Re: [PATCH 1/1] iommu/iova: Make rcache flush optional on IOVA allocation failure

From: Tomasz Nowicki
Date: Tue Sep 19 2017 - 04:04:12 EST


Hi Robin,

On 18.09.2017 18:02, Robin Murphy wrote:
Hi Tomasz,

On 18/09/17 11:56, Tomasz Nowicki wrote:
Since IOVA allocation failure is not unusual case we need to flush
CPUs' rcache in hope we will succeed in next round.

However, it is useful to decide whether we need rcache flush step because
of two reasons:
- Not scalability. On large system with ~100 CPUs iterating and flushing
rcache for each CPU becomes serious bottleneck so we may want to deffer it.
- free_cpu_cached_iovas() does not care about max PFN we are interested in.
Thus we may flush our rcaches and still get no new IOVA like in the
commonly used scenario:

if (dma_limit > DMA_BIT_MASK(32) && dev_is_pci(dev))
iova = alloc_iova_fast(iovad, iova_len, DMA_BIT_MASK(32) >> shift);

if (!iova)
iova = alloc_iova_fast(iovad, iova_len, dma_limit >> shift);

1. First alloc_iova_fast() call is limited to DMA_BIT_MASK(32) to get
PCI devices a SAC address
2. alloc_iova() fails due to full 32-bit space
3. rcaches contain PFNs out of 32-bit space so free_cpu_cached_iovas()
throws entries away for nothing and alloc_iova() fails again
4. Next alloc_iova_fast() call cannot take advantage of rcache since we
have just defeated caches. In this case we pick the slowest option
to proceed.

This patch reworks flushed_rcache local flag to be additional function
argument instead and control rcache flush step. Also, it updates all users
to do the flush as the last chance.

Looks like you've run into the same thing Nate found[1] - I came up with
almost the exact same patch, only with separate alloc_iova_fast() and
alloc_iova_fast_noretry() wrapper functions, but on reflection, just
exposing the bool to callers is probably simpler. One nit, can you
document it in the kerneldoc comment too? With that:

Reviewed-by: Robin Murphy <robin.murphy@xxxxxxx>

Thanks! I will add missing comment.

Tomasz