From: Stephen Hemminger <shemminger@linux-foundation.org>
To: Michael Breuer <mbreuer@majjas.com>
Cc: Jarek Poplawski <jarkao2@gmail.com>,
David Miller <davem@davemloft.net>,
akpm@linux-foundation.org, flyboy@gmail.com,
linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
Michael Chan <mchan@broadcom.com>, Don Fry <pcnet32@verizon.net>,
Francois Romieu <romieu@fr.zoreil.com>,
Matt Carlson <mcarlson@broadcom.com>
Subject: Re: Hang: 2.6.32.4 sky2/DMAR (was [PATCH] sky2: Fix WARNING: at lib/dma-debug.c:902 check_sync)
Date: Wed, 27 Jan 2010 09:56:14 -0800 [thread overview]
Message-ID: <20100127095614.14313677@nehalam> (raw)
In-Reply-To: <4B60707F.1000608@majjas.com>
On Wed, 27 Jan 2010 11:57:35 -0500
Michael Breuer <mbreuer@majjas.com> wrote:
> On 1/27/2010 11:50 AM, Stephen Hemminger wrote:
> > On Wed, 27 Jan 2010 10:34:51 -0500
> > Michael Breuer<mbreuer@majjas.com> wrote:
> >
> >
> >> On 01/23/2010 06:21 PM, Jarek Poplawski wrote:
> >>
> >>> On Fri, Jan 22, 2010 at 06:50:21PM -0500, Michael Breuer wrote:
> >>>
> >>>
> >>>> When the packets were dropped, there was a different sequence in the
> >>>> log - DISCOVER/OFFER repeated. The "normal" is that the sequence
> >>>> appeared correct and complete - DISCOVER/OFFER/REQUEST/ACK - or
> >>>> INFORM/ACK (vs. INFORM repeatedly sans ACK) as the case may be.
> >>>>
> >>>>
> >>> Anyway, I'd be intersted if the switch matters here.
> >>>
> >>> Plus one more test: could you try to load sky2 with the parameter:
> >>> "copybreak=1" (the rest as in any recent test, which gave you dmar
> >>> errors; any switch).
> >>>
> >>> Thanks,
> >>> Jarek P.
> >>>
> >>>
> >> Ok - now up 80+ hours with copybreak=1. I'm going to redo w/o copybreak
> >> to confirm that I haven't inadvertently fixed something. However, given
> >> that it might be copybreak-related, I looked at sky2.c again and I'm
> >> wondering about the copybreak max size in sky2_rx_start:
> >>
> >> size = roundup(sky2->netdev->mtu + ETH_HLEN + VLAN_HLEN, 8);
> >>
> >> /* Stopping point for hardware truncation */
> >> thresh = (size - 8) / sizeof(u32);
> >>
> >> sky2->rx_nfrags = size>> PAGE_SHIFT;
> >> BUG_ON(sky2->rx_nfrags> ARRAY_SIZE(re->frag_addr));
> >>
> >> /* Compute residue after pages */
> >> size -= sky2->rx_nfrags<< PAGE_SHIFT;
> >>
> >> /* Optimize to handle small packets and headers */
> >> if (size< copybreak)
> >> size = copybreak;
> >> if (size< ETH_HLEN)
> >> size = ETH_HLEN;
> >>
> >>
> >> Why would increasing size to copybreak be valid here?
> >>
> >> Guessing a bit as I'm not sure about rx_nfrags, but if I read this
> >> correctly, if size is ever less than copybreak it's because there isn't
> >> enough space left for anything larger. If so, wouldn't increasing size
> >> potentially corrupt something? I'd further guess that the resulting
> >> condition manifests sooner (or at least with a more visible effect) when
> >> using DMAR.
> >>
> >> In any event, why "copybreak" as the minimum buffer size? I'd suggest
> >> that if it isn't possible to allocate at least MTU + overhead that
> >> sky2_rx_start ought to be delayed until there is room.
> >>
> > This code is where driver decides how much data will be received in skb
> > data area and the remaining data spills over into skb frags.
> > Copybreak is the threshold so that packets less than size are copied
> > to a new skb. The code doing the copying there assumes the data is
> > totally contained in the skb (not in frags). The size increase there
> > is to make sure that assumption is always true. I suppose you
> > could do something perverse like setting copybreak really huge
> > and confuse driver, but that is a user error.
> >
> >
> Ok - but I'm wondering under what circumstances size would be <
> copybreak in the first place after computing the residue. If size ends
> up being unreasonably small, is simply increasing the number to whatever
> copybreak is correct? Assuming my testing is correct, then the crash
> I've been experiencing when using dmar (only) seems related to the value
> of copybreak. I don't think the other use (skb reuse) is the issue (but
> hey, I could have missed something). The crash occurs when copybreak is
> the default of 128, didn't happen when I set copybreak to 1.
Does this change it? If so the dma code is (not sky2) is buggy and not
rounding up properly.
--- a/drivers/net/sky2.c 2010-01-27 09:46:10.940005248 -0800
+++ b/drivers/net/sky2.c 2010-01-27 09:53:47.141267850 -0800
@@ -2257,13 +2257,16 @@ static struct sk_buff *receive_copy(stru
skb = netdev_alloc_skb_ip_align(sky2->netdev, length);
if (likely(skb)) {
+ unsigned dma_align = dma_get_cache_alignment();
+ unsigned dma_size = ALIGN(length+1, dma_align);
+
pci_dma_sync_single_for_cpu(sky2->hw->pdev, re->data_addr,
- length, PCI_DMA_FROMDEVICE);
+ dma_size, PCI_DMA_FROMDEVICE);
skb_copy_from_linear_data(re->skb, skb->data, length);
skb->ip_summed = re->skb->ip_summed;
skb->csum = re->skb->csum;
pci_dma_sync_single_for_device(sky2->hw->pdev, re->data_addr,
- length, PCI_DMA_FROMDEVICE);
+ dma_size, PCI_DMA_FROMDEVICE);
re->skb->ip_summed = CHECKSUM_NONE;
skb_put(skb, length);
}
next prev parent reply other threads:[~2010-01-27 17:56 UTC|newest]
Thread overview: 95+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-01-20 9:41 [PATCH] sky2: Fix WARNING: at lib/dma-debug.c:902 check_sync Jarek Poplawski
2010-01-20 18:03 ` Stephen Hemminger
2010-01-20 20:11 ` Michael Chan
2010-01-20 20:30 ` Stephen Hemminger
2010-01-20 20:58 ` Jarek Poplawski
2010-01-20 22:50 ` David Miller
2010-01-20 22:45 ` David Miller
2010-01-20 18:09 ` Stephen Hemminger
2010-01-20 22:24 ` Alan Cox
2010-01-20 22:53 ` David Miller
2010-01-20 22:53 ` Jarek Poplawski
2010-01-21 15:22 ` FUJITA Tomonori
2010-01-21 18:41 ` Jarek Poplawski
2010-01-22 5:11 ` FUJITA Tomonori
2010-01-22 6:38 ` David Miller
2010-02-03 1:18 ` FUJITA Tomonori
2010-02-03 1:27 ` David Miller
2010-01-21 19:59 ` Michael Breuer
2010-01-21 20:41 ` Jarek Poplawski
2010-01-21 20:46 ` Michael Breuer
2010-01-21 21:02 ` Jarek Poplawski
2010-01-22 18:01 ` Hang: 2.6.32.4 sky2/DMAR (was [PATCH] sky2: Fix WARNING: at lib/dma-debug.c:902 check_sync) Michael Breuer
2010-01-22 21:53 ` Jarek Poplawski
2010-01-22 22:14 ` Michael Breuer
2010-01-22 23:06 ` Jarek Poplawski
2010-01-22 23:25 ` Michael Breuer
2010-01-22 23:46 ` Jarek Poplawski
2010-01-22 23:50 ` Michael Breuer
2010-01-23 23:21 ` Jarek Poplawski
2010-01-24 1:53 ` Michael Breuer
2010-01-27 15:34 ` Michael Breuer
2010-01-27 16:50 ` Stephen Hemminger
2010-01-27 16:57 ` Michael Breuer
2010-01-27 17:45 ` Stephen Hemminger
2010-01-27 17:57 ` Michael Breuer
2010-01-27 18:33 ` Michael Breuer
2010-01-27 23:54 ` Hang: 2.6.32.4 sky2/DMAR David Miller
2010-01-27 17:56 ` Stephen Hemminger [this message]
2010-01-27 17:58 ` Hang: 2.6.32.4 sky2/DMAR (was [PATCH] sky2: Fix WARNING: at lib/dma-debug.c:902 check_sync) Michael Breuer
2010-01-27 18:08 ` Michael Breuer
2010-01-27 18:45 ` Michael Breuer
2010-01-27 19:23 ` Jarek Poplawski
2010-01-27 19:32 ` Jarek Poplawski
2010-01-28 15:32 ` Michael Breuer
2010-01-28 16:43 ` Michael Breuer
2010-01-28 17:08 ` Stephen Hemminger
2010-01-28 18:46 ` Michael Breuer
2010-01-28 22:34 ` Jarek Poplawski
2010-01-28 22:43 ` Michael Breuer
2010-01-28 22:56 ` Jarek Poplawski
2010-01-28 22:59 ` Michael Breuer
2010-01-28 23:36 ` [PATCH] sky2: receive dma mapping error handling Stephen Hemminger
2010-01-29 0:05 ` Michael Breuer
2010-01-30 16:30 ` Michael Breuer
2010-01-30 16:31 ` Michael Breuer
2010-01-31 0:34 ` Jarek Poplawski
2010-01-31 4:17 ` Michael Breuer
2010-01-31 22:25 ` Jarek Poplawski
2010-01-31 23:58 ` Michael Breuer
2010-01-31 4:55 ` Michael Breuer
2010-01-31 18:50 ` Michael Breuer
2010-01-31 21:58 ` Michael Breuer
2010-01-31 22:18 ` Jarek Poplawski
2010-02-01 0:19 ` Michael Breuer
2010-02-01 4:26 ` Michael Breuer
2010-02-01 10:47 ` Jarek Poplawski
2010-02-01 9:17 ` [PATCH v2] sky2: Fix transmit dma mapping handling Jarek Poplawski
2010-02-01 17:52 ` Michael Breuer
2010-02-01 18:08 ` [PATCH] sky2: receive dma mapping error handling Stephen Hemminger
2010-02-01 18:20 ` Stephen Hemminger
2010-02-01 18:44 ` Michael Breuer
2010-02-01 20:13 ` Jarek Poplawski
2010-02-01 20:41 ` Jarek Poplawski
2010-02-01 21:27 ` [PATCH v3] " Jarek Poplawski
2010-02-01 22:29 ` Stephen Hemminger
2010-02-01 22:46 ` Jarek Poplawski
2010-02-01 22:51 ` Stephen Hemminger
2010-02-01 21:42 ` [PATCH v3b resent] sky2: Fix transmit dma mapping handling Jarek Poplawski
2010-02-03 4:07 ` [PATCH] sky2: receive dma mapping error handling Michael Breuer
2010-02-03 16:47 ` Michael Breuer
2010-02-03 16:56 ` Stephen Hemminger
2010-02-03 17:07 ` Michael Breuer
2010-02-03 18:23 ` Justin P. Mattock
2010-02-03 18:25 ` Stephen Hemminger
2010-02-03 18:48 ` Justin P. Mattock
2010-02-03 17:16 ` Justin P. Mattock
2010-02-02 22:44 ` Andi Kleen
2012-01-16 16:39 ` Regression: sky2 kernel between 3.1 and 3.2.1 (last known good 3.0.9) Michael Breuer
2012-01-20 14:24 ` Michael Breuer
2012-01-20 16:10 ` Stephen Hemminger
2012-01-20 16:17 ` Michael Breuer
2012-01-20 16:26 ` Stephen Hemminger
2012-01-20 16:44 ` Michael Breuer
2012-01-21 15:29 ` Michael Breuer
2012-01-22 18:03 ` Stephen Hemminger
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20100127095614.14313677@nehalam \
--to=shemminger@linux-foundation.org \
--cc=akpm@linux-foundation.org \
--cc=davem@davemloft.net \
--cc=flyboy@gmail.com \
--cc=jarkao2@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mbreuer@majjas.com \
--cc=mcarlson@broadcom.com \
--cc=mchan@broadcom.com \
--cc=netdev@vger.kernel.org \
--cc=pcnet32@verizon.net \
--cc=romieu@fr.zoreil.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).