From: tosiara <to...@gm...> - 2018-11-12 10:03:29
|
Hi, when I was trying to flash the spi through jtag, I was writing 256 bytes blocks and immediately reading them to verify - data never changed. I didn't see any errors, but the statement on Openwrt that WR1043ND JTAG has no write access to the flash forced me to go further and de-solder the chip. Now, since you asked me to look once again, and I know for sure that the router is alive and there is a valid bootloader, I tried to dump it and verify: dump_image 20181112_uboot.bin 0x00000000 0x100 This returns garbage I tried to use a command from Openwrt instructions: dump_image 20181112_uboot.bin 0x9f000000 0x100 And it does return a valid bootloader's header (first 256 bytes) Unfortunately, I'm pretty new to openocd and jtag stuff, and excuse me my question, but I'm now completely puzzled: why the offset for read and write is different? Openwrt suggests reading the bootloader from 0x9f000000, but writing from 0xbf000000 (or as pre previous email 0x00000000). I tried to flash a modified U-Boot at 0x00000000, re-read it from 0x9f000000- and now I see that my changes have been successfully written I did a smaller test with flashing just 256 bytes: dump_image 20181112_uboot_orig.bin 0x9f017400 0x100 **** modify 20181112_uboot_orig.bin and copy it to 20181112_uboot_alt.bin**** flash write_image unlock 20181112_uboot_alt.bin 0x00017400 dump_image 20181112_uboot_after_check.bin 0x9f017400 0x100 **** changes confirmed **** So this was my mistake and I hope my analysis will help someone in the future Thanks! On Fri, Nov 9, 2018 at 5:33 PM Paul Fertser <fer...@gm...> wrote: > Hello, > > On Fri, Nov 09, 2018 at 01:53:15PM +0200, tosiara wrote: > > I found an OpenWRT page saying that "At the moment it is NOT possible to > UNBRICK > > the router by JTAG (no write access to flashrom)"- > > [1]https://openwrt.org/toh/tp-link/tl-wr1043nd > > Your log suggested otherwise. The flash was detected and it looks like > it was written. > > Have you tried verifing it? > > Is it possible you connect it to jtag again and do more > experimentation to fix the underlying issue? I can try assisting with > it in a more interactive way via IRC. > > -- > Be free, use free (http://www.gnu.org/philosophy/free-sw.html) software! > mailto:fer...@gm... > |