This GPIO driver is common to SoCs that have in-order ARM cores
(BCM2835) as well as superscalar (BCM2836-7) and speculative
out-of-order cores (BCM2711).
For BCM2837 and BCM2711, the processor can dual-issue stores and
is free to merge writes to peripheral memory for pages mapped
MT_NORMAL_NC, which is the default provided by /dev/[gpio]mem.
This can cause glitches (or missing edges) on GPIO pins when
toggled with no delay, as pipelined writes to the same address
can get arbitrarily squelched.
To prevent this happening, make sure the preceding write ops are
flushed outside the shareable domain by using a memory barrier.
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.com>
Change-Id: I8805cc0911667bcb9b7f4ca340d7f4f1cb25d096
Reviewed-on: https://review.openocd.org/c/openocd/+/7258
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Currently 'number' field is zero in the register cache and
this causes an issue on `rtos get_thread_reg_list` calls.
Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: Iaef11e01f55d012969bbc1933f82847d5e02fec5
Reviewed-on: https://review.openocd.org/c/openocd/+/7246
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
arp_waitstate will not work on not-examined state
Change-Id: I56c3e1c7e63af108e4ed1dbacebb567f9bf46264
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7230
Tested-by: jenkins
Reviewed-by: Erwan Gouriou
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
An application often idling in real sleep mode may make a Cortex-M target
hard to access as CPU clock are gated and debug requests are responded
by WAIT ack.
Try to examine the target under reset as the last resort.
Change-Id: I7c3de39fb1e6c23b76e2a0a85ab75f23aac94c4d
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7229
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
An application often idling in real sleep mode may make a Cortex-M target
hard to access as CPU clock are gated and debug requests are responded
by WAIT ack.
Try to examine the target under reset as the last resort.
Change-Id: Ife875a966a838c37dde987bc584ad0a1f4d020d6
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7228
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
After merging [1] 'reset halt' does not work on not responding Cortex-M.
Relax the examined tests and try to set vector catch VC_CORERESET
if debug_ap is available.
While on it add an info about examination state to debug logs.
Fixes: [1] commit 98d9f1168c ("target: reset target examined flag if target::examine() fails")
Change-Id: Ie2e018610026180af5997d70231061a275f05c76
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/6745
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
For 64-bit TAR setup, if 'tar_valid == false' perform the upper 32-bit
write even if the cached copy matches the upper TAR value to be written.
Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com>
Change-Id: I320377dc90a9d1d7b64cbb281b2527e56c7621ee
Reviewed-on: https://review.openocd.org/c/openocd/+/7245
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Read values must be at albuff so that can be copied to buffer
on function exit.
Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: I74a533e8f12f1002ca06a98a7c7cd928552b4cc5
Reviewed-on: https://review.openocd.org/c/openocd/+/7226
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Actually they are the base of epc and eps
Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: I4f43b9609a9929399fb5d3fa0203efc8a98e94c9
Reviewed-on: https://review.openocd.org/c/openocd/+/7227
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
ST-Link v2 dongles can be used with many cortex-m parts, but they have
one limitation -- they can only perform 8-bit and 32-bit writes to the
target. 16-bit writes are done using a pair of 8-bit writes. While not
usually an issue, in the case of the at91samd flash driver, the 16-bit
'command' register must have both halves written in the same
operation.
Fortunately, this register has two pad bytes above it in the address
space, making it safe to always access with 32-bit operations.
Change-Id: I44b0db9406982a8db5818c0533d3101618741db2
Signed-off-by: Keith Packard <keithp@keithp.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7234
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
There is no need to count number of examined threads twice.
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Change-Id: Id32ead853d1ddcd4e67062d6f795700feb20cb4b
Reviewed-on: https://review.openocd.org/c/openocd/+/7223
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
The bitbang driver floods the log by many messages with very
little informational value.
Remove some LOG_DEBUGs, convert some others to LOG_DEBUG_IO.
Change-Id: I0c7539467b45543e12932c67dc71e86d58c8c6cd
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7220
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
Tested-by: jenkins
Make write to DP_TARGETSEL is logged the similar way as other DP register
read/writes.
While on it fix checkpatch message
'Concatenated strings should use spaces between elements'
Change-Id: I98f724c984e8c4610cc461340f4c4a7cc9627ed9
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7219
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
Tested-by: jenkins
armv7m_start_algorithm() stored all non-debug execution
registers from register cache without checking validity.
Check if the register cache is valid.
Try to read from CPU if not valid.
Issue a warning if register read fails.
Change-Id: I365f86d65243230cf521b13909575e5986a87a50
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7240
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
The num_reg_params parameter of target_run_algorithm() was not
updated when setting "sp" was introduced. Therefore "sp" as the last
register parameter was not passed to a target algo.
Introduce a new helper variable with correct count of register parameters
and use it everywhere needed.
Change-Id: I934a71380783d98917167f1569145808ef23540f
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7225
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
Change-Id: Idf3bce842b4507c1f12692b5fbcd6730637de9db
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7216
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
Tested-by: jenkins
Flash read_id/erase/write operation on running target failed
in target_run_algorithm() anyway. It generated lot of error messages.
Check the target state and bail out early if target is running.
Change-Id: I903f5f38c8e61016e5002b235e5f07803bd2ec4e
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7215
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
SPI flash erase often takes longer than the fixed timeout 3 seconds.
Introduce a configurable timeout_ms parameter to rp2040_call_rom_func().
Compute the erase timeout from the number of blocks to be erased.
While on it make the timeouts shorter for connect flash, flush cache and
enter/exit xip (1 second is enough).
Change-Id: I552bfa317ee17064de3a54ec2f0c63e84ba87222
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7214
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
The size of the flash write buffer should be rounded
down to the multiply of flash page size.
Using write chunks of unadjusted size results in write of chunks
unaligned to flash pages.
Change-Id: If7931362ee193dff4dc2df7ec78f13530658cf08
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7187
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
While on it restore memory-mapped mode also after flash erase
(originally was restored after flash write only).
Change-Id: I5e153b79ac27a8439f57239ce90ce8a79c0bb8a1
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7186
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Prepend stack_grab_and_prep() function name by rp2040_ prefix.
Introduce target helper variable in rp2040_stack_grab_and_prep()
and use it instead of dereferencing bank->target several times.
Move flash ID reading code to the new rp2040_spi_read_flash_id()
function.
Change-Id: I9d6e51e17e36e6230155a586065499f2f260089a
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7185
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
swd_multidrop_select() uses its own retry loop.
If select fails, do_reconnect flag remains set on exit and causes
useless reconnect.
Clear do_reconnect flag in retry loop.
Change-Id: Ie06d6967d7f4a977774c8530bb8d4b3e5ab4f62c
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7217
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
Tested-by: jenkins
Implementation of ADI v6 introduced banking of DP reg 0.
The accompanying change preventing DP SELECT write before
DP IDR read during connect was added to swd_connect_single() only.
Unchanged swd_connect_multidrop() / swd_multidrop_select_inner()
was broken as it emited DP SELECT and put DP to protocol error state.
Copy dap->select handling to swd_multidrop_select_inner().
Fixes: 72fb88613f (adiv6: add low level swd transport)
Change-Id: I514cd6d9ae2ba97ce3657b459df22638c278a0b1
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7213
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
Increase "value_buf" size so it can hold a 64-bit number represented
as a string. Previous size could only hold a 32-bit number string.
Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com>
Change-Id: If6fbc875236e6ddc59522fbc25db0129eb60ee27
Reviewed-on: https://review.openocd.org/c/openocd/+/7221
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Scan-build reports:
Logic error: Uninitialized argument value
riscv.c:2688 2nd function call argument is an uninitialized value
This is a real error cause by running the command "riscv
authdata_write" without arguments. In such case 'value' is not
initialized and is passed to and used by r->authdata_write().
Reorganize the code to:
- detect the correct amount or command's arguments;
- drop the LOG_ERROR() on ERROR_COMMAND_SYNTAX_ERROR;
- drop the 'else' after 'return'.
Change-Id: I62e031220593b8308bc674b753e15d16d4c5c9ac
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7210
Tested-by: jenkins
Reviewed-by: Jan Matyas <matyas@codasip.com>
Reviewed-by: Tim Newsome <tim@sifive.com>
Scan-build reports:
Logic error: Result of operation is garbage or undefined
riscv.c:1614 The result of the left shift is undefined due
to shifting by '4294967281', which is greater or
equal to the width of type 'target_addr_t'
This is a false warning due to clang that considers the impossible
case of 32 bits hart (xlen = 32) in SATP_MODE_SV48 mode
(info->va_bits = 48).
Under such case:
riscv.c:1614 ... ((target_addr_t)1 << (xlen - (info->va_bits - 1))) ...
the shift amount wraps around the unsigned type and assumes the
value 4294967281 (0xfffffff1).
Use assert() to prevent clang from complaining.
Change-Id: I08fdd2a806c350d061641e28cf15a51b397db099
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7209
Reviewed-by: Tim Newsome <tim@sifive.com>
Reviewed-by: Jan Matyas <matyas@codasip.com>
Tested-by: jenkins
Scan-build reports:
Unused code: Dead nested assignment
riscv.c:459 Although the value stored to 'ir_user4_raw' is
used in the enclosing expression, the value is
never actually read from 'ir_user4_raw'
This is caused by the value reassigned in 'ir_user4_raw':
riscv.c:459 ir_user4[3] = (uint8_t)(ir_user4_raw >>= 8);
but never used.
Drop the DIY conversion in favor of h_u32_to_le() that does not
reassign the input value.
Change-Id: Ifad29f4c46d4a2d0a2f5a5c4104d768cc3db2794
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7208
Reviewed-by: Tim Newsome <tim@sifive.com>
Reviewed-by: Jan Matyas <matyas@codasip.com>
Tested-by: jenkins
Scan-build reports:
Unused code: Dead assignment
riscv.c:716 Value stored to 'result' is never read
Remove the initialization of variable 'result'.
Change-Id: Ied67bb4fcfa5bace186522074247ead43a5d5cd5
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7207
Reviewed-by: Tim Newsome <tim@sifive.com>
Reviewed-by: Jan Matyas <matyas@codasip.com>
Tested-by: jenkins
Scan-build reports:
Unused code: Dead initialization
riscv-013.c:2362 Value stored to 'control' during its
initialization is never read
Remove the initialization of variable 'control'.
Change-Id: I548f8175530b9a2aa4c1788549d6467bf9824584
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7206
Reviewed-by: Tim Newsome <tim@sifive.com>
Reviewed-by: Jan Matyas <matyas@codasip.com>
Tested-by: jenkins
Scan-build complains about variable 'sbcs_orig' that can be used
not initialized.
Logic error: Assigned value is garbage or undefined
riscv-013.c:4468 Assigned value is garbage or undefined
This is caused by not checking the return value of the call
riscv-013.c:4466 dmi_read(target, &sbcs_orig, DM_SBCS);
In fact when dmi_read() returns error, the variable 'sbcs_orig' is
not assigned.
Check the returned value.
Change-Id: Ia9032a0229aa243138f95f4e13f765726a4ceae9
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7205
Reviewed-by: Tim Newsome <tim@sifive.com>
Reviewed-by: Jan Matyas <matyas@codasip.com>
Tested-by: jenkins
Scan-build triggers a warning:
Unix API: Allocator sizeof operand mismatch
dsp563xx.c:2143 Result of 'calloc' is converted to a pointer
of type 'uint8_t', which is incompatible with sizeof
operand type 'uint32_t'
It's a false positive because calloc() is properly used in this
case, as the uint8_t array is used in blocks of 4 elements to read
or write uint32_t values.
Either
calloc(sizeof(uint32_t), count);
and
malloc(count * sizeof(uint32_t));
keep triggering the same warning.
Drop the warning by using the constant '4' as size of uint32_t, as
already used few lines below.
Change-Id: I5bb1ece177774eefdc5d9cd049338f8f2be87cd7
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7203
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Doxygen complains about:
adapter.h:120: warning: Unsupported xml/html tag <signal_name> found
Move the text in double quote to remove the warning.
Change-Id: Ia4ead5caa83c84e10ee2b2359048c282892170b0
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: 82fd400542 ("jtag/adapter: Add command 'adapter gpio'")
Reviewed-on: https://review.openocd.org/c/openocd/+/7202
Tested-by: jenkins
With the old checkpatch we cannot use the correct format for the
SPDX tags in the file .c, in fact the C99 comments are not allowed
and we had to use the block comment.
With the new checkpatch, let's switch to the correct SPDX format.
Change created automatically through the command:
sed -i \
's,^/\* *\(SPDX-License-Identifier: .*[^ ]\) *\*/$,// \1,' \
$(find src/ contrib/ -name \*.c)
Change-Id: I6da16506baa7af718947562505dd49606d124171
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7153
Tested-by: jenkins
Resolves issues where registers are accessed when poll() logic is inactive or has not yet been triggered.
Signed-off-by: Ian Thompson <ianst@cadence.com>
Change-Id: If7a4d00938fb188b008325249627f7773c3484c5
Reviewed-on: https://review.openocd.org/c/openocd/+/7197
Tested-by: jenkins
Reviewed-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
When FreeRTOS (at least) is compiled with -flto, this leaves certain
static symbols with .lto_priv.0 appended to their name. Arguably this
could be considered to be a gdb or gcc bug, but one place to resolve it
for OpenOCD usage is here at symbol lookup time.
Note that the ".0" is for the first such instance of the variable as a
static; additional ones would end up as ".1", ".2", etc, and are not
considered here.
Signed-off-by: Tim Nordell <tnordell@airgain.com>
Change-Id: I03580b45e8ea364392ef4e05c96276416b390cb0
Reviewed-on: https://review.openocd.org/c/openocd/+/7179
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
This is in preparation for a future commit that looks for an optional
suffix of .lto_priv.0 on the symbol name.
Signed-off-by: Tim Nordell <tnordell@airgain.com>
Change-Id: If803332373825b73bc986bd4ea7dfaee9c625c0a
Reviewed-on: https://review.openocd.org/c/openocd/+/7178
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
This will be utilized for an upcoming refactorization to support -flto
compiled programs.
Signed-off-by: Tim Nordell <tnordell@airgain.com>
Change-Id: Id523c0b3ac2dad8b248ea0d2cac7b4dd2f83d293
Reviewed-on: https://review.openocd.org/c/openocd/+/7177
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Read PS from EPS[debuglevel] during fetch such that it reflects the correct value when read via telnet (not just via gdb_server).
Signed-off-by: Ian Thompson <ianst@cadence.com>
Change-Id: I8504f68989bc6d5fe451a8cb69d01c86f4ec0100
Reviewed-on: https://review.openocd.org/c/openocd/+/7195
Tested-by: jenkins
Reviewed-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
This updates the FreeRTOS module to use the M4F FPU stacking also for the
FPV5_SP and FPV5_DP FPUs, which are found on the Cortex-M7. The FPUs are
in fact different than the FPV4_SP found on the M4, but the register
stacking is the same.
Signed-off-by: Frank Dischner <frank.dischner@gmail.com>
Change-Id: I74c45d2cfb55f55e6c557f2450068ad3c2fe9497
Reviewed-on: https://review.openocd.org/c/openocd/+/6939
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
The method soft_reset_halt is optional; no need to add an
empty function.
Remove mips_mips64_soft_reset_halt() and move the TODO comment
in struct target_type.
Change-Id: Id541a75e7a08645568961d59b73a120c2238701f
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7184
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Jonathan Bell
Erhan Kurubas
Tomas Vanek
Daniel Goehring
Daniel Anselmi
Keith Packard
Evgeniy Naydanov
Ian Thompson
Antonio Borneo
Marc Schink
Tim Nordell
Toms Stūrmanis
Frank Dischner