Re: git: f44ff2aba2d6 - main - bhyve: Treat the COMMAND register for PCI passthru devices as emulated
Date: Mon, 19 Aug 2024 13:22:01 UTC
On 8/13/24 21:17, Navdeep Parhar wrote: > On 7/31/24 7:50 AM, John Baldwin wrote: >> The branch main has been updated by jhb: >> >> URL: https://cgit.FreeBSD.org/src/commit/?id=f44ff2aba2d64d1d9312cb55008dc90275ccdc04 >> >> commit f44ff2aba2d64d1d9312cb55008dc90275ccdc04 >> Author: John Baldwin <jhb@FreeBSD.org> >> AuthorDate: 2024-07-31 14:50:33 +0000 >> Commit: John Baldwin <jhb@FreeBSD.org> >> CommitDate: 2024-07-31 14:50:33 +0000 >> >> bhyve: Treat the COMMAND register for PCI passthru devices as emulated >> >> Don't pass through writes of the command register through to the >> physical device. These registers do not need to be in sync, and in >> some cases (e.g. when the guest is sizing the BAR and temporarily >> disables decoding), the states need to diverge. > > This seems to break MSI-X interrupts for a device that is passed through > to a VM. That is, the device generates an interrupt but the driver in > the VM never receives it. Reverting this change fixes the problem. Yes, https://reviews.freebsd.org/D46245 is what I have been using. It turns out the ppt(4) driver wasn't fully managing the command register. -- John Baldwin