On 12/27/2023 5:06 PM, Duan, Zhenzhong wrote:
-----Original Message-----Is it necessary to return fault to user if qi_check_fault() return -EAGAIN and
From: Liu, Yi L <yi.l.liu@xxxxxxxxx>
Sent: Tuesday, December 26, 2023 4:44 PM
Subject: Re: [PATCH v7 7/9] iommu/vt-d: Allow qi_submit_sync() to return
the QI faults
On 2023/12/26 14:15, Yi Liu wrote:
hmmm, replied too soon. The qi_check_fault() would be called multiple
On 2023/12/26 12:13, Tian, Kevin wrote:
I guess no such a reason. :) let me modify it.From: Liu, Yi L <yi.l.liu@xxxxxxxxx>I meant:
Sent: Tuesday, December 26, 2023 12:03 PM
On 2023/12/22 12:23, Tian, Kevin wrote:
not quite get. do you mean the fault should not be cleared in the callerFrom: Liu, Yi L <yi.l.liu@xxxxxxxxx>do we expect the fault to be accumulated? otherwise it's clearer to
Sent: Thursday, December 21, 2023 11:40 PM
+ fault &= DMA_FSTS_IQE | DMA_FSTS_ITE | DMA_FSTS_ICE;
+ if (fault) {
+ if (fsts)
+ *fsts |= fault;
just do direct assignment instead of asking for the caller to clear
the variable before invocation.
side?
if (fsts)
*fsts = fault;
unless there is a reason to *OR* the original value.
times in one invalidation circle as qi_submit_sync() needs to see if any
fault happened before the hw writes back QI_DONE to the wait descriptor.
There can be ICE which may eventually result in ITE. So caller of
qi_check_fault()
would continue to wait for QI_DONE. So qi_check_fault() returns 0 to let
qi_submit_sync() go on though ICE detected. If we use '*fsts = fault;',
then ICE would be missed since the input fsts pointer is the same in
one qi_submit_sync() call.
a restart run succeeds?
Issue a device-TLB invalidation to no response device there is possibility
will be trapped there loop for ITE , never get return.