qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Cédric Le Goater" <clg@redhat.com>
To: "Paolo Bonzini" <pbonzini@redhat.com>,
	"Cédric Le Goater" <clg@kaod.org>,
	qemu-devel@nongnu.org
Cc: "Stefan Hajnoczi" <stefanha@redhat.com>,
	"Daniel P . Berrangé" <berrange@redhat.com>
Subject: Re: [PATCH for-8.0 1/3] async: Suppress GCC13 false positive in aio_bh_poll()
Date: Tue, 21 Mar 2023 13:16:11 +0100	[thread overview]
Message-ID: <fabc9edf-697e-41d3-101a-bd7a6ddda84a@redhat.com> (raw)
In-Reply-To: <1dc87cd2-2f86-3580-3f29-a89de375c79d@redhat.com>

On 3/21/23 12:57, Paolo Bonzini wrote:
> On 3/21/23 09:33, Cédric Le Goater wrote:
>> +static void aio_bh_slice_insert(AioContext *ctx, BHListSlice *slice)
>> +{
>> +    QSIMPLEQ_INSERT_TAIL(&ctx->bh_slice_list, slice, next);
>> +}
>> +
>>   /* Multiple occurrences of aio_bh_poll cannot be called concurrently. */
>>   int aio_bh_poll(AioContext *ctx)
>>   {
>> @@ -164,7 +169,13 @@ int aio_bh_poll(AioContext *ctx)
>>       /* Synchronizes with QSLIST_INSERT_HEAD_ATOMIC in aio_bh_enqueue().  */
>>       QSLIST_MOVE_ATOMIC(&slice.bh_list, &ctx->bh_list);
>> -    QSIMPLEQ_INSERT_TAIL(&ctx->bh_slice_list, &slice, next);
>> +
>> +    /*
>> +     * GCC13 [-Werror=dangling-pointer=] complains that the local variable
>> +     * 'slice' is being stored in a global list in 'ctx->bh_slice_list'.
>> +     * Use a helper to silent the compiler
>> +     */
>> +    aio_bh_slice_insert(ctx, &slice);
>>       while ((s = QSIMPLEQ_FIRST(&ctx->bh_slice_list))) {
>>           QEMUBH *bh;
>> -- 
> 
> Sorry, but an API that has "insert" and not "remove", and where the argument is *expected to be* a local variable (which must be removed to avoid a dangling pointer---and the warning is exactly -Wdangling-pointer), ranks at least -7 in the bad API ranking[1].

:)

> I tried wrapping the BHListSlice and BHListSlice* into an iterator struct (which is also really overkill, but at least---in theory---it's idiomatic), but the code was hard to follow.
> 
> The fact that the workaround is so ugly, in my opinion, points even more strongly at the compiler being in the wrong here.

It was initially called slice_dangling_pointer_fixup() how's that ?

An alternative could be :

@@ -164,7 +164,14 @@ int aio_bh_poll(AioContext *ctx)
  
      /* Synchronizes with QSLIST_INSERT_HEAD_ATOMIC in aio_bh_enqueue().  */
      QSLIST_MOVE_ATOMIC(&slice.bh_list, &ctx->bh_list);
+    /*
+     * GCC13 [-Werror=dangling-pointer=] complains that the local variable
+     * 'slice' is being stored in the global list 'ctx->bh_slice_list'.
+     */
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wdangling-pointer="
      QSIMPLEQ_INSERT_TAIL(&ctx->bh_slice_list, &slice, next);
+#pragma GCC diagnostic pop
  
      while ((s = QSIMPLEQ_FIRST(&ctx->bh_slice_list))) {
          QEMUBH *bh;

May be that's more explicit. I wonder if we need to ifdef clang also.

Thanks,

C.



  reply	other threads:[~2023-03-21 12:17 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-03-21  8:33 [PATCH for-8.0 0/3] Fixes for GCC13 Cédric Le Goater
2023-03-21  8:33 ` [PATCH for-8.0 1/3] async: Suppress GCC13 false positive in aio_bh_poll() Cédric Le Goater
2023-03-21 10:22   ` Paolo Bonzini
2023-03-21 10:30     ` Daniel P. Berrangé
2023-03-21 10:57       ` Paolo Bonzini
2023-03-21 11:15   ` Stefan Hajnoczi
2023-03-21 11:57   ` Paolo Bonzini
2023-03-21 12:16     ` Cédric Le Goater [this message]
2023-03-21 13:02       ` Paolo Bonzini
2023-03-21  8:33 ` [PATCH for-8.0 2/3] target/s390x: Fix float_comp_to_cc() prototype Cédric Le Goater
2023-03-21  9:37   ` Philippe Mathieu-Daudé
2023-03-22  3:48   ` Richard Henderson
2023-03-22  6:51   ` Thomas Huth
2023-03-21  8:33 ` [PATCH for-8.0 3/3] target/ppc: Fix helper_pminsn() prototype Cédric Le Goater
2023-03-21 11:07   ` Daniel Henrique Barboza
2023-03-22  3:49   ` Richard Henderson

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=fabc9edf-697e-41d3-101a-bd7a6ddda84a@redhat.com \
    --to=clg@redhat.com \
    --cc=berrange@redhat.com \
    --cc=clg@kaod.org \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@redhat.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;
as well as URLs for NNTP newsgroup(s).