qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: P J P <ppandit@redhat.com>
To: Dmitry Fleytman <dmitry@daynix.com>
Cc: Qemu Developers <qemu-devel@nongnu.org>,
	Paolo Bonzini <pbonzini@redhat.com>, Li Qiang <liqiang6-s@360.cn>
Subject: Re: [Qemu-devel] [PATCH] scsi: pvscsi: limit process IO loop to maximum page count
Date: Tue, 13 Sep 2016 16:18:13 +0530 (IST)	[thread overview]
Message-ID: <alpine.LFD.2.20.1609131555030.29381@wniryva> (raw)
In-Reply-To: <36CEB473-4A85-4C0E-AF37-668D3F21B373@daynix.com>

  Hello Dmitry,

+-- On Tue, 13 Sep 2016, Dmitry Fleytman wrote --+
| > On 13 Sep 2016, at 10:00 AM, P J P <ppandit@redhat.com> wrote:
| > 
| > +-- On Wed, 7 Sep 2016, P J P wrote --+
| > | From: Prasad J Pandit <pjp@fedoraproject.org>
| > | 
| > | Vmware Paravirtual SCSI emulator while processing IO requests
| > | could run into an infinite loop if 'pvscsi_ring_pop_req_descr'
| > | always returned positive value. Limit IO loop to the maximum
| 
| Do you see any specific scenario why this might happen?

  A guest user could set the 'ready_ptr' and 'PVSCSIRingInfo *mgr' parameter 
in 'pvscsi_ring_pop_req_descr', such that it always returns true.

| > | Reported-by: Li Qiang <liqiang6-s@360.cn>
| > |  pvscsi_process_io(PVSCSIState *s)
| > |  {
| > | +    int descr_pa_cnt = 0;
| > |      PVSCSIRingReqDesc descr;
| > |      hwaddr next_descr_pa;
| > |  
| > |      assert(s->rings_info_valid);
| > | -    while ((next_descr_pa = pvscsi_ring_pop_req_descr(&s->rings)) != 0) {
| > | +    while (((next_descr_pa = pvscsi_ring_pop_req_descr(&s->rings)) != 0)
| > | +            && descr_pa_cnt++ < PVSCSI_SETUP_RINGS_MAX_NUM_PAGES) {
| 
| Why do you limit number of processed descriptors by maximal number of pages 
| in data exchange ring? What will happen to requests still waiting in the 
| ring after this function exits?

  I limit it to maximum page count thinking that the descriptor value returned 
by pvscsi_ring_pop_req_descr() is derived from the mgr->req_ring_pages_pa[] 
array, which is of size 'PVSCSI_SETUP_RINGS_MAX_NUM_PAGES'. If 
pvscsi_process_io() was to go into an infinite loop, it'd continue processing 
the same set of req_ring_pages?

--
Prasad J Pandit / Red Hat Product Security Team
47AF CE69 3A90 54AA 9045 1053 DD13 3D32 FE5B 041F

  reply	other threads:[~2016-09-13 10:48 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-09-07  4:43 [Qemu-devel] [PATCH] scsi: pvscsi: limit process IO loop to maximum page count P J P
2016-09-13  7:00 ` P J P
2016-09-13  8:31   ` Dmitry Fleytman
2016-09-13 10:48     ` P J P [this message]
2016-09-13 12:10       ` Dmitry Fleytman
2016-09-13 13:09         ` P J P

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=alpine.LFD.2.20.1609131555030.29381@wniryva \
    --to=ppandit@redhat.com \
    --cc=dmitry@daynix.com \
    --cc=liqiang6-s@360.cn \
    --cc=pbonzini@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).