Re: [PATCH v5] x86/sgx: Free backing memory after faulting the enclave page

[Jarkko Sakkinen]

On Wed, Mar 02, 2022 at 08:40:29AM -0800, Dave Hansen wrote:
> On 3/1/22 17:41, Jarkko Sakkinen wrote:
> > On Tue, Mar 01, 2022 at 09:54:14AM -0800, Dave Hansen wrote:
> >> On 3/1/22 04:58, Jarkko Sakkinen wrote:
> >>> @@ -32,14 +58,16 @@ static int __sgx_encl_eldu(struct sgx_encl_page *encl_page,
> >>>  	else
> >>>  		page_index = PFN_DOWN(encl->size);
> >>>  
> >>> +	page_pcmd_off = sgx_encl_get_backing_page_pcmd_offset(encl, page_index);
> >>> +
> >>>  	ret = sgx_encl_lookup_backing(encl, page_index, &b);
> >>>  	if (ret)
> >>>  		return ret;
> >> What tree is this against?  It looks like it might be on top of
> >> Kristen's overcommit series.
> >>
> >> It would be best if you could test this on top of tip/sgx.  Kristen
> >> changed code in this area as well.
> > I rebased this against latest stuff and now I did a sanity check:
> > 
> > $ git fetch tip
> > remote: Enumerating objects: 75, done.
> > remote: Counting objects: 100% (75/75), done.
> > remote: Compressing objects: 100% (12/12), done.
> > remote: Total 77 (delta 65), reused 65 (delta 63), pack-reused 2
> > Unpacking objects: 100% (77/77), 34.07 KiB | 157.00 KiB/s, done.
> >>From git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
> >    161a9a33702a..cedd3614e5d9  perf/core  -> tip/perf/core
> >    6255b48aebfd..25795ef6299f  sched/core -> tip/sched/core
> > 
> > $ git rebase tip/x86/sgx 
> > Current branch master is up to date.
> 
> Are you sure you didn't also rebase all of Kristen's work along with
> your one patch?  How many patches did you rebase?
> 
> You might want to do:
> 
> 	git log tip/x86/sgx..

I did "git reset --hard tip/x86/sgx" and reapplied patch, and fixed the
merge conflict, i.e. "lookup" to "get" :-) You're right.

BR, Jarkko