aboutsummaryrefslogtreecommitdiff
path: root/src/target
AgeCommit message (Collapse)AuthorFilesLines
2022-07-23openocd: src/target: replace the GPL-2.0-or-later license tagAntonio Borneo192-2501/+384
Replace the FSF boilerplate with the SPDX tag. The SPDX tag on files *.c is incorrect, as it should use the C99 single line comment using '//'. But current checkpatch doesn't allow C99 comments, so keep using standard C comments, by now. Change-Id: I255ad17235ff1e01bf0aa4deed4d944e1d693ddb Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/7071 Tested-by: jenkins
2022-07-23openocd: src: fix incorrect SPDX tagsAntonio Borneo11-22/+22
The SPDX tag is aimed at machine handling and it's thus expected to be placed in the first line in specific format. Move the SPDX tag to the first line and fix it where needed. The SPDX tag on files *.c is incorrect, as it should use the C99 single line comment using '//'. But current checkpatch doesn't allow C99 comments, so keep using standard C comments, by now. Change-Id: Ie9a05f530009d482a4116eebd147fd7e1ee3d41e Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/7066 Tested-by: jenkins
2022-07-23openocd: src: replace the incomplete GPL-2.0-or-later license tagAntonio Borneo6-71/+20
Few files have the FSF boilerplate without the latest statement on where to get the GPL license. Manually replace the FSF boilerplate with the SPDX tag. While there, reorganize the copyright statement. The SPDX tag on files *.c is incorrect, as it should use the C99 single line comment using '//'. But current checkpatch doesn't allow C99 comments, so keep using standard C comments, by now. Change-Id: I0c908d01c010e24f9c7e94885e7fbed4ecf26a86 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/7063 Tested-by: jenkins
2022-07-23openocd: src: add GPL license tag on files that miss itAntonio Borneo1-0/+2
Some file miss completely the license tag. Add the SPDX tag, using the same GPL-2.0-or-later license of the OpenOCD project. The SPDX tag on files *.c is incorrect, as it should use the C99 single line comment using '//'. But current checkpatch doesn't allow C99 comments, so keep using standard C comments, by now. Change-Id: I1fb51e722232d14f050458a820c3041de3dc9138 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/7058 Tested-by: jenkins
2022-07-23openocd: src: remove duplicated GPL license tagAntonio Borneo1-13/+0
SPDX tag is replacing the FSF boilerplate. For files that have both, remove the FSF boilerplate and keep only the SPDX tag. Change-Id: I04c4764d13d4ca92453f30ed16aeae53cd4c3fc8 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/7057 Tested-by: jenkins
2022-07-23openocd: build: add SPDX tagAntonio Borneo6-0/+12
Add the SPDX tag to makefiles, configuration scripts and tcl files present in the folders under src/ Change-Id: I1e4552aafe46ef4893d510da9d732c5f181784a4 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/7051 Tested-by: jenkins
2022-07-02semihosting: fix return value of SYS_READ and SYS_WRITEPavel Kirienko1-9/+3
ARM/RISC-V semihosting calls SYS_READ/SYS_WRITE require inversion of the result value as described in "Semihosting for AArch32 and AArch64". Prior to this patch, this was done correctly only if (semihosting->is_fileio==false). This patch has been tested with STM32F446. Change-Id: I1b34c8d8393f7dfa66ee6539904a2eaf8f9154b0 Signed-off-by: Pavel Kirienko <pavel.kirienko@gmail.com> Fixes: https://sourceforge.net/p/openocd/tickets/232/ Reviewed-on: https://review.openocd.org/c/openocd/+/6803 Tested-by: jenkins Reviewed-by: Tim Newsome <tim@sifive.com> Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-07-02target: esp_xtensa_smp: fix clang reportAntonio Borneo1-1/+1
The variable 'smp_break' is only set and used in case of smp. But clang cannot track if 'target->smp' get modified between the set and the use of 'smp_break', so it consider possible to use 'smp_break' uninitialized. Initialize 'smp_break' to silent clang. Change-Id: Ifa25a84fe9ffa25b8b58d7920ec77994c3b7ebfe Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Fixes: 77287b8d47b4 ("target: add Espressif ESP32 basic support") Reviewed-on: https://review.openocd.org/c/openocd/+/7050 Reviewed-by: Erhan Kurubas <erhan.kurubas@espressif.com> Tested-by: jenkins
2022-07-02semihosting: move semihosting_result_t from riscv.h to the semihosting_common.hErhan Kurubas4-29/+31
These enum values are useful for the arch level semihosting call handlers. Currently riscv uses them, we also need similar return codes for the xtensa. Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com> Change-Id: I8f63749cc203c59b07862f33edf3c393cd7e33a9 Reviewed-on: https://review.openocd.org/c/openocd/+/7039 Tested-by: jenkins Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-06-24semihosting: remove comparison with NULLAntonio Borneo1-1/+1
Commit b9526f140127 ("semihosting: permit redirection of semihosting I/O to TCP") introduces a new comparison with NULL. Remove it. Change-Id: Ice4333c50d16f7592f0ff86b1640217fa42e34f6 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Fixes: b9526f140127 ("semihosting: permit redirection of semihosting I/O to TCP") Reviewed-on: https://review.openocd.org/c/openocd/+/7031 Tested-by: jenkins
2022-06-24aarch64: fix handling of 'reset halt'Antonio Borneo1-5/+5
Commit 6c0151623cb0 ("aarch64: add support for "reset halt"") introduces the register setting to halt at reset vector, but: - does not consider the case 'srst_pulls_trst' that makes useless setting the registers as they will be erased by the pulled trst; - does not clean sticky errors in case of 'srst_gates_jtag'. Avoid any register initialization on 'srst_pulls_trst' and move the cleaning of sticky errors in the common block. Change-Id: I6f839f06f7b091e234ede31ec18096e51f017bcd Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Fixes: 6c0151623cb0 ("aarch64: add support for "reset halt"") Reviewed-on: https://review.openocd.org/c/openocd/+/7034 Tested-by: jenkins Reviewed-by: Christian Hoff <christian.hoff@advantest.com>
2022-06-24semihosting: add custom user command handlerErhan Kurubas2-19/+33
Custom user syscalls can be handled with target events in the TCL scripts. This patch gives another opportunity to handle custom syscalls in the c files. Besides that some utility functions are also exported for the custom handlers. Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com> Change-Id: Ice13d527540a0de0b2a8abda912ae3dcff3834b7 Reviewed-on: https://review.openocd.org/c/openocd/+/6889 Tested-by: jenkins Reviewed-by: Ian Thompson <ianst@cadence.com> Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-06-24esp32s2: convert counted timeout to timeval_msErhan Kurubas1-6/+8
Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com> Change-Id: Id685408281478cec0e7e886dbedb3b8972c7b652 Reviewed-on: https://review.openocd.org/c/openocd/+/7020 Tested-by: jenkins Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-by: Ian Thompson <ianst@cadence.com>
2022-06-24target: add Espressif ESP32-S3 basic supportErhan Kurubas4-2/+730
ESP32-S3 is a dual core Xtensa SoC Not full featured yet. Some of the missing functionality: -Semihosting -Flash breakpoints -Flash loader -Apptrace -FreeRTOS Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com> Change-Id: I44e17088030c96a9be9809f6579a4f16dbfc5794 Reviewed-on: https://review.openocd.org/c/openocd/+/6990 Tested-by: jenkins Reviewed-by: Ian Thompson <ianst@cadence.com> Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-06-24target: add Espressif ESP32 basic supportErhan Kurubas8-3/+1546
ESP32 is a dual core Xtensa SoC Not full featured yet. Some of the missing functionality: -Semihosting -Flash breakpoints -Flash loader -Apptrace -FreeRTOS Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com> Change-Id: I76fb184aa38ab9f4e30290c038b5ff8850060750 Reviewed-on: https://review.openocd.org/c/openocd/+/6989 Tested-by: jenkins Reviewed-by: Ian Thompson <ianst@cadence.com> Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-06-24arm_adi_v5: parse ROM tables behind SoC-600 APv1 adapterAntonio Borneo1-3/+12
Arm "CoreSight System-on-Chip SoC-600" specification describes a bridge "Access Port v1 adapter" aimed to "connect a legacy Access Port (AP) ... into an CoreSight Architecture v3 system". A ROM table can be located in the "legacy" part of the system, on the legacy AP behind the APv1 adapter. For the purpose of scanning the ROM tables, consider an ADIv6 SoC-600 APv1 adapter as an ADIv5 AP. Change-Id: I97d42fb77013c1251fb68d0caa4274086bf38a70 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6827 Tested-by: jenkins
2022-06-24adiv6: stay in same AP during dap_lookup_cs_component()Antonio Borneo1-0/+7
Configuration file can specify, as target's debug AP, an AP that contains a ROM table that points, in turn, to other APs. Current code in cortex_a and aarch64 is not able to handle a return from dap_lookup_cs_component() that points to another AP. While it could be interesting to specify 'root' as target's debug AP, drop any found value if it's not in the starting AP. Change-Id: Id206e4fa7a29e9402c8e2393026817b410bbb8bd Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6826 Tested-by: jenkins
2022-06-24adiv6: add support for ROM tables in APAntonio Borneo1-31/+141
ADIv6 adds AP that only contain a ROM table in the AP itself, that can point to other AP containing either another AP level ROM table or a MEM-AP to be parsed as usual. Add support for parsing AP level ROM tables. Change-Id: Ic25863b16463b8a6adc3b15e26db7fdca858d6df Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6467 Tested-by: jenkins
2022-06-24adiv6: prepare for AP level ROM tablesAntonio Borneo1-16/+28
ADIv6 adds AP that only contain a ROM table in the AP itself, that can point to other AP containing either another AP level ROM table or a MEM-AP to be parsed as usual. To handle recursive AP access, reorganize the code to: - pass the depth==0 from the command 'dap info'; - print the AP number as first line, adding proper indentation on depth>0; - align the following print with proper indentation. Change-Id: I5b811810c807fc51b307bd60f67817d9de2aa095 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6466 Tested-by: jenkins
2022-06-24arm_adi_v5: add option 'root' to 'dap info' commandAntonio Borneo3-6/+62
On ADIv6 the system root ROM table is found by reading the DAP DP registers BASEPTR0 and BASEPTR1. Add option 'root' to the commands 'dap info' to let it retrieve the system root ROM table's AP from DAP DP, then use such AP for following dump. Change-Id: I1789457a005faa3870c5d14f763378d2f6a5f095 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6462 Tested-by: jenkins
2022-06-24adiv6: use struct adiv5_ap->ap_num to contain the AP base addressAntonio Borneo7-63/+128
ADIv5 DAP can only have 256 AP, while ADIv6 can provide till 2**40 (1,099,511,627,776) AP per DAP. Reuse the field ap_num in struct adiv5_ap, currently used on ADIv5 to hold the ADIv5 AP number (apsel), to contain the ADIv6 AP base address. Convert struct adiv5_ap->ap_num to 64 bit and initialize it to DP_APSEL_INVALID for unused AP. Restrict dap_find_get_ap() to ADIv5 only. To be enhanced. On ADIv6, let dap_get_ap() return an already allocated AP, or allocate and return an unused AP. Add function is_ap_num_valid() and use it. Change-Id: Ib2fe8c7ec0d08393cd91c29fdac5d632dfc1e438 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6461 Reviewed-by: Daniel Goehring <dgoehrin@os.amperecomputing.com> Tested-by: jenkins
2022-06-24adiv6: add low level swd transportAntonio Borneo1-26/+52
During enter in SWD read DP_DPIDR without selecting the register bank through DP_SELECT_DPBANK. Handle the different format of DP_SELECT register. Change-Id: Iea1b8eb6ec94177e16a430d5885595a38e833eeb Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6697 Tested-by: jenkins
2022-06-24adi_v5_jtag: extend memaccess_tck to every AP accessKevin Burke1-8/+5
ADIv5 reports: Accessing AP registers or debug resources in connected device through an AP can be subjected to other variable response delays in the system. A debugger that can adapt to these delays and avoid wasting WAIT scans operates more efficiently and provides higher maximum data throughput. The existing code in OpenOCD uses extra tck only for accessing resources through an AP. Extend the use of extra tck also for accessing an AP register. Split from change https://review.openocd.org/6077/ Change-Id: I2082362e098d09f4ba0668e01f5196afc965c8f3 Signed-off-by: Kevin Burke <kevinb@os.amperecomputing.com> Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com> Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6460 Tested-by: jenkins
2022-06-24adiv6: add low level jtag transportKevin Burke3-19/+102
swd and dap-direct are not implemented yet Split from change https://review.openocd.org/6077/ Change-Id: I6d73d8adf6a6090001c5d4771325fb1d63c45e3c Signed-off-by: Kevin Burke <kevinb@os.amperecomputing.com> Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com> Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6459 Tested-by: jenkins
2022-06-24adi_v5_jtag: clear sticky overrun errorAntonio Borneo1-2/+2
By accessing invalid AP in JTAG mode, it's possible to trigger the error: JTAG-DP STICKY ERROR After that the sticky error is never cleared and the whole DAP gets not anymore accessible. Clean-up the sticky error once detected. Change-Id: I8b07263b30f9e46645f0c29084b8f1626e241f45 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6430 Tested-by: jenkins
2022-06-24adiv6: read ROM Table address sizeKevin Burke2-0/+20
Required for parsing ADIv6 ROM tables. Split from change https://review.openocd.org/6077/ Change-Id: I849543b7b4a4455b10bd9fc7da38a37849d71700 Signed-off-by: Kevin Burke <kevinb@os.amperecomputing.com> Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com> Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6458 Tested-by: jenkins
2022-06-24adiv6: re-organize mem_ap registers definitionKevin Burke3-86/+114
ADIv5 MEM-AP registers are a subset of ADIv6 MEM-AP registers and are located at different offset. To prepare for introducing ADIv6, add 'struct adiv5_dap *' as argument to ADIv5 registers macro. Check the ADI version and use the proper address. Both adapter drivers rshim and stlink are ADIv5 only, so let them use the ADIv5 macros only. Split from change https://review.openocd.org/6077/ Change-Id: Ib861ddcdab74637b2082cc9f2612dea0007d77b1 Signed-off-by: Kevin Burke <kevinb@os.amperecomputing.com> Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com> Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6457 Tested-by: jenkins
2022-06-24adiv6: add dap flags -adiv5 and -adiv6Kevin Burke2-0/+33
Add flags to 'dap create' command and set the field adi_version in struct adiv5_dap. Actually only ADIv5 is functional. Other patches are needed to get ADIv6 working. Split from change https://review.openocd.org/6077/ Change-Id: I63d3902f99a7f139c15ee4e07c19eae9ed4534b9 Signed-off-by: Kevin Burke <kevinb@os.amperecomputing.com> Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com> Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6456 Tested-by: jenkins
2022-06-24arm_adi_v5: add ap refcount and add get/put around ap useAntonio Borneo9-56/+271
While an ADIv5 DAP can only have 256 AP, ADIv6 can provide till 2**40 (1,099,511,627,776) AP per DAP. The actual trivial code implementation for ADIv5 (that uses an array of 256 ap in the struct adiv5_dap) cannot be extended as-is to handle ADIv6. The simple array of 256 AP can be reused as a dynamic storage for ADIv6 ap: - the ADIv5 AP number is replaced by the ADIv6 base address; - the index of the array (equal to ADIv5 AP number) has no link to any ADIv6 property; - the ADIv6 base_address has to be searched in the array of AP. The 256 elements in the AP array should be enough for any device available today. In future it can be easily increased, if needed. To efficiently use the 256 elements in the AP array, the code should associate one element of the array to an ADIv6 AP (through the AP base address), then cancel the association when the AP is not anymore needed. This is important to avoid saturating the AP array while exploring the device through 'dap apreg' commands. Add a reference counter in the struct adiv5_ap to track how many times the struct has been associated with the same base address. Introduce the function dap_get_ap() to associate and return the struct, and dap_put_ap() to release the struct. For the moment the code covers ADIv5 only, so the association is through the index. Use the two functions above and dap_find_get_ap() throughout the code. Check the return value of dap_get_ap(). It is always not NULL in the current ADIv5-only implementation, but can be NULL for ADIv6 when there are no more available AP in the array. Instrument dap_queue_ap_read() and dap_queue_ap_write() to log an error message if the AP has reference counter zero, meaning that the AP has not been 'get' yet. This helps identifying AP used without get/put, e.g. code missed by this patch, or merged later. Instrument dap_cleanup_all() to log an error message if an AP has reference counter not zero at openocd exit, meaning that the AP has not been 'put' yet. Change-Id: I98316eb42b9f3d9c9bbbb6c73b1091b53f629092 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6455 Reviewed-by: Daniel Goehring <dgoehrin@os.amperecomputing.com> Tested-by: jenkins
2022-06-10semihosting: fix accessing memory outside the bounds of the fn arrayErhan Kurubas1-3/+5
There is an accsess to wrong index, when arm semihosting_basedir command not used or basedir set to empty string. Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com> Change-Id: I3afa049d74b30496f5c03ba4ef67431784f81bdc Fixes: ce5027ab019a ("semihosting: add semihosting_basedir command") Reviewed-on: https://review.openocd.org/c/openocd/+/7005 Tested-by: jenkins Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com> Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-06-10arm_adi_v5: check for calloc() return valueAntonio Borneo1-0/+4
In function adiv5_jim_configure() check that calloc() returns a valid allocated memory pointer. Change-Id: I97287e168834693900341add9d9eb9a5f38c55b4 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reported-by: Tomas Vanek <vanekt@fbl.cz> Reviewed-on: https://review.openocd.org/c/openocd/+/7014 Tested-by: jenkins Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
2022-06-10arm_adi_v5: fix scan-build warning [3/3]Antonio Borneo1-0/+2
While scan-build complains that dap_p or ap_num_p could be NULL, the current code never passes NULL pointers. Add an assert() to silent scan-build and prevent any further use of the function with incorrect parameters. Change-Id: I656810dddcea61e85d85b13efb114f7607ef837c Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/7012 Tested-by: jenkins Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
2022-06-10arm_adi_v5: fix scan-build warning [2/3]Antonio Borneo1-1/+3
Commit d01b3d69ec17 ("arm_adi_v5: separate ROM table parsing from command output [3/3]") introduces a new scan-build warning because removing one return in case of error causes using uninitialized values. Add back the return on error. Change-Id: I10ddc548b756d34aaccc0511f091b4caa5ec271a Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Fixes: d01b3d69ec17 ("arm_adi_v5: separate ROM table parsing from command output [3/3]") Reviewed-on: https://review.openocd.org/c/openocd/+/7011 Tested-by: jenkins Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
2022-06-10arm_adi_v5: fix scan-build warning [1/3]Antonio Borneo1-2/+3
Commit 21f7885d1c2a ("arm_adi_v5: separate ROM table parsing from command output [1/3]") introduces a new scan-build warning because continues the execution even when dap_get_debugbase() returns error. The value of 'apid' can be uninitialized: 5th function call argument is an uninitialized value Check the return value and quit on error. While there, remove the useless initialization of 'dbgbase' that was apparently required for the same problem. Change-Id: Iade26a152925ee0f1bf114ed829b94f7ed5b254f Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Fixes: 21f7885d1c2a ("arm_adi_v5: separate ROM table parsing from command output [1/3]") Reviewed-on: https://review.openocd.org/c/openocd/+/7010 Tested-by: jenkins Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
2022-06-04target/riscv: drop unused variable registers_initializedTomas Vanek2-8/+0
Change-Id: If7bfe38ac273ce9e54003e003807e128cced1568 Signed-off-by: Tomas Vanek <vanekt@fbl.cz> Reviewed-on: https://review.openocd.org/c/openocd/+/6995 Tested-by: jenkins Reviewed-by: Tim Newsome <tim@sifive.com> Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-06-04target: add Espressif ESP32-S2 basic supportErhan Kurubas13-1/+4956
ESP32-S2 is a single core Xtensa chip. Not full featured yet. Some of the missing functionality: -Semihosting -Flash breakpoints -Flash loader -Apptrace -FreeRTOS Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com> Change-Id: I2fb32978e801af5aa21616c581691406ad7cd6bb Reviewed-on: https://review.openocd.org/c/openocd/+/6940 Reviewed-by: Tomas Vanek <vanekt@fbl.cz> Reviewed-by: Ian Thompson <ianst@cadence.com> Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com> Tested-by: jenkins
2022-06-03cortex_a: get rid of not needed log messagesTarek BOCHKATI1-1/+1
when using semi-hosting with cortex_a this LOG_INFO pollutes openocd console, so just reduce the log level of this message. Change-Id: I91aa70492f4e361b25a0e5517d0cf73f2f8ed599 Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com> Reviewed-on: https://gerrit.st.com/c/stm32ide/official/openocd/+/248225 Tested-by: Tarek BOCHKATI <tarek.bouchkati@st.com> Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@st.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6993 Tested-by: jenkins Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-05-29flash/nor/core, target: don't ask for working mem if no target algoTomas Vanek1-1/+1
The command 'flash erase_check' showed the message 'Running slow fallback erase check - add working memory' even in the case the target didn't implement blank_check_memory. Change return code of target_blank_check_memory() in this case and sense it in default_flash_blank_check() and show a message without a request for working memory. Change-Id: I7cf9bf77742964b4f377c9ce48ca689e57d0882f Signed-off-by: Tomas Vanek <vanekt@fbl.cz> Reviewed-on: https://review.openocd.org/c/openocd/+/6765 Tested-by: jenkins Reviewed-by: Tim Newsome <tim@sifive.com>
2022-05-27Give each SMP group a unique number.Tim Newsome2-2/+4
This helps e.g. if there are 8 cores, and cores 0--3 are in one SMP group while 4--7 are in another group. (And there are 2 gdb instances connected, one debugging the first group, and one the second.) Signed-off-by: Tim Newsome <tim@sifive.com> Change-Id: I7b6c9382eadf964529105eaf0411a42d48768668 Reviewed-on: https://review.openocd.org/c/openocd/+/6979 Tested-by: jenkins Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-05-21semihosting: add semihosting_basedir commandErhan Kurubas3-3/+59
This command allows users to set base working directory for the semihosting I/O operations.Default is the current OpenOCD directory. Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com> Change-Id: I80c5979e4c96d66cccdd12cc6fcd5f353e5c6b4d Reviewed-on: https://review.openocd.org/c/openocd/+/6888 Tested-by: jenkins Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-05-21target: fix clang static analyzer warningTomas Vanek1-1/+1
Removes Warning: line 6482, column 12 1st function call argument is an uninitialized value Use target ptr directly as checked in previous lines instead of dereferencing head->target Change-Id: I6804b776fd493af71f3098d702f9cdc7acb50151 Signed-off-by: Tomas Vanek <vanekt@fbl.cz> Reviewed-on: https://review.openocd.org/c/openocd/+/6970 Tested-by: jenkins Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
2022-05-18target/riscv: fix 'reset run' after 'reset halt'Tomas Vanek1-5/+5
'reset halt' does not clear DM_DMCONTROL_HALTREQ at deassert_reset(). If hw reset line is configured e.g. 'reset_config srst_only' the folowing 'reset run' halts: > gd32v.cpu curstate running > reset halt JTAG tap: gd32v.cpu tap/device found: 0x1000563d (mfg: 0x31e ... > gd32v.cpu curstate halted > reset JTAG tap: gd32v.cpu tap/device found: 0x1000563d (mfg: 0x31e ... > gd32v.cpu curstate halted <<<<---- wrong!!! > reset JTAG tap: gd32v.cpu tap/device found: 0x1000563d (mfg: 0x31e ... > gd32v.cpu curstate running Clear DM_DMCONTROL_HALTREQ when acking reset. Change-Id: Iae0454b425e81e64774b9785bb5ba1d4564d940b Signed-off-by: Tomas Vanek <vanekt@fbl.cz> Reviewed-on: https://review.openocd.org/c/openocd/+/6961 Tested-by: jenkins Reviewed-by: Tim Newsome <tim@sifive.com>
2022-05-14aarch64: don't wait for smp targets halted in deassert resetAntonio Borneo1-3/+7
The function target_type::deassert_reset() is called for every target after reset is deasserted. If the target fails to get halted, we log a warning and issue a halt request for the target itself. Current code calls the generic target_halt() that: - extends the halt to all the targets in the SMP node; - waits for targets to halt. This breaks the logic of running target_type::deassert_reset() per target. Plus, waiting for targets to halt delays the call of target_type::deassert_reset() for the next targets. Replace the call to target_halt() with the aarch64 specific function to halt the single target. Pass the parameter HALT_LAZY to prevent the wait for target halted. Similar solution is already implemented for cortex_a. Change-Id: I446dc03cb91524c6d388db485bc2388177af77b6 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6947 Tested-by: jenkins
2022-05-14arm_adi_v5: add support for 64bit Class 0x9 ROM tablesAntonio Borneo1-11/+32
Arm documentation does not explicitly report the order of the two 32bit words that compose the 64bit value. But both ADIv5 and ADIv6 specify that only little-endian is supported (ADIv5.2 obsoletes the big-endian support). This change reads the 64bit value in little-endian. Detect the 64bit content and use it. Change-Id: I723ec099c7e8c70c1f9a568e32ea867fcbf1f1db Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6465 Tested-by: jenkins
2022-05-14arm_adi_v5: replace dap_lookup_cs_component()Antonio Borneo2-76/+85
With the generic function for ROM table walk-through, reimplement dap_lookup_cs_component(). Catch the code CORESIGHT_COMPONENT_FOUND and halt the search. While there, drop two macros in arm_coresight.h, now unused. Change-Id: I589ef7ae8a651d0c422ce7d0b4ed913713a8217e Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6824 Tested-by: jenkins
2022-05-14arm_adi_v5: let dap_lookup_cs_component() to get AP dbgbaseAntonio Borneo4-28/+24
Simplify the code in cortex_a and aarch64 by moving the call to dap_get_debugbase() inside dap_lookup_cs_component(). This has the further effects: - dap_get_debugbase() is not referenced outside arm_adi_v5.c and becomes static; - dap_lookup_cs_component() looses one parameter; - the coreid parameter 'idx' is passed as value; - the caller in aarch64 don't have and don't print the irrelevant value of AP register APID; - fixes the debug message in the caller in aarch64 to print the coreid value instead of always zero. Change-Id: Ic7f0f643fdf067c059c8f2455a02ff18a3fed054 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6823 Tested-by: jenkins
2022-05-14arm_adi_v5: abstract actions during ROM table parsingAntonio Borneo1-30/+123
Now all the actions (build command output) of command "dap info" are decoupled from the ROM walk-through. Pass the actions as a generic parameter to ROM walk-through code. Put as private data every information that is only required by the actions and not by the ROM walk-through. Change-Id: I3b6ad112ea21296458c94aebbf91bf65bf6657a7 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6822 Tested-by: jenkins
2022-05-14arm_adi_v5: separate ROM table parsing from command output [3/3]Antonio Borneo1-17/+41
This change only targets the output of rtp_rom_loop(). Change-Id: If9ac013798923428c3b897a969887e98b6935a2b Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6821 Tested-by: jenkins
2022-05-14arm_adi_v5: separate ROM table parsing from command output [2/3]Antonio Borneo2-65/+107
This change only targets the output of rtp_cs_component(). To easily propagate the coordinates of the CoreSight component, add them in the struct that holds the register values. While there, define a macro for the max depth of ROM tables. Change-Id: I75e5ef4f9419da3192123aebcd61471c2af9374f Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6820 Tested-by: jenkins
2022-05-14arm_adi_v5: separate ROM table parsing from command output [1/3]Antonio Borneo1-6/+44
In OpenOCD arm_adi_v5 we have already two implementations of code for parsing the ADIv5 ROM table: - in the commands "dap info" and "$dap_name info"; - in the function dap_lookup_cs_component(). Adding support for ADIv6 requires extending both implementations. Moreover, current code does not handle few aspects of the ROM parsing, e.g. the "Power Domain IDs". To add such extensions both implementations should be touched. I plan to add a command to parses (again) the ROM table and dump a simple prototype of a configuration script for the target, useful while analysing a new target. Keeping aligned all these implementation would be too complex. With focus to "dap info" command, decouple the part of code to walk-through the ROM table from the code that creates the command output. The idea is to keep a single implementation for the walk-through code, while parametrizing the output code to handle the generation of a configuration script or the result of the function dap_lookup_cs_component(). This change only targets the output of MEM-AP header Further changes will target other parts of the code. While there, add a message if MEM-AP is not accessible. Change-Id: I112f637edfdb8688afb4e631297f6536da9604f1 Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com> Reviewed-on: https://review.openocd.org/c/openocd/+/6819 Tested-by: jenkins
generated by cgit v1.1 at 2024-07-17 23:18:19 +0000