Re: Strange network timeouts w/ 2.6.30.5

[Walt Holman <walt@holmansrus.com>]

----- "Krzysztof Halasa" <khc@pm.waw.pl> wrote:

> Walt Holman <walt@holmansrus.com> writes:
> 
> > dmesg is attached. This box does have >2GB Ram (6GB total).  The
> dmesg
> > will show e100 init'd 3 times since the first is the stock
> modprobe,
> > 2nd was forced with use_io and the 3rd modprobe was after reverting
> > the patch.
> 
> You most probably can't test without swiotlb (RAM has to be limited
> to
> 2 GB or so), can you? That would (dis)prove my theory. Alternatively
> (or
> better), a test on IOMMU-equipped system would do.

Would something like passing a mem=xx cmdline on x86_64 be sufficient to test this?

-Walt



> 
> Since swiotlb is x86-only thing (though other 64-bit archs may have
> something similar), I think the correct work around is to enable the
> "for_device" handoff on !X86.
> 
> Something like maybe:
> 
> Signed-off-by: Krzysztof Hałasa <khc@pm.waw.pl>
> 
> diff --git a/drivers/net/e100.c b/drivers/net/e100.c
> index 014dfb6..b610088 100644
> --- a/drivers/net/e100.c
> +++ b/drivers/net/e100.c
> @@ -1762,9 +1762,12 @@ static int e100_rx_indicate(struct nic *nic,
> struct rx *rx,
>  
>  			if (ioread8(&nic->csr->scb.status) & rus_no_res)
>  				nic->ru_running = RU_SUSPENDED;
> +#ifndef CONFIG_X86
> +		/* FIXME interferes with swiotlb. */
>  		pci_dma_sync_single_for_device(nic->pdev, rx->dma_addr,
>  					       sizeof(struct rfd),
>  					       PCI_DMA_BIDIRECTIONAL);
> +#endif
>  		return -ENODATA;
>  	}
>  
> -- 
> Krzysztof Halasa