Hi Julien, > Let me start with some process. In general, we expect multi-patch series > to have a cover letter.
Sorry, my bad. I'll keep in mind while submitting v3. > > The second one is addressed by using IPA-based TLBI (IPAS2E1) in > > combination with VMALLE1 whenever the IPA range is known instead of > > using VMALLS12E1. There is an upper cap placed on number of IPA-based > > TLBI. This factor for execution time of VMALLS12E1 vs IPAS2E1 is > > found to be 70K on Graviton4. > > Is this running Xen on baremetal or in nested virt? Yes, this is with Xen running in nested virtualization. I'll add this information in the commit message. > > + /* > > + * If IPA range is too big (empirically found to be 256M), then > > fallback to > > + * full TLB flush > > + */ > > Bertrand, Michal, Stefano, Luca, can you check this limit would be ok > for your setup? Yes, please, if anyone of you can evaluate it in baremetal setup, then it would be great. Also, I have currently used the same limit in Arm32 which needs to be evaluated and it is not possible in my current nested virtualization setup. I'll update my patches as per the feedback comments and send out v3 soon. Regards, Haseeb
