qemu-devel
[Top][All Lists]
Advanced

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

Re: [Qemu-devel] [PATCH v2 6/6] nbd: use generic trace subsystem instead


From: Eric Blake
Subject: Re: [Qemu-devel] [PATCH v2 6/6] nbd: use generic trace subsystem instead of TRACE macro
Date: Thu, 29 Jun 2017 15:12:54 -0500
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.2.0

On 06/21/2017 10:34 AM, Vladimir Sementsov-Ogievskiy wrote:
> Starting from this patch to enable traces use -trace option of qemu or
> -T, --trace option of qemu-img, qemu-io and qemu-nbd. For qemu traces
> also can be managed by qmp commands trace-event-{get,set}-state.

Maybe:

Let NBD use the trace mechanisms already present in qemu. Now you can
use the -trace optino of qemu, or the -T/--trace option of qemu-img,
qemu-io, and qemu-nbd, to select nbd traces.  For qemu, the QMP commands
trace-event-{get,set}-state can also toggle tracing on the fly.

I also mentioned in v1 that an actual command line example might be
helpful (for example, is it qemu-nbd --trace "nbd_*" to get ALL nbd
traces enabled?)

> 
> Recompilation with CFLAGS=-DDEBUG_NBD is no more needed, furthermore,
> DEBUG_NBD macro is removed from the code.

Yay!

> 
> Signed-off-by: Vladimir Sementsov-Ogievskiy <address@hidden>


> @@ -462,7 +460,7 @@ int nbd_receive_negotiate(QIOChannel *ioc, const char 
> *name, uint16_t *flags,
>  
>      memcpy(&nbd_magic, buf, 8);
>      nbd_magic = be64_to_cpu(nbd_magic);
> -    TRACE("Magic is 0x%" PRIx64, nbd_magic);
> +    trace_nbd_receive_negotiate_magic(nbd_magic);

Might be some churn based on the resolution to 5/6.

>  
>      if (memcmp(buf, "NBDMAGIC", 8) != 0) {
>          error_setg(errp, "Invalid magic received");
> @@ -474,7 +472,7 @@ int nbd_receive_negotiate(QIOChannel *ioc, const char 
> *name, uint16_t *flags,
>          goto fail;
>      }
>      magic = be64_to_cpu(magic);
> -    TRACE("Magic is 0x%" PRIx64, magic);
> +    trace_nbd_receive_negotiate_magic2(magic);

I'm not sure you need two different trace functions, since the message
is identical.  I'd just have trace_nbd_receive_magic(magic) and use that
at both call sites.

But overall worth having, even if we fine-tune it later.

Reviewed-by: Eric Blake <address@hidden>

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org

Attachment: signature.asc
Description: OpenPGP digital signature


reply via email to

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