Hello Julien, > On 9 Dec 2020, at 9:18 am, Julien Grall <[email protected]> wrote: > > > > On 09/12/2020 07:55, Bertrand Marquis wrote: >> Hi, > > Hi, > >> I also agree with the issue on the spinlock but we have no equivalent of >> something >> looking like a mutex for now in Xen so this would require some major >> redesign and >> will take us far from the linux driver. > > I agree that keeping the Xen implementation close to the Linux one is > important. However, I view this has a the secondary goal. The primary goal is > to have a safe and secure driver. > > If it means introducing a new set of lock or diverging from Linux, then so it > be. >> I would suggest to add a comment before this part of the code with a “TODO” >> so that >> it is clear inside the code. >> We could also add some comment in Kconfig to mention this possible “faulty” >> behaviour. > > I think it would be enough to write in the Kconfig that the driver is > "experimental and should not be used in production”.
Ok . I will add in Kconfig. > > Ideally, I would like a list of known issues for the driver (could be in the > cover letter or/and at the top of the source file) so we can track what's > missing to get it supported. I will list all known issues and short coming of the driver at the top of the source file. Regards, Rahul > > Cheers, > > -- > Julien Grall
