weekly report of July 8

zhixiang yu zxyu.core at gmail.com
Tue Jul 16 13:31:30 UTC 2013

Hi all,

This week, I focused on design decision about multithread,kqueue or
not. As we know, pci_virtio_block doesn't make use of multithread and
kqueue, but pci_virtio_net fully makes use of these technologies.
Guided by my mentor Alexander, I know it's better to make it

Then we talked with Neel and Peter and found that Peter has
implemented one excellent multithreaded general block_if which can be
used in my project.

Finally, I merged Peter's patch and fixed one problem of the block_if.
Per my test, the block_if now works perfectly. However, the
pci_virtio_block still can't work after switching to block_if.

I'll bypass pci_virtio_block since it doesn't impact my SoC project.


More information about the soc-status mailing list