state of the testsuite
When running the testsuite on Debian -testing I constanyly get the 12 failing testcases (016 041 049 064 071 082 084 104 111 136 140 166). Is this expected or are other people seeing better results?
Quoted 3 lines · Christoph Hellwig
When running the testsuite on Debian -testing I constanyly get the 12 failing testcases (016 041 049 064 071 082 084 104 111 136 140 166). Is this expected or are other people seeing better results?
What kernel+platform and what version of the tests?
Prior to 2.6.22-rc2, 016 and 144 are the only ones i saw failing regularly. 166 will fail until we get ->page_mkwrite sorted out (but passes in my tree).
With 2.6.22-rc2, all the loopback based tests are failing because someone broke the loopback device so lots of tests fail because of that. The other tests I haven't seen fail for some time...
Cheers,
Dave.
Signature · 3 lines
Dave Chinner Principal Engineer SGI Australian Software Group
Quoted 10 lines · David Chinner
Quoted 3 lines · Christoph Hellwig
When running the testsuite on Debian -testing I constanyly get the 12 failing testcases (016 041 049 064 071 082 084 104 111 136 140 166). Is this expected or are other people seeing better results?
What kernel+platform and what version of the tests?
Prior to 2.6.22-rc2, 016 and 144 are the only ones i saw failing regularly. 166 will fail until we get ->page_mkwrite sorted out (but passes in my tree).
looks like 082 and 088 have been failing for a while in the automated QA as well but there's not a lot of other regular failures....
Cheers,
Dave.
Signature · 3 lines
Dave Chinner Principal Engineer SGI Australian Software Group
Quoted 6 lines · David Chinner
Quoted 3 lines · Christoph Hellwig
When running the testsuite on Debian -testing I constanyly get the 12 failing testcases (016 041 049 064 071 082 084 104 111 136 140 166). Is this expected or are other people seeing better results?
What kernel+platform and what version of the tests?
Current mainline on i386 (running in qemu)
Quoted 4 lines
With 2.6.22-rc2, all the loopback based tests are failing because someone broke the loopback device so lots of tests fail because of that. The other tests I haven't seen fail for some time...
Ok, that should account for the loop stuff. I already suspected Ken's unlimited number of loop devices patch breaks something.
Quoted 4 lines · David Chinner
With 2.6.22-rc2, all the loopback based tests are failing because someone broke the loopback device so lots of tests fail because of that. The other tests I haven't seen fail for some time...
Backing out the dynamic loop device allocations gets me 049 back to pass.