From: Mark Kettenis <mark.kettenis@xs4all.nl>
Subject: Re: bounce buffer mbuf defrag
To: Alexander Bluhm <bluhm@openbsd.org>
Cc: tech@openbsd.org
Date: Thu, 29 Aug 2024 16:57:41 +0200

> Date: Wed, 28 Aug 2024 14:37:45 +0200
> From: Alexander Bluhm <bluhm@openbsd.org>
> 
> Hi,
> 
> In my tests I have seen packet drops vio_encap() when using bounce
> buffers.  Some mbufs seem to be too fragmented for the pre allocated
> bounce buffer pages.  Calling m_defrag() fixes the problem.

Hmm.  How does this happen?  Is this a case for which the function
would have return EFBIG at the end in the non-bounce-buffer case?

> ok?
> 
> bluhm
> 
> Index: arch/amd64/amd64/bus_dma.c
> ===================================================================
> RCS file: /data/mirror/openbsd/cvs/src/sys/arch/amd64/amd64/bus_dma.c,v
> diff -u -p -r1.57 bus_dma.c
> --- arch/amd64/amd64/bus_dma.c	22 Aug 2024 11:36:24 -0000	1.57
> +++ arch/amd64/amd64/bus_dma.c	27 Aug 2024 20:48:12 -0000
> @@ -726,7 +726,7 @@ _bus_dmamap_load_buffer(bus_dma_tag_t t,
>  
>  		if (use_bounce_buffer) {
>  			if (page >= map->_dm_npages)
> -				return (ENOMEM);
> +				return (EFBIG);
>  
>  			off = vaddr & PAGE_MASK;
>  			pg = map->_dm_pages[page];
> 
>