This is the patch [1] to address the issue on a general level,
but it had a suggestion to be done at the driver level, which
may or may not interfere with the patch being applied -- both
approaches are not mutually exclusive.
I'm working on the driver-level patch for bcache as well (it's
a bit more complex), and should provide you another test kernel.
Hopefully you may be able to test this one too; I can reproduce
the bug now, but feedback from real reporters is always welcome.
...
Ryan, yes, I noticed you had filed up an upstream bug (thanks!)
I should post on it with the new patch approach.
Sebastian, BTW, when you have a chance, could you please upload
the output of this command? Should be safe from sensitive data.
Hi Ryan and Sebastian,
Thanks for testing! Glad it helps.
This is the patch [1] to address the issue on a general level,
but it had a suggestion to be done at the driver level, which
may or may not interfere with the patch being applied -- both
approaches are not mutually exclusive.
I'm working on the driver-level patch for bcache as well (it's
a bit more complex), and should provide you another test kernel.
Hopefully you may be able to test this one too; I can reproduce
the bug now, but feedback from real reporters is always welcome.
...
Ryan, yes, I noticed you had filed up an upstream bug (thanks!)
I should post on it with the new patch approach.
Sebastian, BTW, when you have a chance, could you please upload
the output of this command? Should be safe from sensitive data.
$ sudo grep ^ /sys/block/ */queue/ *_block_ size > lp1867916- queue_block_ size.seb 2>&1
cheers,
Mauricio
[1] https:/ /lore.kernel. org<email address hidden>/T/#t