qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: "Daniel P. Berrangé" <berrange@redhat.com>
Cc: qemu-devel@nongnu.org
Subject: Re: [PATCH v2 00/26] block: fix coroutine_fn annotations
Date: Mon, 9 May 2022 15:30:31 +0200	[thread overview]
Message-ID: <65ed1c57-82d5-21aa-c7cf-63695c9d2b99@redhat.com> (raw)
In-Reply-To: <YnkAsxsrvan0yCWY@redhat.com>

On 5/9/22 13:53, Daniel P. Berrangé wrote:
>> Apart from this, I also identified the following functions that
>> can be called both in coroutine context and outside:
> snip
> 
>> - qio_channel_readv_full_all_eof
>> - qio_channel_writev_full_all
>
> I'm trying to understand what criteria makes those two functions
> liable for the annotation, but not others in the I/O code ?

These two are _not_ annotated as coroutine_fn: this is a list of 
functions that can be called both in coroutine context and outside, 
while 'coroutine_fn' functions can be called only within coroutines.

The only 'coroutine_fn' function in the I/O code is qio_channel_yield.

> What is the actual rule for when to apply 'coroutine_fn' annotation
> to a function, and does it apply transitively to up and/or down the
> call stack ?

The only rule is that callers of coroutine_fn must be coroutine_fn 
themselves, or the call must be within "if (qemu_in_coroutine())".  For 
example:

- qio_channel_readv_full_all_eof() calls qio_channel_yield() within such 
an "if", therefore it need not be coroutine_fn.

- qio_channel_yield() unconditionally calls qemu_coroutine_yield() which 
is coroutine_fn, and therefore must be coroutine_fn as well.

After this series, the only exception to the rule is that 
qemu_coroutine_self() is occasionally called from tracepoints.

Paolo


      reply	other threads:[~2022-05-09 13:47 UTC|newest]

Thread overview: 66+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-09 10:29 [PATCH v2 00/26] block: fix coroutine_fn annotations Paolo Bonzini
2022-05-09 10:29 ` [PATCH v2 01/26] block: remove incorrect " Paolo Bonzini
2022-05-10 15:03   ` Eric Blake
2022-07-04 12:54   ` Alberto Faria
2022-05-09 10:29 ` [PATCH v2 02/26] qcow2: " Paolo Bonzini
2022-05-10 15:04   ` Eric Blake
2022-07-04 12:56   ` Alberto Faria
2022-05-09 10:29 ` [PATCH v2 03/26] nbd: " Paolo Bonzini
2022-05-10 15:05   ` Eric Blake
2022-07-04 12:56   ` Alberto Faria
2022-05-09 10:29 ` [PATCH v2 04/26] coroutine: " Paolo Bonzini
2022-05-10 15:12   ` Eric Blake
2022-07-04 12:57   ` Alberto Faria
2022-05-09 10:29 ` [PATCH v2 05/26] blkdebug: add missing " Paolo Bonzini
2022-05-10 15:30   ` Eric Blake
2022-05-09 10:29 ` [PATCH v2 06/26] blkverify: " Paolo Bonzini
2022-05-10 15:59   ` Eric Blake
2022-07-04 12:57   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 07/26] block: " Paolo Bonzini
2022-05-13 21:26   ` Eric Blake
2022-05-14  9:54     ` Paolo Bonzini
2022-05-16 18:49       ` Alberto Faria
2022-05-17  9:51         ` Paolo Bonzini
2022-05-09 10:30 ` [PATCH v2 08/26] file-posix: " Paolo Bonzini
2022-05-13 21:52   ` Eric Blake
2022-07-04 12:57   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 09/26] iscsi: " Paolo Bonzini
2022-05-18 14:44   ` Eric Blake
2022-07-04 12:58   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 10/26] nbd: " Paolo Bonzini
2022-05-18 14:50   ` Eric Blake
2022-07-04 12:58   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 11/26] nfs: " Paolo Bonzini
2022-07-04 12:59   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 12/26] nvme: " Paolo Bonzini
2022-07-04 12:59   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 13/26] parallels: " Paolo Bonzini
2022-07-04 12:59   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 14/26] qcow2: " Paolo Bonzini
2022-07-04 12:59   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 15/26] copy-before-write: " Paolo Bonzini
2022-07-04 12:59   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 16/26] curl: " Paolo Bonzini
2022-07-04 13:00   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 17/26] qed: " Paolo Bonzini
2022-07-04 13:01   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 18/26] quorum: " Paolo Bonzini
2022-07-04 13:01   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 19/26] throttle: " Paolo Bonzini
2022-07-04 13:02   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 20/26] vmdk: " Paolo Bonzini
2022-07-04 13:02   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 21/26] job: " Paolo Bonzini
2022-07-04 13:02   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 22/26] coroutine-lock: " Paolo Bonzini
2022-07-04 13:03   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 23/26] raw-format: " Paolo Bonzini
2022-07-04 13:03   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 24/26] 9p: " Paolo Bonzini
2022-07-04 13:03   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 25/26] migration: " Paolo Bonzini
2022-07-04 13:03   ` Alberto Faria
2022-05-09 10:30 ` [PATCH v2 26/26] test-coroutine: " Paolo Bonzini
2022-07-04 13:03   ` Alberto Faria
2022-05-09 11:53 ` [PATCH v2 00/26] block: fix " Daniel P. Berrangé
2022-05-09 13:30   ` Paolo Bonzini [this message]

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=65ed1c57-82d5-21aa-c7cf-63695c9d2b99@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=berrange@redhat.com \
    --cc=qemu-devel@nongnu.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 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).