b91c4a42d14ad8 Bart Van Assche   2022-09-27 @340  	    bio->bi_io_vec && bio->bi_io_vec->bv_len > lim->max_segment_size)
b91c4a42d14ad8 Bart Van Assche   2022-09-27  341  		return true;
c55ddd9082f757 Christoph Hellwig 2022-07-27  342  	return lim->chunk_sectors || bio->bi_vcnt != 1 ||
abd45c159df5fc Jens Axboe        2021-10-13 @343  		bio->bi_io_vec->bv_len + bio->bi_io_vec->bv_offset > PAGE_SIZE;
abd45c159df5fc Jens Axboe        2021-10-13  344  }
abd45c159df5fc Jens Axboe        2021-10-13  345  

:::::: The code at line 343 was first introduced by commit
:::::: abd45c159df5fcb7ac820e2825dac85de7c01c21 block: handle fast path of bio splitting inline

:::::: TO: Jens Axboe <axboe@kernel.dk>
:::::: CC: Jens Axboe <axboe@kernel.dk>

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests

^ permalink raw reply	[flat|nested] 3+ messages in thread
* [bvanassche:block-small-segments 10/13] block/blk.h:343 bio_may_exceed_limits() error: we previously assumed 'bio->bi_io_vec' could be null (see line 340)
@ 2023-01-30 11:36 ` Dan Carpenter
  0 siblings, 0 replies; 3+ messages in thread
From: Dan Carpenter @ 2023-01-30 11:36 UTC (permalink / raw)
  To: oe-kbuild, Bart Van Assche; +Cc: lkp, oe-kbuild-all

tree:   https://github.com/bvanassche/linux block-small-segments
head:   c776ad1c2e1e0afaefb3ef5f1710dc3886f5348f
commit: b91c4a42d14ad8bfd74387c10c8c21db9fb2b998 [10/13] block: Add support for filesystem requests and small segments
config: m68k-randconfig-m041-20230129 (https://download.01.org/0day-ci/archive/20230129/202301292052.ofpUPNlC-lkp@intel.com/config)
compiler: m68k-linux-gcc (GCC) 12.1.0

If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@intel.com>
| Reported-by: Dan Carpenter <error27@gmail.com>

smatch warnings:
block/blk.h:343 bio_may_exceed_limits() error: we previously assumed 'bio->bi_io_vec' could be null (see line 340)

vim +343 block/blk.h

c55ddd9082f757 Christoph Hellwig 2022-07-27  323  static inline bool bio_may_exceed_limits(struct bio *bio,
aa261f20589d89 Bart Van Assche   2022-10-25  324  					 const struct queue_limits *lim)
abd45c159df5fc Jens Axboe        2021-10-13  325  {
abd45c159df5fc Jens Axboe        2021-10-13  326  	switch (bio_op(bio)) {
abd45c159df5fc Jens Axboe        2021-10-13  327  	case REQ_OP_DISCARD:
abd45c159df5fc Jens Axboe        2021-10-13  328  	case REQ_OP_SECURE_ERASE:
abd45c159df5fc Jens Axboe        2021-10-13  329  	case REQ_OP_WRITE_ZEROES:
abd45c159df5fc Jens Axboe        2021-10-13  330  		return true; /* non-trivial splitting decisions */
abd45c159df5fc Jens Axboe        2021-10-13  331  	default:
abd45c159df5fc Jens Axboe        2021-10-13  332  		break;
abd45c159df5fc Jens Axboe        2021-10-13  333  	}
abd45c159df5fc Jens Axboe        2021-10-13  334  
abd45c159df5fc Jens Axboe        2021-10-13  335  	/*
b91c4a42d14ad8 Bart Van Assche   2022-09-27  336  	 * Check whether bio splitting should be performed. This check may
b91c4a42d14ad8 Bart Van Assche   2022-09-27  337  	 * trigger the bio splitting code even if splitting is not necessary.
abd45c159df5fc Jens Axboe        2021-10-13  338  	 */
b91c4a42d14ad8 Bart Van Assche   2022-09-27  339  	if (blk_queue_sub_page_limits(lim) &&
b91c4a42d14ad8 Bart Van Assche   2022-09-27 @340  	    bio->bi_io_vec && bio->bi_io_vec->bv_len > lim->max_segment_size)
                                                            ^^^^^^^^^^^^^^
Check for NULL

b91c4a42d14ad8 Bart Van Assche   2022-09-27  341  		return true;
c55ddd9082f757 Christoph Hellwig 2022-07-27  342  	return lim->chunk_sectors || bio->bi_vcnt != 1 ||
abd45c159df5fc Jens Axboe        2021-10-13 @343  		bio->bi_io_vec->bv_len + bio->bi_io_vec->bv_offset > PAGE_SIZE;
                                                                ^^^^^^^^^^^^^^^^
Unchecked dereference

abd45c159df5fc Jens Axboe        2021-10-13  344  }

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests


^ permalink raw reply	[flat|nested] 3+ messages in thread
* Re: [bvanassche:block-small-segments 10/13] block/blk.h:343 bio_may_exceed_limits() error: we previously assumed 'bio->bi_io_vec' could be null (see line 340)
  2023-01-30 11:36 ` Dan Carpenter
  (?)
@ 2023-01-30 17:17 ` Bart Van Assche
  -1 siblings, 0 replies; 3+ messages in thread
From: Bart Van Assche @ 2023-01-30 17:17 UTC (permalink / raw)
  To: Dan Carpenter, oe-kbuild; +Cc: lkp, oe-kbuild-all

On 1/30/23 03:36, Dan Carpenter wrote:
> tree:   https://github.com/bvanassche/linux block-small-segments
> head:   c776ad1c2e1e0afaefb3ef5f1710dc3886f5348f
> commit: b91c4a42d14ad8bfd74387c10c8c21db9fb2b998 [10/13] block: Add support for filesystem requests and small segments
> config: m68k-randconfig-m041-20230129 (https://download.01.org/0day-ci/archive/20230129/202301292052.ofpUPNlC-lkp@intel.com/config)
> compiler: m68k-linux-gcc (GCC) 12.1.0
> 
> If you fix the issue, kindly add following tag where applicable
> | Reported-by: kernel test robot <lkp@intel.com>
> | Reported-by: Dan Carpenter <error27@gmail.com>
> 
> smatch warnings:
> block/blk.h:343 bio_may_exceed_limits() error: we previously assumed 'bio->bi_io_vec' could be null (see line 340)
> 
> vim +343 block/blk.h
> 
> c55ddd9082f757 Christoph Hellwig 2022-07-27  323  static inline bool bio_may_exceed_limits(struct bio *bio,
> aa261f20589d89 Bart Van Assche   2022-10-25  324  					 const struct queue_limits *lim)
> abd45c159df5fc Jens Axboe        2021-10-13  325  {
> abd45c159df5fc Jens Axboe        2021-10-13  326  	switch (bio_op(bio)) {
> abd45c159df5fc Jens Axboe        2021-10-13  327  	case REQ_OP_DISCARD:
> abd45c159df5fc Jens Axboe        2021-10-13  328  	case REQ_OP_SECURE_ERASE:
> abd45c159df5fc Jens Axboe        2021-10-13  329  	case REQ_OP_WRITE_ZEROES:
> abd45c159df5fc Jens Axboe        2021-10-13  330  		return true; /* non-trivial splitting decisions */
> abd45c159df5fc Jens Axboe        2021-10-13  331  	default:
> abd45c159df5fc Jens Axboe        2021-10-13  332  		break;
> abd45c159df5fc Jens Axboe        2021-10-13  333  	}
> abd45c159df5fc Jens Axboe        2021-10-13  334
> abd45c159df5fc Jens Axboe        2021-10-13  335  	/*
> b91c4a42d14ad8 Bart Van Assche   2022-09-27  336  	 * Check whether bio splitting should be performed. This check may
> b91c4a42d14ad8 Bart Van Assche   2022-09-27  337  	 * trigger the bio splitting code even if splitting is not necessary.
> abd45c159df5fc Jens Axboe        2021-10-13  338  	 */
> b91c4a42d14ad8 Bart Van Assche   2022-09-27  339  	if (blk_queue_sub_page_limits(lim) &&
> b91c4a42d14ad8 Bart Van Assche   2022-09-27 @340  	    bio->bi_io_vec && bio->bi_io_vec->bv_len > lim->max_segment_size)
>                                                              ^^^^^^^^^^^^^^
> Check for NULL
> 
> b91c4a42d14ad8 Bart Van Assche   2022-09-27  341  		return true;
> c55ddd9082f757 Christoph Hellwig 2022-07-27  342  	return lim->chunk_sectors || bio->bi_vcnt != 1 ||
> abd45c159df5fc Jens Axboe        2021-10-13 @343  		bio->bi_io_vec->bv_len + bio->bi_io_vec->bv_offset > PAGE_SIZE;
>                                                                  ^^^^^^^^^^^^^^^^
> Unchecked dereference
> 
> abd45c159df5fc Jens Axboe        2021-10-13  344  }

Hi Dan,

I think in this case the unchecked dereference is safe.
bio->bi_vcnt == 1 implies that bio->bi_io_vec is not NULL.

Thanks,

Bart.



^ permalink raw reply	[flat|nested] 3+ messages in thread
end of thread, other threads:[~2023-01-30 17:17 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-01-29 12:21 [bvanassche:block-small-segments 10/13] block/blk.h:343 bio_may_exceed_limits() error: we previously assumed 'bio->bi_io_vec' could be null (see line 340) kernel test robot
2023-01-30 11:36 ` Dan Carpenter
2023-01-30 17:17 ` Bart Van Assche

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.