This is an automated email from Gerrit. Alamy Liu ([email protected]) just uploaded a new patch set to Gerrit, which you can find at http://openocd.zylin.com/3347
-- gerrit commit 4915db5c454be4a37c384977b46224aba6353c7e Author: Alamy Liu <[email protected]> Date: Tue Feb 16 17:47:26 2016 -0800 The TAG of a progress of ARMv8 on Juno R1 Note: These patches are meant for reference. Althoug it might work, but unlikely. No time to re-organize all the lot patches, No time to verify the new re-based code and fixing all the stuff. Just checked in, in case of ... I believe there are some good stuff/idea in it. i.e.: The use of uintmax_t to solve n-bit problem forever co-exist of target_type & target_type_64 (v8 could switch mode on exception) armv8_cti.* is fully tested except armv8_cti_enable_cross_halt_restart() simplify the single/smp handling in aarch64.c aarch64_halt_one / aarch64_halt_smp aarch64_poll_one / aarch64_poll_smp ... That might have benefit to other developers. Just my little hope~ ^_^ ------------------------------ Status * Use different configuration files for Single core/SMP debugging > armv8_juno_r1.a57x1.cfg (for Single core --- i.e.: UEFI) > armv8_juno_r1.a57x2.cfg (for SMP --- i.e.: Linux kernel) Fix: Should be able to combined into one configuration file, enable/disable from GDB client * Registers --- good > GDB "J/j" protocol not implemented ("info reg" always shows one core's register, in SMP) > Alternative: use "monitor targets [a57.cpu0|a57.cpu1|...]" to switch target. use "monitor reg" to examine register values. > In SMP, "set $<reg> = <value>" always set the same target (because "J/j" protocol not implemented _yet_) Fix: Implement the "J/j" command protocol. * Memory --- good > Only through debugging interface, no AHB, no MMU on/off handling, ...etc. Fix: review/fix all the other memory related functions. * Halt --- works * Resume --- works * Breakpoint --- works * ECT/CTI (single / smp) --- works > Embedded Cross Trigger / Cross Trigger Interface * Source code level debugging > UEFI --- code is relocated, there is problem to re-define symbol address. > Kernel --- good. * Step --- Not verified > If Debug State was entered due to breakpoint, delete breakpoint before "continue". Reason: When "continue" on a breakpoint, GDB issues a "step" command to step over that instruction before "continue" * Watchpoint --- NOT implemented (same status as v7:cortex_a) (That was the work based on 2dcf7bf77d09f139 svf: flush the queue before reallocing memory) ------------------------------ These patches are rebased on Feb-15,2016 source 50d4f76e13 Add board config for ST NUCLEO L476RG board * Resolve merging conflict > Some fixes are already in arm_adi_v5.* and adi_v5_jtag.c during the time. But the definitions are a little bit different > The moving of dap_ap_select() into mem_ap_sel_xxx(). Need to reivew aarch64.c, armv8_cti.c, and/or arm_adi_v5.c Change-Id: Ia44909855870926836dce247de6ec215b65e6023 Signed-off-by: Alamy Liu <[email protected]> -- ------------------------------------------------------------------------------ Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140 _______________________________________________ OpenOCD-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/openocd-devel
