On 31 March 2014 08:36, Tim Sander <[email protected]> wrote: > Hi Paul >> On Mon, Mar 31, 2014 at 08:56:08AM +0200, Tim Sander wrote: >> > I am pretty confident, that the ARM_MODE_MON enum value is invalid. >> > at least for Cortex A9. I don't have an idea where the 26 comes from. >> > Have not found any references. Perhaps some Hex Decimal confusion? >> >> This should have better been sent to the devel mailing list, as many >> devs are not reading the users' list at all. > Mh, as you say it. I was currently not aware that this was the users list, > even if this is obvious from the mailing list address. >> Can you please provide a reference (document and page number) to the >> official arm documentation you're reading? > Arm Architecture Reference Manual ARMv7-a and ARMv7-R edition: > Table B1-1 Arm (Page B1-6,1156) column "Mode encoding" Row Monitor. > The Mode encoding "10110" looks pretty binary to me. And if you search for > 10110 this is all over the place. But there is no entry for "11010" aka 26. >
This enum value was added in ec93209f and looking at the code the actual value is set to arm->core_type and never changed. It appears it is more of an identifier and its value is not important as long as it does not conflict with the other valid modes. However changing it will probably make it clearer to anyone reading the code. Are you able to send a patch to gerrit? Cheers Spen ------------------------------------------------------------------------------ _______________________________________________ OpenOCD-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/openocd-devel
