Re: [PATCH 0/2] iommu/io-pgfault: reject partial faults after device removal

From: Baolu Lu

Date: Sun Nov 22 2026 - 02:39:29 EST


On 11/21/2026 10:13 AM, Qihang wrote:
Hi Baolu,

Thanks for confirming the VT-d ordering bug. I'll keep this series
withdrawn and won't resend it.

I still have one question about AMD. detach_device() calls
iopf_queue_flush_dev() just before amd_iommu_iopf_remove_device().
That flush only waits on the iopf workqueue, while the PPR IRQ thread
runs amd_iommu_poll_ppr_log() and reports the fault through
iommu_report_device_fault() directly. What ensures that the thread has
finished reporting faults for this device, and cannot report another
one, before the device is removed from the IOPF queue?

+ Vasant

Thanks,
Qihang

Thanks,
baolu