This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] new MI command for pattern filling of memory regions
- From: Tom Tromey <tromey at redhat dot com>
- To: Giuseppe MONTALTO <giuseppe dot montalto at st dot com>
- Cc: "gdb-patches\ at sourceware dot org" <gdb-patches at sourceware dot org>
- Date: Wed, 09 May 2012 12:11:15 -0600
- Subject: Re: [PATCH] new MI command for pattern filling of memory regions
- References: <76FE3225DF13124EA2D05B290B624C95E65F2D466E@SAFEX1MAIL1.st.com>
>>>>> "Giuseppe" == Giuseppe MONTALTO <giuseppe.montalto@st.com> writes:
Giuseppe> it's the first time I attempt a contribution to gdb, so feel
Giuseppe> free to ask for details and, please, let me know if I'm doing
Giuseppe> anything wrong!
I was told off-list that you are covered by your employer's assignment.
So, good.
Giuseppe> the proposed enhancement is about memory filling.
Giuseppe> Filling a memory region with a pattern of bytes so far, can
Giuseppe> only be done by sending a potentially high number of repeated
Giuseppe> commands to gdb (depending on the size of the region that we
Giuseppe> need to fill). If gdb is controlled through a GUI
Giuseppe> (e.g. Eclipse/CDT), this may cause the UI to freeze for a
Giuseppe> quite long time. This is especially true when you're
Giuseppe> debugging a remote target, due to the amount of generated
Giuseppe> network traffic.
I started to review this, but I wondered why a new command is necessary.
It seems like you could send any hex string to want to
-data-write-memory-bytes.
Alternatively, if you really want the repeat count, why not just add an
optional argument to -data-write-memory-bytes? That would seem to be
simpler.
Tom