[Top][All Lists]

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

Re: [Qemu-devel] [PATCH v7 00/74] linux-user: Split do_syscall

From: Richard Henderson
Subject: Re: [Qemu-devel] [PATCH v7 00/74] linux-user: Split do_syscall
Date: Wed, 22 May 2019 07:29:08 -0400
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.6.1

[Adding the mailing list back in.]

On 5/22/19 2:47 AM, Aleksandar Markovic wrote:
>> > Version 6 was
>> > https://lists.gnu.org/archive/html/qemu-devel/2019-01/msg04794.html
>> >
>> > Since v5, I've fixed the ioctl failure that Laurent found, and
>> > have done a few more syscalls.
>> >
>> > I've tried to do more testing with LTP....
> Running LTP is certainly a very important way of testing this series. However,
> it is not sufficient. This is because this series completely replaces internal
> QEMU strace mechanism for involved system calls,

What you missed because of the very large time lag between iterations, and
because I was lazy and only referenced v6 in this cover letter, is that in an
early revision the major objection was that I *wasn't* replacing strace, and
thus we were retaining lots of duplicate logic in multiple places.

> and any bug in such
> replacement would not be captured by LTP results. How did you test new strace
> bits and peaces? Can you provide logs of such tests (for example, before/after
> comparisons of strace output for relevant system calls, or similar)?

My strace testing is totally ad-hoc.  Run something with -strace and eyeball it.

Here are before/after logs for busybox ls (aka linux-user-test-0.3, which has
for some reason vanished from the qemu wiki?) for alpha-linux-user.

Attachment: ls-strace-before
Description: Text document

Attachment: ls-strace-after
Description: Text document

reply via email to

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