The periodic poll scans all the targets in the same order they
have been declared in the configuration file.
When one target in a SMP node halts, the transition is detected
in the following poll and this triggers a halt request to all the
other cores of the SMP node.
The targets that will be polled afterwards will be identified as
"halted", but the targets already scanned will remain as
"running" until the next periodic poll.
This creates a race condition with GDB; GDB sets the breakpoints
when runs the target and removes them as soon as the target is
halted. When it receives the halt event, it starts removing the
breakpoints and fails on the targets that are still reported as
"running".
Fixed by polling all the targets in the SMP node before informing
GDB about the halt event.
This implementation is almost copy/paste from the one in aarch64.
Change-Id: Id2bd99f1e56b014e48e9e34ccb891b4219c518f8
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4622
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
There is no reason to restrict the command "cortex_a dacrfixup"
to the EXEC phase only.
Change the command mode to ANY so the command can be used in
the initialization phase too.
Change-Id: I498cc6b2dbdc48b3b2dd5f0445519a51857b295f
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4623
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Commit ad6c71e151 introduced the
variable "mmu_enabled" whose pointer is passed to cortex_a_mmu()
that initialises it.
This initialization is not visible to the compiler that issue
a compile error.
The same situation is common across the same file and the usual
workaround is to initialize it to zero; thus the same fix i
applied here.
Ticket: https://sourceforge.net/p/openocd/tickets/197/
Fixes: commit ad6c71e151 ("cortex_a: fix virt2phys when mmu is disabled")
Change-Id: I77dec41acdf4c715b45ae37b72e36719d96d9283
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4619
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
When the MMU is not enabled on debug state entry, virt2phys cannot
perform a translation since it is unknown whether a valid MMU
configuration existed before. In this case, return the virtual
address as physical address.
Change-Id: I6f85a7a5dbc200be1a4b5badf10a1a717f1c79c0
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Reviewed-on: http://openocd.zylin.com/4480
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Commit bfc5c764df avoids reading
ttbcr and ttb0/1 at every virt2phys translation by caching them,
and it updates the cached values in armv7a_arch_state().
But the purpose of any (*arch_state)() method, thus including
armv7a_arch_state(), is to only print out and inform the user
about some architecture specific status.
Moreover, to reduce the verbosity during a GDB session, the
method (*arch_state)() is not executed anymore at debug state
entry (check use of target->verbose_halt_msg in src/openocd.c),
thus the state of translation table gets out-of-sync triggering
Error: Address translation failure
or even using a wrong address in the memory R/W operation.
In addition, the commit above breaks the case of armv7r by
calling armv7a_read_ttbcr() unconditionally.
Fixed by moving in cortex_a_post_debug_entry() the call to
armv7a_read_ttbcr() on armv7a case only.
Remove the call to armv7a_read_ttbcr() in armv7a_identify_cache()
since it is (conditionally) called only in the same procedure
cortex_a_post_debug_entry().
Fixes: bfc5c764df ("armv7a: cache ttbcr and ttb0/1 on debug
state entry")
Change-Id: Ifc20eca190111832e339a01b7f85d28c1547c8ba
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4601
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
- add 'dap create' command to create dap instances
- move all dap subcmmand into the dap instance commands
- keep 'dap info' for convenience
- change all armv7 and armv8 targets to take a dap
instance instead of a jtag chain position
- restructure tap/dap/target relations, jtag tap no
longer references the dap, daps are now independently
created and initialized.
- clean up swd connect
- re-initialize DAP also on JTAG errors (e.g. after reset,
power cycle)
- update documentation
- update target files
Change-Id: I322cf3969b5407c25d1d3962f9d9b9bc1df067d9
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Reviewed-on: http://openocd.zylin.com/4468
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
and move declaration of all transport_is_xxx() functions to transport.h
Change-Id: Ib229115b5017507b49655bc43b517ab6fb32f7a6
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4469
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
This patch adds support in openOCD to read/write Arm vector/floating
point registers. This is compatible with Arm vfp v3 target xml in GDB.
Please refer to binutils-gdb/gdb/features/arm/arm-vfpv3.xml
Change-Id: Id4dd1bddef51c558f1a86300c1a876d159463f18
Signed-off-by: Omair Javaid <omair.javaid@linaro.org>
Reviewed-on: http://openocd.zylin.com/4421
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Tested-by: jenkins
When debugging Thumb-2 code, Gdb will at times send a
breakpoint packet 'Z0,<addr>,3', the number 3 denoting that
the instruction to break on is 32 bits long. Handle this by
replacing it with two consecutive 16bit Thumb BKPTs and make
sure to save and restore the full, original 32bit
instruction.
Note that this fix is only applicable if you debug a bare-metal program
(like the linux kernel) with the 'wrong' gdb, e.g. use an
"arm-linux" gdb instead of an "arm-eabi" gdb. But since most people
may not know about the subtle differences between gdb configurations
regarding thumb2 breakpoints it's still valid.
Change-Id: Ib93025faf35b11f0dba747a8c1fc36fd09a4c0f8
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Reviewed-on: http://openocd.zylin.com/4241
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Tested-by: jenkins
Make DSCR_RUN_MODE() usable for armv8 and arm7 debug
Change-Id: Ib3ba3000d5b6aa03e590f3ca4969e677474eb12c
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Add new enum ARM_STATE_AARCH64 to the list of possible states.
Change-Id: I3cb2df70f8d5803a63d8374bf3eb75de988e24f8
Signed-off-by: David Ung <david.ung.42@gmail.com>
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Define a target_addr_t type to support 32-bit and 64-bit addresses at
the same time. Also define matching TARGET_PRI*ADDR format macros as
well as a convenient TARGET_ADDR_FMT.
In targets that are 32-bit (avr32, nds32, arm7/9/11, fm4, xmc1000)
be least invasive by leaving the formatting unchanged apart from the
type;
for generic code adopt TARGET_ADDR_FMT as unified address format.
Don't silently change gdb formatting here, leave that to later.
Add COMMAND_PARSE_ADDRESS() macro to abstract the address type.
Implement it using its own parse_target_addr() function, in the hopes
of catching pointer type mismatches better.
Add '--disable-target64' configure option to revert to previous 32-bit
target address behavior.
Change-Id: I2e91d205862ceb14f94b3e72a7e99ee0373a85d5
Signed-off-by: Dongxue Zhang <elta.era@gmail.com>
Signed-off-by: David Ung <david.ung.42@gmail.com>
[AF: Default to enabling (Paul Fertser), rename macros, simplify]
Signed-off-by: Andreas Färber <afaerber@suse.de>
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
OpenOCD doesn't (yet) know how to handle HYP mode properly so spsr
register is not getting initialised when OpenOCD connects to a target
stopped in this mode.
Reported on IRC by thinkfat and nearffxx.
Change-Id: I4bda9ba0c582c8e9cacefe708cc4a3d947151f84
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Reviewed-on: http://openocd.zylin.com/3906
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-by: Chengyu Zheng <chengyu.zheng@polimi.it>
Change b0698501b0 fixed
reset for i.MX6 and TI Sitara SoCs but broke reset for
cortex-a targets that use SWD. This patch is a work-
around that forces asserting SRST when SWD is used.
Change-Id: I7e39f2a469b9b4b2b74ad48ba49f2eeb58528921
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Reviewed-on: http://openocd.zylin.com/3641
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
This patch adds support for bridging semihosting to GDB's File-I/O
remote protocol extension. For the most part operations match up 1:1,
however some require a working area to complete successfully, namely
operations that devolve to read, stat, and gettimeofday.
A new command was added to enable support for fileio named `arm
semihosting_fileio`, which ensures that the default behavior remains
intact for those that prefer it.
Finally, redundant logging was removed from the target_arch_state
function; this permits ARM targets to quiesce log output when polling
for a fileio reply. This prevents filling the logs with halt/resume
messages when using semihosting fileio.
Change-Id: Ifbb864fc2373336a501cc0332675b887b552e1ee
Signed-off-by: Steven Stallion <stallion@squareup.com>
Reviewed-on: http://openocd.zylin.com/3566
Tested-by: jenkins
Reviewed-by: Steven Stallion <sstallion@gmail.com>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Depending on the Debug implementation the "OS Lock" feature might be
implemented or not. It is not actually depending on the part number of the
implemented ARM core but on the DBGOSLSR.OSLM bits. This patch removes
querying the part number and implements proper parsing of OSLM. Result is
a more generic approach that will work out-of-box on more devices.
Change-Id: I79e052869c2f9af1d7fdedef42faddb7292e7332
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3213
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
On multi-core systems, with some cores in power-down state, examination
will fail for these cores. Make sure assert- and deassert_reset functions
don't crash due to uninitialized variables.
Change-Id: I472f8d19af2cd3c770c05f3e57a31b35a863b687
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Reviewed-on: http://openocd.zylin.com/3552
Tested-by: jenkins
Reviewed-by: Jiri Kastner <cz172638@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Add semihosting support for ARMv7-A based processors.
Tested with custom Vybrid VF610 based board
and Pandaboard ES (Rev. B1) board (Cortex-A9).
Change-Id: I6b896a61c1c6a1c5dcf89de834486f82dd6c80a2
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Signed-off-by: Tsung-Han Lin <tsunghan.tw@gmail.com>
Reviewed-on: http://openocd.zylin.com/2908
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Tested-by: jenkins
Also make GPL notices consistent according to:
https://www.gnu.org/licenses/gpl-howto.html
Change-Id: I84c9df40a774958a7ed91460c5d931cfab9f45ba
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/3488
Tested-by: jenkins
Reviewed-by: Andreas Färber <afaerber@suse.de>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
It's Cortex-Xn, not Cortex Xn or cortex xn or cortex-xn or CORTEX-Xn
or CortexXn. Further it's Cortex-M0+, not M0plus.
Cf. http://www.arm.com/products/processors/index.php
Consistently write it the official way, so that it stops propagating.
Originally spotted in the documentation, it mainly affects code comments
but also Atmel SAM3/SAM4/SAMV, NiietCM4 and SiM3x flash driver output.
Found via:
git grep -i "Cortex "
git grep -i "Cortex-" | grep -v "Cortex-" | grep -v ".cpu"
git grep -i "CortexM"
Change-Id: Ic7b6ca85253e027f6f0f751c628d1a2a391fe914
Signed-off-by: Andreas Färber <afaerber@suse.de>
Reviewed-on: http://openocd.zylin.com/3483
Tested-by: jenkins
Reviewed-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Memory accesses are not made through the APB-AP, they are made through
the CPU (which happens to be controlled over the APB-AP). Rename all
irrelevant uses of the APB-AP term. And fix the long standing typo in
the function names...
Change-Id: Ide466fb2728930968bdba698f0dd9012cc9dbdf9
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/3216
Tested-by: jenkins
Reviewed-by: Andreas Färber <afaerber@suse.de>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
The cortex_a specific assert_reset function must only apply nSRST if
the reset configuration states that JTAG can be used while nSRST is
asserted.
Change-Id: If604a65fdea5bcb46ec723ada547a4e8d6fa8c59
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Reviewed-on: http://openocd.zylin.com/3356
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Before this change jim_target_reset() checked examined state of a target
and failed without calling .assert_reset in particular target layer
(and without comprehensible warning to user).
Cortex-M target (which refuses access to DP under active SRST):
If connection is lost then reset process fails before asserting SRST
and connection with MCU is not restored.
This resulted in:
1) A lot of Cortex-M MCUs required use of reset button or cycling power
after firmware blocked SWD access somehow (sleep, misconfigured clock etc).
If firmware blocks SWD access early during initialization, a MCU could
become completely inaccessible by SWD.
2) If OpenOCD is (re)started and a MCU is in a broken state unresponsive
to SWD, reset command does not work even if it could help to restore communication.
Hopefully this scenario is not possible under full JTAG.
jim_target_reset() in target.c now does not check examined state
and delegates this task to a particular target. All targets have been checked
and xx_assert_reset() (or xx_deassert_reset()) procedures were changed
to check examined state if needed. Targets except arm11, cortex_a and cortex_m
just fail if target is not examined although it may be possible to use
at least hw reset. Left as TODO for developers familiar with these targets.
cortex_m_assert_reset(): memory access errors are stored
instead of immediate returning them to a higher level.
Errors from less important reads/writes are ignored.
Requested reset always leads to a configured action.
arm11_assert_reset() just asserts hw reset in case of not examined target.
cortex_a_assert_reset() works as usual in case of not examined target.
Change-Id: I84fa869f4f58e2fa83b6ea75de84440d9dc3d929
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/2606
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
This feature is required for boards that use a programmatical way
to reset the cpu, like the TI Pandaboard with OMAP4. The board only
has a 14 pin JTAG header that doesn't feature SRST and is reset by
direct write to the PRM_RSTCTL register.
iMX6 can be reset through triggering the on-chip watchdog, but for these
methods to work reliably, access through the AHB-AP without interaction
with the CPU core is necessary.
Change-Id: I9a07a536adda83cc2f93e504384c8c7f0306220b
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Reviewed-on: http://openocd.zylin.com/3359
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
work around issues with software breakpoints when the text segment
is mapped read-only by the OS. Set DACR to "all-manager" to bypass
TLB permission checks on memory access.
Change-Id: I79fd9b32b04a4d538d489896470ee30b26b72b30
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3107
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Debug initialization blindly selects AP#0 as default, which is the AHB-AP
in many cases. This sets the default for target_read/write functions.
However, AHB-AP is the wrong choice, because it bypasses caches on read
and write and also makes some peripherals inaccessible (e.g. l2 outer
caches). This patch explicitely selects the APB-AP (debug_ap) as the
default.
Change-Id: I13f9e0750186d35dcfc135c8d67d437c5884d9c4
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3113
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
It's currently set during target creation but the AP that will be used
for the target is not even known.
Change-Id: I4502e7eb1fa8d90f746445b8cf8a4c21cb7d519e
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/3155
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
All mem_ap_* functions now make sure the SELECT register is updated with
the AP number that it's operating on. This shouldn't have to be handled
explicitly.
Change-Id: Ib193d8930fabb6a25715064355f98258c9580b5d
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/3153
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
This function does two separate things, powering up the DP and setting
up a MEM-AP. But the DP needs to be powered before even searching for a
MEM-AP to initialize and targets may have multiple MEM-APs that need
initializing.
Split the function into dap_dp_init() and mem_ap_init() and change all
call sites to use the appropriate one.
Change-Id: I92f55e09754a93f3f01dd8e5aa1ffdf60c856126
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/3151
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Change the debug_ap and memory_ap fields of the cortex_a target and
the debug_ap field of the cortex_m target to be pointers to the
struct adiv5_ap instead of AP numbers in some known DAP.
This reduces the dependency on the DAP struct in the targets and
enables MEM-AP accesses to take the relevant AP as parameter.
Change-Id: I39d7b134d78000564b7eec5bff464adf0ef89147
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/3147
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
The Cortex-A and Cortex-M keeps an arm_jtag struct around just to be
able to pass a pointer to it to one common JTAG function which anyway
only uses the TAP field.
Refactor the function to take a TAP directly, remove the legacy struct
from cortex instances and store the TAP pointer only in the DAP.
Cortex-M makes a call to arm_jtag_setup_connection() with the struct
but the function does nothing useful for a Cortex-M target so remove
the call.
Change-Id: I3b33709ef55372ef14522ed4337e9f2e817ae3ab
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/3142
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Separate out the values from adiv5_dap that are associated with a specific AP into a new struct, so we can properly support multiple APs. Remove the DAP struct from the armv7* structs, because we can have multiple CPUs per DAP, and we shouldn't have multiple DAP structs. Tidy up a few places where ap_current is used incorrectly.
Change-Id: I0c6ef4b49cc86b140366347aaf9b76c07cbab0a8
Signed-off-by: Patrick Stewart <patstew@gmail.com>
Reviewed-on: http://openocd.zylin.com/2984
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Prepare to support multiple cortex-m cores on one DAP. Uses mem_ap_sel_*
functions and removes mem_ap_* functions. Adds a new debug_ap
parameter to the cortex_m (currently set to zero as in existing code).
Change-Id: I6926029d1e7bf44a42d453d1aff349bda824ba72
Signed-off-by: Patrick Stewart <patstew@gmail.com>
Reviewed-on: http://openocd.zylin.com/2983
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Problem
No matter what target->coreid is, it always shows
Detected core 0 dbgbase: ...
In dap_lookup_cs_component(), it decreases the core index value to zero
in order to find the desired core.
The reference to coreidx is necessary considering "a device which has nested
ROM tables, with each core described in its own table." (by Paul Fertser).
Change-Id: I9b56d45d6edf6639e748a625ab27787f8e5a5776
Signed-off-by: Alamy Liu <alamy.liu@gmail.com>
Reviewed-on: http://openocd.zylin.com/2902
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
cortex_a_read_apb_ab_memory_fast() uses the wrong order of ITR and DSCR
writes when setting up the transfer. ARM DDI0406C says in C8.2 regarding
"Fast mode" operation to first switch to fast mode and then latch the
instruction in ITR. Current implementation first wrote ITR, causing
the instruction to be executed immediately, then switched to fast mode
without an instruction latched. Repeated reading of DTRTX didn't
execute LDC and thus replicated its current content into the whole buffer.
This patch uses the following, revised algorithm:
1) switch to non-blocking mode and issue the LDC for the first word
2) if more than one word is to be read:
- switch to fast mode
- latch the LDC instruction into ITR (it is _not_ executed)
- issue (count-1) reads of DTRTX register, each read returns the current
content of DTRTX and re-issues the latched instruction
-> now the second-to-last word is in the buffer and the LDC for the last
word has been issued.
3) wait for the last instruction to complete
4) switch back to non-blocking mode
5) Read DTRTX for the last (or: only) word and put it into the buffer
Change-Id: I44f5c585962ffa5af257c3d5a2a802c122b6b1e4
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3122
Tested-by: jenkins
Reviewed-by: Christopher Head <chead@zaber.com>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
When accessing memory through the ARM core, privilege levels and mmu
access permissions observed. Thus it depends on the current mode of the
ARM core whether an access is possible or not. the ARM in USR mode can
not access memory mapped to a higher privilege level. This means, if the
ARM core is halted while executing at PL0, the debugger would be
prevented from setting a breakpoint at an address with a higher privilege
level, e.g. in the OS kernel. This is not desirable.
cortex_a_check_address() tried to work around this by predicting if an
access would fail and switched the ARM core to SVC mode. However, the
prediction was based on hardcoded address ranges and only worked for
Linux and a 3G/1G user/kernel space split.
This patch changes the policy to always switch to SVC mode for memory
accesses. It introduces two functions cortex_a_prep_memaccess() and
cortex_a_post_memaccess() which bracket memory reads and writes. These
function encapsulate all actions necessary for preparation and cleanup.
Change-Id: I4ccdb5fd17eadeb2b66ae28caaf0ccd2d014eaa9
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3119
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Tested-by: jenkins
when disabling the mmu to access physical addresses, normally the d-cache
must be disabled as well. Disabling the d-cache also requires a full
clean&invalidate. However, since all memory writes are treated as write-
through no-allocate and memory reads do not allocate cache lines,
effectively the d-cache state does not change at all. We can therefore
save the the d-cache disabling and flushing.
This patch also simplifies the function a bit.
Change-Id: Ia17c56a28f432156429cd4596107e3652b788e63
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3114
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
for minimal impact on the hardware state, force all memory accesses to
bypass the caches and tlbs. This may actually be the default, but ARM
recommends in DDI0406C to set proper default values on debug init.
Change-Id: If5ac097b6ee725c047b1e86c2f90eabe16b98c7b
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3079
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Tested-by: jenkins
Call armv7a_l1_d_cache_flush_virt() before writing the breakpoint,
to make sure the d-cache is clean and invalid at the breakpoint
location down to PoC.
Call armv7a_l1_d_cache_inval_virt() after writing the breakpoint
again, so that d-cache will pick up the modified code.
Call armv7a_l1_i_cache_inval_virt() after writing the breakpoint
to memory to make the change visible to the CPU.
Change-Id: I24fc27058d99cb00d7f6002ccb623cab66b0d234
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3033
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Tested-by: jenkins
The following changes are implemented:
- Clean&Invalidate the VA range to PoC *before* the write takes place
- Remove SMP handling since DCCIMVA instruction already maintains SMP
coherence.
- Remove separate Invalidate step
Change-Id: I19fd3cc226d8ecf2937276fc63258b6a26e369a7
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3027
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Tested-by: jenkins
ARMv7 architecture allows up to 7 cache levels that are architecturally
visible, as opposed to "system caches", which are outside of the domain
defined by ARMv7 and require separate management. This patch enables
detection and identification of caches at all levels. It also implements
a new "flush-all" function that cleans & invalidates all cache levels to
the "Point of Coherence".
Change-Id: Ib77115d6044d39845907941c6f031e208f6e0aa5
Signed-off-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-on: http://openocd.zylin.com/3024
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Tested-by: jenkins
Antonio Borneo
Paul Bartholomew
Matthias Welwarsky
Tomas Vanek
Omair Javaid
Evan Hunter
David Ung
Dongxue Zhang
Paul Fertser
Steven Stallion
Andrey Smirnov
Marc Schink
Andreas Färber
Andreas Fritiofson
Matthias Welwarsky
Patrick Stewart
Alamy Liu