Original Submission:
I'm not sure if this is a bug so to speak, but when I set a breakpoint on a piece of code, I expect the code execution to stop before that line of code is executed, e.g. if a breakpoint is set on the very first instruction, I expect no instructions to have been executed when the debugger gains control.

Perhaps it would be nice to offer users the options to break after or before the instruction highlighted, in some systems setting dialog perhaps?

Please log in, so followups can be emailed to you.

If you cannot login, then enter your email address here:

Followups

Comment	Date	By
Doesn't this mean shifting all the code down to make room for the nop nop pair? Again, that could be impossible (for one, there may not be any memory for it, for another, jump addresses would have to be rewritten).	2003-Oct-08 14:05	sauce
OK, I half-expected that would be the case. I'll have to just get used to it and work around it :) Wouldn't it be possible to have an option where setting a breakpoint actually inserts a NOP NOP pair before the instruction yo uwish to break on, and breaks there instead? Can you set breakpoints on NOPs? On second thoughts, this is probably a bad road to go down...	2003-Oct-08 12:38	eratosthenes
Breaking before an instruction is intractible -- there could be a number of instructions before any given instruction -- the only way to know is at runtime. The debugger would have to break on the prior instruction and all candidate instructions (basically every branch/jump instruction) and try to determine whether or not the instruction would be executed next. This could cause a tremendous slow-down. The VU's execute an instruction which is marked as a breakpoint and stop immediately afterwards.	2003-Oct-07 14:27	sauce

Bug Change History

Field	Old Value	Date	By
assigned_to	nobody	2003-Oct-07 14:27	sauce