[RFC][PATCH 01/15] Fix endian problem for tracepoint enabled flag
Pedro Alves
palves@redhat.com
Thu Oct 15 17:16:00 GMT 2015
On 10/14/2015 12:14 PM, henrik.wallin@windriver.com wrote:
> From: Par Olsson <par.olsson@windriver.com>
>
> When running big endian machines there is a problem with
> the enabled flag for tracepoints as it is defined as a
> int8_t but written from gdbserver as an integer and then
> read in the agent as 8-bit value.
> This caused problem when tracepoint was disabled and
> re-enabled.
>
> gdb/gdbserver/ChangeLog:
>
> * tracepoint.c (struct tracepoint): Change type of enabled.
Seems wasteful for no obvious reason. Why not make gdbserver write
one byte then?
Thanks,
Pedro Alves
More information about the Gdb-patches
mailing list