Jack,
Thanks for your response. Ali had also responded on the Q&A site with the same suggestion. It took me a couple days to get around to trying it out, but I can now confirm that this does indeed work. I may add something to the wiki (http://www.m5sim.org/Debugger_Based_Debugging) to make it more clear that you have to actually modify code and recompile in order to have this work. Along those lines, I found that you also need to modify a variable (current_debugger) if you want the "call debugger()" to work. A few comments on the Q&A thread (http://qa.gem5.org//351/gdb-debugging-in-syscall-mode<http://qa.gem5.org/351/gdb-debugging-in-syscall-mode>) are suggesting that something should be done to make this a little easier to use (e.g. command line option) and avoid needing to recompile - I agree J Thanks! -Dan From: gem5-users-boun...@gem5.org [mailto:gem5-users-boun...@gem5.org] On Behalf Of Jack Harvard Sent: Friday, February 07, 2014 3:01 PM To: gem5 users mailing list Subject: Re: [gem5-users] GDB debugging of simulated program Hi Dan, Does the following not work? If you would like to have the simulator wait for gdb to attach before executing you just need to change a line in src/sim/system.cc: int rgdb_wait = -1; to rgdb_wait = 0 (If you only have one cpu or the cupid for the one you want to wait for it to attach to). Jack Jack Harvard On Wed, Feb 5, 2014 at 10:16 PM, Walters, Dan <dwalt...@mitre.org<mailto:dwalt...@mitre.org>> wrote: What is the recommended way to attach a remote GDB to Gem5 in order to debug the program being run in simulation, particularly when the simulated program finishes execution very quickly? I've read through http://www.m5sim.org/Debugger_Based_Debugging and was able to get it working with an example program (hello-world in an infinite loop), but the program that I actually want to debug does not have an infinite loop, so it finishes executing in the simulation before I have a chance to connect the remote debugger. I've tried using gem5.debug with the -debug-break cmdline option, but since that causes gem5 itself to hit a breakpoint, the remote GDB's attempt to connect times-out. I've tried various experiments with trying to resume gem5 at the same time as telling the remote GDB to connect, but the race condition is too tight so I never get the simulated program to break before finishing execution... there must be a better way. Has anyone else done something like this with any success? I'd appreciate any ideas. BTW, I'm using gem5-stable to simulate an ARM platform in SE mode. Also, this same question was asked on the gem5 Q&A website a couple weeks ago (http://qa.gem5.org/351/gdb-debugging-in-syscall-mode ), but it hasn't gotten much attention... it seems like all questions to the mailing list in general get more attention... should I avoid the Q&A site? Thanks, Dan _______________________________________________ gem5-users mailing list gem5-users@gem5.org<mailto:gem5-users@gem5.org> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
_______________________________________________ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
