["Martin Mathieson" <martin.r.mathieson@xxxxxxxxxxxxxx>]

On Mon, Aug 4, 2008 at 1:19 PM, Luis EG Ontanon <luis@xxxxxxxxxxx> wrote:

I think it's might be a fake. next_tvb is considered volatile...
because dissector_try_port() does not specify the argument as constant
so the compiler thinks it might change while popping the stack on a
longjmp.

Hi, does the attached patch gets rid of the warning?

Thanks Luis,

Unfortunately not.  I can't see what is different between this case and the one in the built-in packet-ethertype.c which may have been used as a template in part for the plugin packet-infiniband.c.

For now, I've made next_tvb static, and that lets the build complete...

Martin
 

Luis

On Mon, Aug 4, 2008 at 1:13 PM, Martin Mathieson
<martin.r.mathieson@xxxxxxxxxxxxxx> wrote:
> Hi,
>
> I've had this build failure for the past few days.  I haven't looked at it
> deeply, but I don't understand how the exception stuff is working.
>
> packet-infiniband.c: In function `parse_PAYLOAD':
> packet-infiniband.c:806: warning: variable 'next_tvb' might be clobbered by
> `longjmp' or `vfork'
> make[3]: *** [packet-infiniband.lo] Error 1
> make[3]: Leaving directory `/users/martinm/wireshark/
> plugins/infiniband'
>
>
> Any ideas?
> Martin

> _______________________________________________
> Wireshark-dev mailing list
> Wireshark-dev@xxxxxxxxxxxxx
> https://wireshark.org/mailman/listinfo/wireshark-dev
>
>



--
This information is top security. When you have read it, destroy yourself.
-- Marshall McLuhan

_______________________________________________
Wireshark-dev mailing list
Wireshark-dev@xxxxxxxxxxxxx
https://wireshark.org/mailman/listinfo/wireshark-dev