David Brownell wrote: > On Tuesday 19 January 2010, Spencer Oliver wrote: >> David Brownell wrote: >>> On Tuesday 19 January 2010, Spencer Oliver wrote: >>>> Skip over a bkpt instruction if found on resume/step. >>>> Only software breakpoints known to openod are handled. >>>> So this handles the special case of a user added bkpt, eg. semi-hosting. >>>> >>>> Signed-off-by: Spencer Oliver <ntfr...@users.sourceforge.net> >>> OK by me, but I'd change the comment. The "user added" >>> breakpoints come from debugging activity ... users says >>> break here, OpenOCD says "OK!". Semihosting would be >>> more like "library added". >>> >>> And "OpenOCD" is not spelled "openod" except via typo. >>> Too bad spelling checkers are so poor with intermixed >>> C and English! ;) >>> >> Did you want leave this patch until 0.4 is out? > > It's a bugfix we have in hand, ready to merge. Unless you > see it as risky for some reason (what?) which creates a > reason for delaying it, my bias would be to merge the fix > ASAP. > >
i will push then. >> As a side note this issue will also apply to armv5 cores with the bkpt >> instruction. > > Curious ... there's already some logic to skip breakpoints > there, you're saying it doesn't handle this case right? > yes openocd will handle breakpoints it know about, the same as the v7m code did. A user added bkpt instruction is not handled, the only way of knowing is to read the actual instruction like the v7m code does now. I will have a look at the v5 case to see what if anything needs doing. Cheers Spen _______________________________________________ Openocd-development mailing list Openocd-development@lists.berlios.de https://lists.berlios.de/mailman/listinfo/openocd-development
