- Matt (Arm)
- Stefano (Xilinx)
- Bogdan (NXP)
- Srivatsa (QC)
- Diana (NXP)
- Ed Doxat (Arm)
- Get a view of the engineering support available to work on Stratos
Update on the demo progress
Masami : Booted Xen Dom0 with video - on Synquaser - how can we enabe virtio-video, virti-io video != virtio-gpu is 3d capable - Alex has this almost working and will send the details on.
Alex : is this the AGL image, can't boot their kernel, steffano said we need Xen console. Been trying to build since yesterday
Steffano : config-xen ths tiggers other options. if you have two consoles you can tell Dom0 use the second one, to start guest you need config-xen
Alex How many patches 1 for Xen in kernel, Masami - I can share it, Alex Grub ? no changes. Alex need this for Machiatobin
Alex : Gave AGL meeting an overview, appeared happy, wil update slides on Stratos homepage.
Update on RPMB patches
Not huge change - posted, some requests for support for EMMC probing - Illias uncovered the issues, reached out to Ulf to find ways to make things easier.
Did not start on re-implementation using Rust vhost-user libs.
Illias : we can look at kernel but, you expect the same driver as host vs support 1/2 the MMC calls, I have uboot patches, Alex I want to put it in the spec vs just change the driver
Introduction to cheri and some thoughts on its relevance to Stratos
Workign with Arm : been running 10 years
Steffano : cherification is adding capabilities : Joakim yes
Arnd : not sensible to try with Linux, it cant work with t 128 bits
Lots of questions about what happens to pointers across security boundaries (sycalls into kernels, pointers between device/driver, moving to secure world etc).
Please ping Joakim if you have questions
Kevin Brodsky's talk giving answers to many of Arnd's questions statements
The current state of VirtIO-MMU / Limited Memory Sharing
Arnd: discussed 3 different approaches - do they all suck? Which to start prototyping?
Option 1: host manages shared memory, use swiotlb to force use of shared memory
Option 2: guest uses IOMMU, signals pages meant for access, no copying, expensive transactions updating iommu?
Option 3: new virtquue design, shared memory part of queue (inline).
Steffano: can we increase the size of the queue? Arnd: it involves changing the binary layout of the queue
Steffano: had to increase ring size for 9pfs implementation - needed to go to 2Mb (up from 512K)
Arnd: having large queues risks cache thrashing.
Alex: Does swiotlb always imply copying Arnd: yes
Stefano: what about virtio managing page allocation?
Alex: sees the complexities with existing devices, talked about video - size it for "known" memory bandwidth
Stefano: what about expanding option 3 but allowing bigger sizes queus appropriate to the device?
Arnd: Option 3 has s simplicity, but offers a allocation per device.
Srivista: our implementation currently is closer to option 1
Arnd: so which option to implement, or do we see if QC can share their implementation
Alex: I'm leaning towards 3 which could then be expanded to larger queue sizes for given devices. We could then compare with QC swiotlb approach
What is the development platform ?