On 5/30/2024 12:37, Jani Nikula wrote:
On Thu, 30 May 2024, John Harrison <john.c.harri...@intel.com> wrote:
Except that the whole reason this was started was because Michal
absolutely refuses to allow line counted output in a sysfs/debugfs
file because "it is unnecessary and breaks the decoding tools".
I'm only looking at this patch at face value, it seems useful to me, but
I'm not aware of any deeper context. And judging by the tone of the
discussion, I also don't want to get into it.
That said, might be prudent to post a patch using the line printer along
with its addition.
See:
https://patchwork.freedesktop.org/series/134695/
Patches #5 and #6 make use of the line printer. Although as mentioned in
that patch set, the usage there is still just a stop-gap measure and
both those usages are intended to be replaced once the issues with the
devcoredump version are resolved. The ultimate aim would be to use the
line printer from just one function which dumps a devcoredump capture to
dmesg. That function can then be used from the dead CT code and any
other random debug effort a random developer is working on locally.
John.
Over and out.
BR,
Jani.
