Closed kimstik closed 2 years ago
Very Nice! I'll test this branch before merge. Thinking about create a "standalone" option in the makefile to supress the lib directory if needed.
Bug: Not receiving the CMD_ACK after software reset request (CMD_EOT), dont know why. Aparently handshake is ok, flash program is ok. Tested with 2 SDCC versions: 3.91 footprint 754 4.1 footprint 723
Bug found: tx function not sending the ACK before software reset. "while(TI==0);" should go to the last line.
I believe that "while(TI==0);" placed correctly This way is more correct and secure. Plus it gives better performance.
May you test last one before merging?
I believe that "while(TI==0);" placed correctly This way is more correct and secure. Plus it gives better performance.
Yes, your way is better.
May you test last one before merging?
Of course, Do you mind if I add your name to the source header as co-author?
Could not make those last 2 commits to work. They simply don't boot. (SDCC 4.1)
Could not make those last 2 commits to work. They simply don't boot. (SDCC 4.1)
Really? No any answer for CMD_SOH ? I see that erase screwed in last commit. is 4ff1bf4 commit works?
Last working version: [7bfc15f] WHITOUT the modified peep file. I'll do some more testing as soon as possible.
Of course, Do you mind if I add your name to the source header as co-author?
I hesitate to change the header. You can add me as a contributor.
Of course, Do you mind if I add your name to the source header as co-author?
I hesitate to change the header. You can add me as a contributor.
In my mind, with the work you have done so far, this bootloader is much yours as it is mine. Please, feel free to change the header as you see fit.
footprint is 727 bytes (instead of 1389 from last iteration) get rid of useless runtime 32bit divisions FIXME: makefile was screwed for non-bootloader apps TODO: lets pack to 512?