Linux SCSI subsystem development
 help / color / mirror / Atom feed
From: Bart Van Assche <bvanassche@acm.org>
To: Christoph Hellwig <hch@lst.de>
Cc: "Martin K . Petersen" <martin.petersen@oracle.com>,
	linux-scsi@vger.kernel.org, John Garry <john.garry@linux.dev>
Subject: Re: [PATCH v5 7/8] scsi: scsi_debug: Improve lock context annotations
Date: Mon, 12 Oct 2026 08:31:01 -0700	[thread overview]
Message-ID: <89e3bb6c-4e8a-4480-a6d4-e7f6a2ddbef1@acm.org> (raw)
In-Reply-To: <20261012122457.GB31394@lst.de>

On 10/12/26 5:24 AM, Christoph Hellwig wrote:
> On Thu, Oct 08, 2026 at 03:42:18PM -0700, Bart Van Assche wrote:
>>   static inline void
>>   sdeb_meta_read_lock(struct sdeb_store_info *sip)
>> +	__acquires_shared(&sip->macc_meta_lck)
>> +	__context_unsafe(/*conditional locking*/)
>>   {
>> -	if (sdebug_no_rwlock) {
>> -		if (sip)
>> -			__acquire(&sip->macc_meta_lck);
>> -		else
>> -			__acquire(&sdeb_fake_rw_lck);
>> -	} else {
>> +	if (!sdebug_no_rwlock) {
> 
> Why can't this be modelled using a __acquire variant?  And if you
> add __context_unsafe(/*conditional locking*/) please add a detailed
> comment why that is needed, otherwise no one is going to understand
> this in the future.

Hi Christoph,

I will change the comment into "can't be annotated with
__{acquire,release}(,_shared}() without triggering undefined behavior".

What Sashiko reported on a Google-internal server for an
earlier attempt to annotate this and similar functions with
__{acquire,release}(,_shared}() has been included below. I think that
Sashiko is right:
=======================================================================
 >  static inline void
 >  sdeb_meta_read_lock(struct sdeb_store_info *sip)
 > +	__acquires_shared(&sip->macc_meta_lck)
 >  {
 >  	if (sdebug_no_rwlock) {
 > -		if (sip)
 > -			__acquire(&sip->macc_meta_lck);
 > -		else
 > -			__acquire(&sdeb_fake_rw_lck);
 > +		__acquire_shared(&sip->macc_meta_lck);
 >  	} else {
 > -		if (sip)
 > +		if (sip) {
 >  			read_lock(&sip->macc_meta_lck);
 > -		else
 > +		} else {
 >  			read_lock(&sdeb_fake_rw_lck);
 > +			__release_shared(&sdeb_fake_rw_lck);
 > +			__acquire_shared(&sip->macc_meta_lck);
                                          ^^^^^^^^^^^^^^^^^^^
Can this regression trigger an undefined behavior splat when sip is NULL?
When the scsi_debug driver is used with the module parameter fake_rw=1,
devip2sip() returns NULL. In this else branch, sip is explicitly NULL,
so evaluating &sip->macc_meta_lck means taking the address of a member
from a NULL pointer.
When compiled with Context Analysis and UBSAN (-fsanitize=null or
-fsanitize=pointer-overflow), the inline function call evaluates the
argument before the call and will trigger a runtime splat. Without UBSAN,
it calculates an invalid offset from a NULL pointer.
=======================================================================

Bart.



  reply	other threads:[~2026-10-12 15:31 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-08 22:42 [PATCH v5 0/8] scsi_debug: Enable lock context analysis Bart Van Assche
2026-10-08 22:42 ` [PATCH v5 1/8] scsi: scsi_debug: Fix a locking bug in resp_write_same() Bart Van Assche
2026-10-08 22:42 ` [PATCH v5 2/8] scsi: scsi_debug: Split resp_write_same() Bart Van Assche
2026-10-08 22:52   ` sashiko-bot
2026-10-08 23:27     ` Bart Van Assche
2026-10-08 22:42 ` [PATCH v5 3/8] scsi: scsi_debug: Split corrupt_lbas() Bart Van Assche
2026-10-08 22:42 ` [PATCH v5 4/8] scsi: scsi_debug: Split resp_read_dt0() Bart Van Assche
2026-10-12 12:23   ` Christoph Hellwig
2026-10-12 17:18   ` John Garry
2026-10-08 22:42 ` [PATCH v5 5/8] scsi: scsi_debug: Split resp_write_dt0() Bart Van Assche
2026-10-08 22:42 ` [PATCH v5 6/8] scsi: scsi_debug: Split resp_atomic_write() Bart Van Assche
2026-10-08 22:42 ` [PATCH v5 7/8] scsi: scsi_debug: Improve lock context annotations Bart Van Assche
2026-10-12 12:24   ` Christoph Hellwig
2026-10-12 15:31     ` Bart Van Assche [this message]
2026-10-08 22:42 ` [PATCH v5 8/8] scsi: core: Enable lock context analysis for the scsi_debug driver Bart Van Assche

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=89e3bb6c-4e8a-4480-a6d4-e7f6a2ddbef1@acm.org \
    --to=bvanassche@acm.org \
    --cc=hch@lst.de \
    --cc=john.garry@linux.dev \
    --cc=linux-scsi@vger.kernel.org \
    --cc=martin.petersen@oracle.com \
    /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