Hi: While evaluating file system alternatives, I ran some tests of the dde_rump file system code (Genode 17.05 master/HEAD on Ubuntu 16.04 / x86_64).
The EXT2 and FAT file system tests work as expected (although the latter benefited from a <default caps="256">).
The ISO file system however left me a puzzle (trace below; default caps line added to run file.) The same results appear under builds x86_{32,64} x KERNEL={linux,nova}.
Any ideas regarding what might be going wrong? The prepared test image fs.iso appears to have the intended file "test.txt" on it.
Best, Steve
make run/rump_iso ... [init] child "rom_blk" announces service "Block" [init] child "timer" announces service "Timer" [init -> rom_blk] Using file=fs.iso as device with block size 2048. [init -> rump_fs] Using cd9660 as file system [init -> rump_fs] RUMP ver: 17 [init -> rump_fs] RUMP_THREADS [init -> rump_fs] RUMP_VERBOSE [init -> rump_fs] _RUMPUSER_NCPU [init -> rump_fs] RUMP_MEMLIMIT [init -> rump_fs] asserting rump kernel 4064 KB of RAM [init -> rump_fs] BOOTSTRAP [init -> rump_fs] RUMP_NVNODES [init -> rump_fs] RUMP_BLKFAIL [init -> rump_fs] RUMP_BLKSECTSHIFT [init -> rump_fs] RUMP_MODULEBASE [init -> rump_fs] _RUMPUSER_HOSTNAME [init] child "rump_fs" announces service "File_system" [init] child "fs_rom" announces service "ROM" [init -> fs_rom] request for test-iso [init -> fs_rom] Error: /test-iso: lookup_failed [init -> fs_rom] request for ld.lib.so [init -> fs_rom] Error: /ld.lib.so: lookup_failed [init -> fs_rom] Error: /ld.lib.so: lookup_failed [init -> fs_rom] Error: /ld.lib.so: lookup_failed [init -> fs_rom] Error: /test-iso: lookup_failed [init -> fs_rom] Error: /test-iso: lookup_failed ... (timeout)
Hi Steve,
On 26.07.2017 02:24, Steven Harp wrote:
The ISO file system however left me a puzzle (trace below; default caps line added to run file.) The same results appear under builds x86_{32,64} x KERNEL={linux,nova}.
Any ideas regarding what might be going wrong? The prepared test image fs.iso appears to have the intended file "test.txt" on it.
thank you very much for reporting the issue. We indeed missed to adjust the run script to the changes of the routing of the environment sessions introduced earlier this year. I just fixed the problem with commit [1] on the staging branch. May you try cherry-picking this commit?
[1] https://github.com/genodelabs/genode/commit/89642795a00a1ed3f404cb4a2e75a1ec...
Cheers Norman
On 7/26/17 1:58 AM, Norman Feske wrote:
thank you very much for reporting the issue. We indeed missed to adjust the run script to the changes of the routing of the environment sessions introduced earlier this year. I just fixed the problem with commit [1] on the staging branch. May you try cherry-picking this commit?
[1] https://github.com/genodelabs/genode/commit/89642795a00a1ed3f404cb4a2e75a1ec...
Thanks for the quick (and instructive) patch -- tested and showing full success here.
// Steve