]>
git.gag.com Git - fw/stlink/log
Jerry Jacobs [Fri, 29 Apr 2016 07:44:22 +0000 (09:44 +0200)]
st-info: Add support for OpenOCD hla_serial printing
texane [Sun, 24 Apr 2016 07:25:51 +0000 (09:25 +0200)]
Fix issue #399: Some Nucleo devices (L476RG) do not return a valid serial
texane [Sat, 16 Apr 2016 06:05:28 +0000 (08:05 +0200)]
Merge pull request #398 from xor-gate/refactoring
Refactoring some code, merge st-probe and fixup #318 serial print
texane [Sat, 16 Apr 2016 06:04:03 +0000 (08:04 +0200)]
Merge pull request #397 from xor-gate/st-probe-libusb-fix
st-probe: Fix hang when user has no libusb access
Jerry Jacobs [Fri, 15 Apr 2016 21:35:11 +0000 (23:35 +0200)]
stlink_open_usb: Fix issue where serial is not cached when serial is not compared in search
Jerry Jacobs [Fri, 15 Apr 2016 21:09:45 +0000 (23:09 +0200)]
* Merge st-probe tool into st-info
* Fixup #318 serial print in stlink probe with trailing zeros
* Refactor stlink-usb.h with some doxygen-style comments
* Refactor some pieces of stlink_open_usb
Jerry Jacobs [Fri, 15 Apr 2016 20:20:22 +0000 (22:20 +0200)]
Fix hang when user has no libusb access because of plugdev under e.g linux.
Add some better sanity checking when using libusb_open and cleanup accordingly.
Fixes hang (debian 8, amd64, libusb 1.0.19):
libusb: warning [add_to_flying_list] failed to arm first timerfd (errno 9)
libusb: warning [libusb_close] internal signalling write failed, closing anyway
texane [Fri, 15 Apr 2016 14:24:56 +0000 (16:24 +0200)]
Merge pull request #396 from xor-gate/cmake-fixes
Add -fPIC compiler flag, remove st-info from libstlink.a
Jerry Jacobs [Fri, 15 Apr 2016 13:53:35 +0000 (15:53 +0200)]
Add -fPIC compiler flag, remove st-info from libstlink.a which is already specified by the st-info tool
texane [Fri, 8 Apr 2016 07:06:30 +0000 (09:06 +0200)]
Merge pull request #395 from halt-hammerzeit/master
Styling
Nikolay [Fri, 8 Apr 2016 00:26:30 +0000 (03:26 +0300)]
driver README minor edits
Nikolay [Fri, 8 Apr 2016 00:16:59 +0000 (03:16 +0300)]
renaming the README file to README.md
Nikolay [Fri, 8 Apr 2016 00:12:48 +0000 (03:12 +0300)]
Beautify + El Capitan confirmation
texane [Thu, 7 Apr 2016 11:58:39 +0000 (13:58 +0200)]
Merge pull request #393 from xor-gate/cmake-better-cflags
cmake: Extend cflags and add Debug build type to be debuggable with gdb
Jerry Jacobs [Thu, 7 Apr 2016 11:01:32 +0000 (13:01 +0200)]
Merge with master
texane [Thu, 7 Apr 2016 10:22:10 +0000 (12:22 +0200)]
Merge pull request #394 from xor-gate/st-probe-busy-segfault
st-probe: Fix segfault when programmer is already busy
texane [Thu, 7 Apr 2016 10:20:44 +0000 (12:20 +0200)]
Merge pull request #392 from xor-gate/travis
travis: Initial continues integration build on linux and osx
Jerry Jacobs [Thu, 7 Apr 2016 10:08:05 +0000 (12:08 +0200)]
st-probe: Fix segfault when programmer is already busy and null pointers are in the list
Jerry Jacobs [Thu, 7 Apr 2016 09:58:16 +0000 (11:58 +0200)]
cmake: Add -ggdb flag when CMAKE_BUILD_TYPE=Debug, add more usefull and stricter compiler flags when supported
Jerry Jacobs [Thu, 7 Apr 2016 08:01:20 +0000 (10:01 +0200)]
travis: Initial continues integration build on linux and osx
* Fixup README to render HTML from Markdown
* Fixup CMakeLists.txt to have a lower working minimum version
* Add .travis.yml and .travis.sh for autobuild
texane [Wed, 6 Apr 2016 15:43:55 +0000 (17:43 +0200)]
Merge pull request #391 from xor-gate/st-probe
st-probe: Initial working stlink_probe_* API and CLI tool
Jerry Jacobs [Wed, 6 Apr 2016 12:35:57 +0000 (14:35 +0200)]
st-probe: Initial working stlink_probe_* API and CLI tool
texane [Sat, 2 Apr 2016 16:52:07 +0000 (18:52 +0200)]
Merge pull request #389 from tobbad/master
Trace the read data in stlink_read_debug32 and not the address of the…
Tobias Badertscher [Sat, 2 Apr 2016 16:18:39 +0000 (18:18 +0200)]
Trace the read data in stlink_read_debug32 and not the address of the variable.
texane [Sat, 2 Apr 2016 14:37:29 +0000 (16:37 +0200)]
Merge pull request #388 from tobbad/master
Proper writing of page 0 of second bank for stm32l476xe.
Tobias Badertscher [Sat, 2 Apr 2016 13:19:27 +0000 (15:19 +0200)]
Proper writing of page 0 of second bank for stm32l476xe.
texane [Sun, 20 Mar 2016 06:19:26 +0000 (07:19 +0100)]
Merge pull request #386 from l29ah/texane-master
avoid reading from NULL
Sergey Alirzaev [Sun, 20 Mar 2016 02:28:07 +0000 (05:28 +0300)]
avoid reading from NULL
texane [Thu, 17 Mar 2016 18:27:14 +0000 (19:27 +0100)]
fix: FLASH_F4_CR_LOCK
texane [Wed, 16 Mar 2016 19:27:37 +0000 (20:27 +0100)]
fix: chip_id now a pointer, dereference
fabien.lementec [Wed, 16 Mar 2016 10:33:57 +0000 (11:33 +0100)]
fix: stlink_version return code
texane [Mon, 14 Mar 2016 16:20:16 +0000 (17:20 +0100)]
Merge pull request #384 from mcoquelin-stm32/cleanups
Cleanups
Maxime Coquelin [Mon, 14 Mar 2016 14:09:22 +0000 (15:09 +0100)]
stlink-common: Ensure flash type is properly declared in device params
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Mon, 14 Mar 2016 13:53:32 +0000 (14:53 +0100)]
stlink-common: Fix STM32L4 loader write count to reflect 64bits granularity
The stm32l4 loader expects a count of 32 bits words while its granularity is
really 64 bits.
This patch fixes this to simplify count calculation in run_flash_loader().
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Mon, 14 Mar 2016 13:22:54 +0000 (14:22 +0100)]
stlink-common: Update STM32L0/1 loaders to return remaining count in r2
All the loaders returns remaining work count in r2, except stm32l0/1.
Make these loaders behaving as the others to simplify run_flash_loader() code.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Mon, 14 Mar 2016 12:32:48 +0000 (13:32 +0100)]
stlink-common: Update STM32L0 and STM32L1 loader ABI
This patch invert source and destination registers in the stm32l0 and stm32l1
loaders, so that it follows the same ABI as other stm32 loaders.
Doing that, the run_flash_loader() function can be simplified a little.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Mon, 14 Mar 2016 12:01:23 +0000 (13:01 +0100)]
stlink-common: Simplify run_flash_loader()
This patch simplifies run_flash_loader() function in preparation for error
propagation from backends.
Doing this, we have less call sites for stlink API.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Fri, 11 Mar 2016 17:51:15 +0000 (18:51 +0100)]
stlink-common: Introduce type of flash controller enum
Having a type of flash information has some advantages:
- Make the code easier to read
- Make adding family derivatives easier (only add a new entry in header file)
- Make the backends error propagation easier to implement (less places to fix)
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Fri, 11 Mar 2016 15:57:34 +0000 (16:57 +0100)]
st-flash: Improve error handling
Now that libusb errors are propagated up to stlink API, we can handle these
errors.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Fri, 11 Mar 2016 14:27:56 +0000 (15:27 +0100)]
stlink-common: Make stlink API propagate backend errors
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Fri, 11 Mar 2016 10:16:41 +0000 (11:16 +0100)]
Make the backends propagate errors
As the libusb returns errors, make the backends propagates them so that
callers can decide to continue or stop task execution.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Thu, 10 Mar 2016 15:39:26 +0000 (16:39 +0100)]
stlink_target_voltage: Check for libusb error
_stlink_usb_target_voltage already returns an error value.
If value return is positive, this is a voltage, if negative this is an error.
Check the return on callers side to inform there is an error in reading the
voltage, instead of notifying of a too low voltage.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Thu, 10 Mar 2016 15:26:48 +0000 (16:26 +0100)]
stlink_core_id: No need to return core_id value
This rework is done in order to prepare for propagating errors returned by
libusb.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
Maxime Coquelin [Thu, 10 Mar 2016 15:15:57 +0000 (16:15 +0100)]
read_debug32: Use a pointer instead of returning the value
This rework is done in order to prepare for propagating errors returned by
libusb.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
texane [Fri, 11 Mar 2016 21:33:06 +0000 (22:33 +0100)]
Merge pull request #383 from xor-gate/cmake-add-st-info
Add st-info to CMakeLists.txt
Jerry Jacobs [Fri, 11 Mar 2016 21:13:09 +0000 (22:13 +0100)]
Add st-info to CMakeLists.txt
texane [Sat, 27 Feb 2016 18:17:19 +0000 (19:17 +0100)]
be quicker about determining erased byte pattern when flashing, more readable code
texane [Sat, 27 Feb 2016 18:12:27 +0000 (19:12 +0100)]
Merge pull request #377 from gtalusan/master
be quicker about determining erased byte pattern when flashing
George Talusan [Sat, 27 Feb 2016 16:41:39 +0000 (11:41 -0500)]
be quicker about determining erased byte pattern when flashing
texane [Tue, 23 Feb 2016 05:52:15 +0000 (06:52 +0100)]
Merge pull request #376 from gtalusan/master
make st-flash clean up on sigint/term/segv, make gdb-server clean up on sigsegv.
George Talusan [Tue, 23 Feb 2016 01:07:43 +0000 (20:07 -0500)]
make st-flash clean up on sigint/term/segv, make gdb-server clean up on sigsegv.
texane [Mon, 22 Feb 2016 17:43:35 +0000 (18:43 +0100)]
Merge pull request #375 from gtalusan/master
make usb backend status check work, remove some dead code
George Talusan [Mon, 22 Feb 2016 16:41:48 +0000 (11:41 -0500)]
make usb backend status check work, remove some dead code
texane [Fri, 19 Feb 2016 17:33:24 +0000 (18:33 +0100)]
Merge pull request #374 from gtalusan/master
use libusb synchronous api
George Talusan [Fri, 19 Feb 2016 16:11:31 +0000 (11:11 -0500)]
use libusb synchronous api
texane [Fri, 19 Feb 2016 06:12:11 +0000 (07:12 +0100)]
Merge pull request #371 from gtalusan/master
remove empty byte truncation from stlink_fread
texane [Fri, 19 Feb 2016 06:10:00 +0000 (07:10 +0100)]
Merge pull request #372 from gtalusan/gdb-bigread
clamp gdb memory reads to 0x1800
George Talusan [Fri, 19 Feb 2016 05:04:44 +0000 (00:04 -0500)]
clamp gdb memory reads to 0x1800
George Talusan [Fri, 19 Feb 2016 03:56:27 +0000 (22:56 -0500)]
remove empty byte truncation from stlink_fread
fabien.lementec [Mon, 15 Feb 2016 13:55:09 +0000 (14:55 +0100)]
Added STM32F051R8T6
texane [Wed, 3 Feb 2016 22:45:52 +0000 (23:45 +0100)]
Merge pull request #370 from brunowonka/STM32L4_flash_fix
Stm32 l4 flash fix
Bruno Dal Bo [Wed, 3 Feb 2016 21:10:54 +0000 (13:10 -0800)]
clean SR flags before programming CR
Bruno Dal Bo [Wed, 3 Feb 2016 21:06:18 +0000 (13:06 -0800)]
fix on stm32l4 to clear flash mass erase flags on CR
fabien.lementec [Mon, 1 Feb 2016 16:02:55 +0000 (17:02 +0100)]
fix signed unsigned comparison
texane [Mon, 1 Feb 2016 16:02:29 +0000 (17:02 +0100)]
Merge pull request #369 from meisteg/stutil_reset
st-util: Fix for connecting to running devices
texane [Mon, 1 Feb 2016 16:01:45 +0000 (17:01 +0100)]
Merge pull request #368 from meisteg/l4_memory_map
Add memory map for STM32L4
texane [Mon, 1 Feb 2016 16:01:13 +0000 (17:01 +0100)]
Merge pull request #367 from meisteg/fix_every_other
Fix issue where "unknown chip id!" is seen every other time
Greg Meiste [Mon, 1 Feb 2016 15:08:59 +0000 (09:08 -0600)]
st-util: Fix for connecting to running devices
Previously, even with the reset enabled, st-util would not connect
to a target if it was running. This change will reset the chip
when opening the connection, which allows st-util to connect
successfully.
Signed-off-by: Greg Meiste <w30289@motorola.com>
Greg Meiste [Wed, 27 May 2015 18:20:07 +0000 (13:20 -0500)]
Add memory map for STM32L4
Signed-off-by: Greg Meiste <w30289@motorola.com>
Greg Meiste [Wed, 27 May 2015 15:12:40 +0000 (10:12 -0500)]
Fix issue where "unknown chip id!" is seen every other time
Previously when running the program, reading the chip version
would fail. Running the program a second time worked. This
sequence is repeated the next time st-flash/st-util is run.
Giving reset 10ms to complete before trying to read the chip
version resolves the issue.
Signed-off-by: Greg Meiste <w30289@motorola.com>
texane [Fri, 29 Jan 2016 15:35:37 +0000 (16:35 +0100)]
Merge pull request #366 from meisteg/l4_voltage
Allow flashing of STM32L4 down to 1.71 V
Greg Meiste [Mon, 19 Oct 2015 14:24:40 +0000 (09:24 -0500)]
Allow flashing of STM32L4 down to 1.71 V
The STM32L4 specification specifies that VDD can be 1.71 - 3.6 V.
Signed-off-by: Greg Meiste <w30289@motorola.com>
texane [Mon, 11 Jan 2016 15:02:48 +0000 (16:02 +0100)]
Merge pull request #363 from gingold-adacore/cache
st-util: synchronize cache for stm32f7
Tristan Gingold [Mon, 11 Jan 2016 14:44:17 +0000 (15:44 +0100)]
st-util: synchronize cache for stm32f7
This patch handles the cache on the cortex-m7 stm32f7 cpu.
When gdb inserts a soft breakpoint (which is the default for
code in RAM), it replaces an instruction with a breakpoint
instruction. But if the caches are enabled, the replacement
may be made only in the D-cache. To reach the I-cache, the
D-cache must be flushed and the I-cache invalidated.
This implementation is coarse: it cleans the whole D-cache
and invalidate the whole I-cache. It is possible to
track which cache lines have to be cleaned and invalidated.
texane [Thu, 17 Dec 2015 11:38:21 +0000 (12:38 +0100)]
Merge pull request #360 from gingold-adacore/stm32f7-regs
stm32f7: also send features to gdb.
texane [Thu, 17 Dec 2015 11:37:50 +0000 (12:37 +0100)]
Merge pull request #359 from gingold-adacore/no-trailing-nul
Do not send a NUL at end of packets to gdb.
Tristan Gingold [Thu, 17 Dec 2015 10:05:00 +0000 (11:05 +0100)]
stm32f7: also send features to gdb.
So that 'extra' registers (msp, psp, ...) are visible to user.
Tristan Gingold [Thu, 17 Dec 2015 10:00:45 +0000 (11:00 +0100)]
Do not send a NUL at end of packets to gdb.
GDB doesn't need it and see the NUL as a junk character
(visible with 'set debug remote 1').
texane [Tue, 15 Dec 2015 16:18:25 +0000 (17:18 +0100)]
Merge pull request #357 from jlbirccyn/patch-1
Added STM32L053R8
Jean-Luc Béchennec [Tue, 15 Dec 2015 14:12:55 +0000 (15:12 +0100)]
Added STM32L053R8
Hello, checked this board with stink. It communicates without flaw with the board:
```
2015-12-15T14:47:03 INFO src/stlink-common.c: Loading device parameters....
2015-12-15T14:47:03 INFO src/stlink-common.c: Device connected is: L0x3 device, id 0x10086417
2015-12-15T14:47:03 INFO src/stlink-common.c: SRAM size: 0x2000 bytes (8 KiB), Flash: 0x10000 bytes (64 KiB) in pages of 128 bytes
2015-12-15T14:47:03 INFO gdbserver/gdb-server.c: Chip ID is
00000417 , Core ID is
0bc11477 .
2015-12-15T14:47:03 INFO gdbserver/gdb-server.c: Target voltage is 3264 mV.
2015-12-15T14:47:03 INFO gdbserver/gdb-server.c: Listening at *:4242...
```
When I launch GDB and connect to the server, I get :
```
2015-12-15T14:51:41 ERROR gdbserver/gdb-server.c: [FP_CTRL] = 0x00000041 expecting 0x00000261
2015-12-15T14:51:41 INFO gdbserver/gdb-server.c: GDB connected.
```
on the st-util side and :
```
Cannot access memory at address 0xffffffff
```
on the GDB side
Reading and writing to memory and registers works however I get
```
Cannot access memory at address 0xffffffff
```
Each time I write to the memory or to a register
texane [Mon, 7 Dec 2015 17:29:58 +0000 (18:29 +0100)]
Merge pull request #354 from gvz/master
Added parameter to specify one stlink v2 of many
Georg von Zengen [Mon, 7 Dec 2015 16:23:15 +0000 (17:23 +0100)]
Added parameter to specify one stlink v2 of many
This adds a parameter to the function stlink_open_usb and to the binary
st-flash to specify one of multiple connected stlinks.
As the identifier the iSerial of the stlink is used.
If no serial is given the function and binary behave as before.
texane [Fri, 20 Nov 2015 16:53:56 +0000 (17:53 +0100)]
Merge pull request #350 from cheshirekow/cmake_build
Cmake build
Josh Bialkowski [Fri, 20 Nov 2015 16:09:12 +0000 (08:09 -0800)]
Merge branch 'master' into cmake_build
Josh Bialkowski [Thu, 19 Nov 2015 17:29:27 +0000 (09:29 -0800)]
fix MINGW test in cmake
texane [Thu, 19 Nov 2015 10:00:20 +0000 (11:00 +0100)]
Merge pull request #349 from ibobalo/master
mingw compilation error (sleep) fixed
bob [Thu, 19 Nov 2015 08:53:21 +0000 (10:53 +0200)]
mingw compilation error (sleep) fixed
texane [Tue, 17 Nov 2015 16:32:39 +0000 (17:32 +0100)]
Merge pull request #348 from cheshirekow/cmake_build
first stab at a cmake build
Josh Bialkowski [Tue, 17 Nov 2015 00:50:08 +0000 (16:50 -0800)]
first stab at a cmake build
texane [Mon, 9 Nov 2015 18:15:13 +0000 (19:15 +0100)]
Merge pull request #347 from jauler/master
Implement simple non-blocking ring buffer for communication over stlink
Rytis Karpuska [Mon, 9 Nov 2015 18:03:42 +0000 (20:03 +0200)]
Implement simple non-blocking ring buffer for communication over stlink
texane [Sat, 24 Oct 2015 05:04:21 +0000 (07:04 +0200)]
Merge pull request #345 from mcoquelin-stm32/stm32f469
Add support to STM32F469/STM32F479
Maxime Coquelin [Fri, 23 Oct 2015 21:14:59 +0000 (23:14 +0200)]
Add support to STM32F469/STM32F479
STM32F469 is a STM32F429 MCU with Dual-QSPI NOR support and DSI screen
support.
Signed-off-by: Maxime Coquelin <mcoquelin.stm32@gmail.com>
lementec [Mon, 12 Oct 2015 06:07:59 +0000 (08:07 +0200)]
README, add working target
texane [Sat, 3 Oct 2015 15:20:35 +0000 (17:20 +0200)]
Merge pull request #341 from mlu/master
Breakpoint handling for Cortex M7
mlundinse [Sat, 3 Oct 2015 14:36:37 +0000 (16:36 +0200)]
Breakpoint handling for Cortex M7
texane [Thu, 3 Sep 2015 08:45:36 +0000 (10:45 +0200)]
Merge pull request #339 from b1y1k/master
correct error message is given upon missing pkg-config package.
Efe Can İçöz [Thu, 3 Sep 2015 08:32:12 +0000 (11:32 +0300)]
correct error message is given upon missing pkg-config package. fixes texane/stlink#338
texane [Fri, 21 Aug 2015 13:46:05 +0000 (15:46 +0200)]
Merge pull request #336 from mlu/revert-328-master
Revert "Make _stlink_usb_reset use hardreset"
mlu [Fri, 21 Aug 2015 13:44:34 +0000 (15:44 +0200)]
Revert "Make _stlink_usb_reset use hardreset"