in memory type calculation for Xmega "boot" memory region.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@1087 81a1dc3b-b13d-400b-aceb-764788c761c2
firmware >= V7.x)
* jtagmkII.c: Add firmware-version dependent handling of Xmega parameters.
V7.x firmware expects the NVM offsets being specified through the Xmega
parameters command, but left out as part of the memory address itself.
* jtagmkII_private.h: Add CMND_SET_XMEGA_PARAMS, and struct xmega_device_desc.
* config_gram.y: Add mcu_base keyword.
* avrpart.h: (Dito.)
* lexer.l: (Dito.)
* avrdude.conf.in (.xmega): add mcu_base, and data memory segment.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@1078 81a1dc3b-b13d-400b-aceb-764788c761c2
firmware < V7.x)
* jtagmkII.c: When going to write to the boot section of flash,
use MTYPE_BOOT_FLASH rather than MTYPE_FLASH
* jtagmkII_private.h: add MTYPE_BOOT_FLASH constant
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@1077 81a1dc3b-b13d-400b-aceb-764788c761c2
"GO" code in again, but only activate it when programming an
Xmega device through PDI.
Wonderfully consistent tools.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@1071 81a1dc3b-b13d-400b-aceb-764788c761c2
turned out to do more harm than good; on the Xmega, it caused exactly
the opposite reaction, the CPU stopped. Even without it, after the
RESET command, the target is running already. So, drop the GO.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@1070 81a1dc3b-b13d-400b-aceb-764788c761c2
will update only the table in pgm_type.c.
* config_gram.y, lexer.l: removed programmer type keywords,
use now locate_programmer_type() function
* pgm_type.[ch]: added new files for table of programmer types
* main.c: allow list of programmer types by -c ?type
* avrdude.conf.in: changed all type keywords to quoted strings
* doc/avrdude.texi: changed description of type definition, list
of valid types is now included from generated file
* doc/Makefile.am: generate list of programmer types for doc
* all programmers [hc]: add xxx_desc string for description of programmer
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@1051 81a1dc3b-b13d-400b-aceb-764788c761c2
Each memory image byte is now tagged as it's being read from a file.
Only bytes read from a file will be written or verified (modulo page
granularity requirements).
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@1007 81a1dc3b-b13d-400b-aceb-764788c761c2
* jtagII.c (jtagmkII_getsync): Instead of exit()ing from
deep within the tree when detecting the "need debugWIRE"
situation, properly pass this up as a return code.
* jtagII_private.h (JTAGII_GETSYNC_FAIL_GRACEFUL): New constant.
* stk500v2.c (stk500v2_jtagmkII_open): Don't tell anything
anymore when receiving a JTAGII_GETSYNC_FAIL_GRACEFUL from
jtagmkII_getsync(); silently give up (all necessary has been
said already).
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@983 81a1dc3b-b13d-400b-aceb-764788c761c2
bug #30268: Debugwire broken in avrdude-5.10
* jtagmkII.c (jtagmkII_initialize): only try setting up a JTAG chain when
the programmer is using JTAG.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@973 81a1dc3b-b13d-400b-aceb-764788c761c2
timeout should depend on the selected clock speed, but changing the
API for that would require quite some effort throughout the entire
code base.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@954 81a1dc3b-b13d-400b-aceb-764788c761c2
external reset applied (in case the target is in sleep mode or has
asserted the JTD bit).
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@926 81a1dc3b-b13d-400b-aceb-764788c761c2
jtagmkII_private.h, drop all #if 0 blocks, fold overly long lines,
move the *_initpgm() functions to the end of the file; while being
here, remove all trailing whitespace.
* jtagmkII_private.h: move AVR32 #defines here.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@909 81a1dc3b-b13d-400b-aceb-764788c761c2
subsequent programming operation will fail (for unknown reasons).
Actually, this was really only required for ancient AVRs, but doesn't
hurt on mega and tiny devices.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@904 81a1dc3b-b13d-400b-aceb-764788c761c2
* jtagmkII_private.h: Add CMND_XMEGA_ERASE as well as
the various XMEGA_ERASE_* definitions (from updated
appnote AVR067)
* jtagmkII.c (jtagmkII_chip_erase): Correctly implement
Xmega chip erase based on CMND_XMEGA_ERASE.
* jtagmkII.c (jtagmkII_pre_write): Remove, this turned out
to be just a chip erase.
* jtagmkII.c (jtagmkII_program_disable): Don't try reading
"hfuse" for Xmega parts; they don't have it.
* main.c (main): Re-enable auto-erase. It's been done
before (as "jtagmkII_pre_write") in jtagmkII_paged_write()
anyway. Xmega boot and application flash areas should be
handled separately in the future, so auto_erase can only
affect the area just being programmed.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@903 81a1dc3b-b13d-400b-aceb-764788c761c2
by recent GCC versions).
* config_gram.y (parse_cmdbits): "brkt possibly used uninitialized"
(GCC errs here)
* jtagmkII.c (jtagmkII_reset32): "status possibly used uninitialized"
(I think GCC errs, too)
* buspirate.c: "pointers differ in signedness" (mismatch between
string processing and the use of "unsigned char" throughought the
AVRDUDE API)
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@882 81a1dc3b-b13d-400b-aceb-764788c761c2
* avr.h: Remove stray semicolon.
* configure.ac: Add check for predefined types uint_t and ulong_t.
* confwin.c: Include "avrdude.h" on top to avoid empty translation
unit warning.
* ppwin.c: (Ditto.)
* ser_win32.c: (Ditto.)
* serbb_win32.c: (Ditto.)
* jtagmkII.c (jtagmkII_recv): remove unreachable "return".
* stk500.c (stk500_initialize): (Ditto.)
* par.c: Test for both, __sun__ and __sun to see whether we are
being compiled on Solaris.
* ppi.c: (Ditto.)
* stk500v2.c: Implement the DEBUG and DEBUGRECV macros in a way
that is compatible with the ISO C99 standard.
* usbtiny.c: Only typedef uint_t and ulong_t if they have not
been found already by the autoconf checks.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@804 81a1dc3b-b13d-400b-aceb-764788c761c2
patch #6261: avrdude won't use dragon/debugwire to write a file
to eeprom
* jtagmkII.c (jtagmkII_paged_write): when in debugWire mode,
implement a paged write to EEPROM as a series of byte writes.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@792 81a1dc3b-b13d-400b-aceb-764788c761c2
mkII (or AVR Dragon) in ISP mode. The wrong set of per-programmer
private data had been allocated (stk500v2 vs. jtagmkII) which was too
small to hold the actual data.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@781 81a1dc3b-b13d-400b-aceb-764788c761c2
for Xmega programming using the JTAG ICE mkII in JTAG mode.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@780 81a1dc3b-b13d-400b-aceb-764788c761c2
definition. If present, call the setup hook immediately after finding
the respective programmer object, and schedule the teardown hook to be
called upon exit. This allows the programmer implementation to
dynamically allocate private programmer data.
avr910.c, butterfly.c, jtagmkI.c, jtagmkII.c, stk500v2.c, usbasp.c,
usbtiny.c: Convert static programmer data into dynamically allocated
data.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@764 81a1dc3b-b13d-400b-aceb-764788c761c2
the programmer backend.
jtagmkII.c: Implement the extended parameter jtagchain=
to support JTAG daisy-chains.
avrdude.1, doc/avrdude.texi: Document all of the above.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@761 81a1dc3b-b13d-400b-aceb-764788c761c2
own (duplicate) copy of it.
other files: Constify char pointers.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@725 81a1dc3b-b13d-400b-aceb-764788c761c2
declarations in each file by a central header file "avrdude.h".
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@721 81a1dc3b-b13d-400b-aceb-764788c761c2
usleep(1000000) that accidentally crept in in rev 1.19.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@702 81a1dc3b-b13d-400b-aceb-764788c761c2
- declare a dummy "struct timezone" for some Win32 systems (MinGW)
- fix a few printf() argument types
- get rid off the prevailing "all filedescriptors are of type int"
attitude
The last item required a large sweep across the code, in order to
replace all "int fd"s by "struct filedescriptor *fd"s, and pass
pointers (as we cannot pass a union directly). In return, the
code is now supposed to be fully 64-bit safe, rather than relying
on a 64-bit pointer being converted to a (32-bit) int and back
to a pointer as we did previously.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@694 81a1dc3b-b13d-400b-aceb-764788c761c2
upon failure now that the upper layers won't fall back to the cmd()
method anymore in that case.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@691 81a1dc3b-b13d-400b-aceb-764788c761c2
imposed by debugWire itself, so effectively, only flash ROM can be
read and written.
Currently, the required changes to avrdude.conf.in are only present
for the ATtiny44.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@688 81a1dc3b-b13d-400b-aceb-764788c761c2
avr_read_byte_default() and avr_write_byte_default (resp.) by directly
calling the latter functions from within all programmers that don't
implement their own read_byte()/write_byte() methods. In turn, make the
read_byte() and write_byte() methods mandatory, and the cmd() method
(direct ISP command) optional instead (it's effectively mandatory for
any programmer using avr_read_byte_default()/avr_write_byte_default()
though). Remove all the pointless cmd() method stubs from those programmers
that don't need it.
Eliminate avr_read_byte() as it was now completely identical to
pgm->read_byte().
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@684 81a1dc3b-b13d-400b-aceb-764788c761c2
that indicates whether the underlying communication can dynamically
change its speed or not. This flag is set for true serial
communication but clear for USB communication. Don't try to adjust
the speed when talking over a communication channel that doesn't
support it. (The Dragon does not even support the respective
parameter.)
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@676 81a1dc3b-b13d-400b-aceb-764788c761c2
Joerg Wunsch
Rene Liebscher
Nils Springob
David Hoerl