qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PULL v5 00/22] virtio, vhost, pci: fixes, features


From: Stefan Weil
Subject: Re: [Qemu-devel] [PULL v5 00/22] virtio, vhost, pci: fixes, features
Date: Thu, 2 Feb 2017 20:01:45 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Icedove/45.6.0

Am 02.02.2017 um 17:25 schrieb Peter Maydell:
On 2 February 2017 at 13:56, Peter Maydell <address@hidden> wrote:
On 31 January 2017 at 20:18, Michael S. Tsirkin <address@hidden> wrote:
virtio, vhost, pci: fixes, features

generic pci root port support
disable shpc by default
safer version of ARRAY_SIZE and QEMU_BUILD_BUG_ON
fixes and cleanups all over the place

Signed-off-by: Michael S. Tsirkin <address@hidden>

Applied, thanks.

...travis builds now fail for the --enable-tcg-interpreter config:
https://travis-ci.org/qemu/qemu/jobs/197648661

In file included from /home/travis/build/qemu/qemu/tcg/tcg.c:255:0:
/home/travis/build/qemu/qemu/tcg/tci/tcg-target.inc.c: In function ‘tcg_out_op’:
/home/travis/build/qemu/qemu/tcg/tci/tcg-target.inc.c:569:117: error:
negative width in bit-field ‘<anonymous>’
/home/travis/build/qemu/qemu/tcg/tci/tcg-target.inc.c:569:255: error:
negative width in bit-field ‘<anonymous>’
In file included from /home/travis/build/qemu/qemu/tcg/tcg.c:255:0:
/home/travis/build/qemu/qemu/tcg/tci/tcg-target.inc.c:578:115: error:
negative width in bit-field ‘<anonymous>’
/home/travis/build/qemu/qemu/tcg/tci/tcg-target.inc.c:578:255: error:
negative width in bit-field ‘<anonymous>’

These look to be because we were trying to use ARRAY_SIZE()
on a non-array, which was previously undetected. The use is
only in an assert() so fairly harmless.

Would somebody who cares about TCI like to provide a fix?

thanks
-- PMM



Other architectures either no longer use an assertion or
use tcg_debug_assert(s->tb_jmp_insn_offset != NULL), see
tcg/aarch64/tcg-target.inc.c and tcg/ppc/tcg-target.inc.c.

As the majority thinks that there is no longer a need for
an assertion here, I think that is the best solution for
TCI, too.

I'll send a patch.

Regards
Stefan



reply via email to

[Prev in Thread] Current Thread [Next in Thread]