On 7/13/22 10:47, Ján Tomko wrote: > On a Tuesday in 2022, Michal Privoznik wrote: >> We will need two attributes of usb-host device to set: >> guest-reset and guest-resets-all. The former was introduced in >> QEMU v4.0.0-rc0~56^2 and the other in v4.2.0-rc1~9^2. Hence, >> track the latter only as it's only starting from that commit when >> QEMU has both attributes. >> >> Our qemucapabilitiesdata are wrong though, as they enable the >> capability for quite old releases (for instance 3.1.0 when >> neither of the attributes existed). But this is because I didn't >> use actual QEMU to refresh .replies files in v6.9.0-rc1~457. I >> just blindly copied particular reply around. But that's okay, it >> doesn't really hurt anything. >> > > It hurts the usefulness of our test data. Just like in v6.9.0-rc1~457 > you only added the relevant attribute only to QEMU versions that have > it, you should delete it from older replies here, to demonstrate > that the capability detection code actually works.
I'm not sure how capability detection could not have worked since I had to add <flag name='usb-host.guest-resets-all'/> to capabiliteis XMLs. Unless it would be set unconditionally, which it is not. > > So there should be no "guest-reset" until QEMU 4.0.0, > no "guest-resets-all" until QEMU 4.2.0 and while you're at it, > "hostdevice" was added in QEMU 5.1.0 by 9f815e83e983d and > so was "suppress-remote-wake" by 7bacaf5fea I guess I just wanted to save myself from useless work since we are going to drop those files soon. But okay, let me post a patch to clean up those files. Michal
