resides above the last non-0xff data value in the address space. Only
do this for flash memory since writing a 0xff to flash is a no-op.
This has the affect of creating smaller output files when dumping
memory contents from flash if the program in flash does not consume
the whole memory space. It also results in shorter programming times
when avrdude is asked to load a file into flash that has lots of 0xff
filled data past the last non-0xff data value.
I think this is basically where Alexey was going with his s-record
routine, but this should have a similar affect for all the I/O
routines. The main difference is that Alexey's also optimized 0xff
from the beginning of the address space and was not limited to flash.
I think that these optimizations should be limited to the flash since
it is currently the only memory that treats 0xff as special.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@331 81a1dc3b-b13d-400b-aceb-764788c761c2
output routines (don't optimize away 0xff data before and after non
0xff data). Also, fix a bug where the data contents sometimes weren't
written out completely.
Initial bug reported by Tom Harris <TomH@optiscan.com>. Fixes
provided by Alexey V.Levdikov <tsar@kemford.com>.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@330 81a1dc3b-b13d-400b-aceb-764788c761c2
main.c : print out '<stdin>' and '<stdout>' instead of '-' when using
stdio for I/O.
Thanks to Francisco T. A. Silva <ftas@geodigitus.com.br> for catching
this, and the error fixed by the previous commit as well.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@325 81a1dc3b-b13d-400b-aceb-764788c761c2
specify byte values on the command line instead of via a file. This
can be good for specifying fuse bytes and eliminates the need to
create single-byte files or using interactive terminal mode for these
single-byte memories. Requested by several folks on the mailing list.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@317 81a1dc3b-b13d-400b-aceb-764788c761c2
* .cvsignore: Ignore autoconf files.
* AUTHORS: New file.
* ChangeLog: New file.
* Makefile: Removed file.
* Makefile.am: New file.
* NEWS: New file.
* README: New file.
* bootstrap: New file.
* configure.ac: New file.
* avr.c: Include ac_cfg.h (generated by autoconf).
* config.c: Include ac_cfg.h.
Include config_gram.h instead of y.tab.h.
* config.h: If HAS_YYSTYPE is not defined, define YYSTYPE.
* config_gram.y: Include ac_cfg.h.
* fileio.c: Include ac_cfg.h.
* lexer.l: Include config_gram.h instead of y.tab.h.
* lists.c: Include ac_cfg.h.
* main.c: Include ac_cfg.h.
* par.c: Include ac_cfg.h.
* pgm.c: Include ac_cfg.h.
* ppi.c: Include ac_cfg.h.
* stk500.c: Include ac_cfg.h.
* term.c: Include ac_cfg.h.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@192 81a1dc3b-b13d-400b-aceb-764788c761c2
This was intended to be used for identifying code in the field for
incoming bug reports, but I've never really found it all that useful.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@171 81a1dc3b-b13d-400b-aceb-764788c761c2
undergone. This utilizes the last two bytes of EEPROM to maintain a
counter that is incremented each time the part is erased.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@138 81a1dc3b-b13d-400b-aceb-764788c761c2
serial programming instructions are not very orthoganal, i.e., the
"read fuse bits" instruction on an ATMega103 is an entirely different
opcode and data format from the _same_ instruction for an ATMega163!
Thus, it becomes impossible to have a single instruction encoding
(varying the data) across the chip lines.
This set of changes allows and requires instruction encodings to be
defined on a per-part basis within the configuration file. Hopefully
I've defined the encoding scheme in a general enough way so it is
useful in describing the instruction formats for yet-to-be invented
Atmel chips. I've tried hard to make it match very closely with the
specification in Atmel's data sheets for their parts. It's a little
more verbose than what I initially hoped for, but I've tried to keep
it as concise as I could, while still remaining reasonably flexible.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@100 81a1dc3b-b13d-400b-aceb-764788c761c2
has a 128K flash.
Due to the bank addressing required, interactive update of the flash
is not supported, though the eeprom can be updated interactively.
Both memories can be programmed via non-interactive mode.
Intel Hex Record type '04' is now generated as required for outputing
memory contents that go beyond 64K.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@78 81a1dc3b-b13d-400b-aceb-764788c761c2
file. This makes supporting other programmers much easier.
Rename AVRprog.pdf to avrprog.pdf.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@67 81a1dc3b-b13d-400b-aceb-764788c761c2
component module. This is intended for support purposes, so that I
can tell unambiguously what version a binary out in the field is.
Additionally, display a revision timestamp along with the version
number. This also is intended for aiding in support and is the Unix
time of the latest component module. Having this, should allow me to
do a "cvs co -D timestamp avrprog" and get exactly the source of the
version that is being reported.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@55 81a1dc3b-b13d-400b-aceb-764788c761c2
bytes written. The presence of an Intel Hex address record can cause
these two number to be different; but the callers of this routine need
the former.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@54 81a1dc3b-b13d-400b-aceb-764788c761c2
fileio.c: Properly handle all the Intel Hex record types that I can
find information about.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@52 81a1dc3b-b13d-400b-aceb-764788c761c2
prompted.
avr.c avr.h fileio.c term.c :
Change the avrpart data structure so that the typedef AVRMEM is
used as an index into an array for the sizes of the memory types
and also for pointers to buffers that represent the chip data for
that memory type. This removes a lot of conditional code of the
form:
switch (memtype) {
case AVR_FLASH :
...
}
Also, re-code avr_read_byte() and avr_write_byte() to properly
handle the flash memory type without having to tell them whether
they should program the high byte or the low byte - figure that
out from the address itself. For flash memory type, these
routines now take the actual byte address instead of the word
address. This _greatly_ simplifies many otherwise simple
operations, such a reading or writing a range of memory, by not
having to worry about whether the address starts on an odd byte
or an even byte.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@45 81a1dc3b-b13d-400b-aceb-764788c761c2
only via main() so that the exitspecs are properly applied.
When reading input data from a file, remember how many bytes were read
and write and verify only that many bytes.
Don't complain when an input file size is smaller than the memory size
we are programming. This is normal.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@44 81a1dc3b-b13d-400b-aceb-764788c761c2
record type was causing non-zero record types to be calculated
incorrectly.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@43 81a1dc3b-b13d-400b-aceb-764788c761c2
more modular pieces.
Also, accept command abbreviations as long as they are not ambiguous.
git-svn-id: svn://svn.savannah.nongnu.org/avrdude/trunk/avrdude@38 81a1dc3b-b13d-400b-aceb-764788c761c2
kiwi64ajs
arcanum
bdean
troth
bsd