lzma data error on a seemingly problematic squashfs filesystem

[brainstorm]

I'm seeing a lot of errors while running backhand(-cli) against a Flash memory I dumped from a home router:

% cargo run --bin unsquashfs-backhand squashfs_part1.sqsh
(...)
          Failed squashfs-root/usr/lib/libiptext6.so : lzma data error
          Failed squashfs-root/usr/lib/libmmdbdtests.so : lzma data error
          Failed squashfs-root/usr/lib/libwlctl.so : lzma data error
          Failed squashfs-root/usr/lib/libiqctl.so : lzma data error
          Failed squashfs-root/usr/lib/libmmdevcallstate.so : lzma data error
          Failed squashfs-root/usr/lib/lua/socket/url.lua : lzma data error
(...)

While a lot of files decompress successfully and their contents are legit, a lot of other files are 0-byte sized, i.e:

% ls -alh squashfs-root/bin
total 0
drwxr-xr-x 60 rvalls staff 1.9K Aug 23  2017 .
drwxr-xr-x 20 rvalls staff  640 Sep 25  2017 ..
lrwxr-xr-x  1 rvalls staff    7 Sep 25  2017 ash -> busybox
lrwxr-xr-x  1 rvalls staff    7 Sep 25  2017 base64 -> busybox
-rw-r--r--  1 rvalls staff    0 Dec  5 15:30 board_detect
-rw-r--r--  1 rvalls staff    0 Dec  5 15:30 busybox
lrwxr-xr-x  1 rvalls staff    7 Sep 25  2017 cat -> busybox

I've tried other tools before such as unsquashfs, sasquatch and sqsh-tools, but none of them handle this filesystem particularly well, perhaps yours yields the most output though...

Here's the full context and how to obtain (and slice) the aforementioned squashfs_part1.sqsh file if you are curious: Gottox/sqsh-tools#298

[wcampbell0x2a]

You can use RUST_LOG=trace to see even more output..

Does this device boot? If so, you need to read the datasheet to fixup from the oob regions.. I think some readers have knowledge about the regions, but they can be costly.

You can at least see if all the corrupted blocks are in one region!

[brainstorm]

Yes, it does boot! I used my FlashCat to dump the firmware... I'll see if there's some guidance about OOB there too.

Thanks for the suggestions! I'll read the datasheet (closest to actual model Micron_MT29F1G08ABAEA).

I also ran the following:

$ export NO_COLOR=1
$ rm -rf squashfs-root && RUST_LOG=trace cargo run --bin unsquashfs-backhand squashfs_part1.sqsh >& trace.log && zip trace.log.zip trace.log

Let me know if you spot something interesting in it, I'll take a peek at it too: trace.log.zip