On Wed Jun 5, 2024 at 6:33 PM EEST, Haitao Huang wrote:
> sgx_cgroup_try_charge() expects sgx_cg_wq, so it would break unless we
> check and return 0 which was the initially implemented in v12. But then
> Kai had some concern on that we expose all the interface files to allow
> user to set limits but we don't enforce. To keep it simple we settled down
~~~~~~~~~~~~~~
Sure:
"Keep it simple and corpse"
> back to BUG_ON(). This would only happen rarely and user can add
> command-line to disable SGX if s/he really wants to start kernel in this
> case, just can't do SGX.
Even disabling all of SGX would be a less catastrophical measure.
> Yes I had a comment but Kai thought it was too obvious and I can't think
> of a better one that's not obvious so I removed:
Not great advice given. Please just document it. In patch, which
BUG_ON() I don't want to see my R-by in it, until I've reviewed an
updated version.
BR, Jarkko
