Jean-Rene Peulve wrote:
Hi Nathan,
Oups: I resend this because the previous email was html and got bounced
by gdb@sources.redhat.com
Thanks for testing the patch out!
I have been forced to change your patch to make it works.
1) missing return in server.c
The return you've added is unnecessary.
2) I also add an fprintf to stderr with the address of the text and data
section to allow the use of add-symbol_file command which requires these
address.
The response to qOffsets being hidden by gdb, I then can get it from
the target console or telnet.
Dan, I'll leave this change for you to decide. I don't understand the
usage here. Why is it not necessary to know the text_end address in this
case, for instance?
+ if (the_target->read_offsets (&text, &data)) {
+ sprintf (own_buf, "Text=%lX;Data=%lX;Bss=%lX",
+ (long)text, (long)data, (long)data);
+ return;
This return is not needed, why did you find it necessary?
+ }
+ else
+ write_enn (own_buf);
+
+ return;
+ }
nathan
--
Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery
nathan@codesourcery.com :: http://www.planetfall.pwp.blueyonder.co.uk