All of lore.kernel.org
 help / color / mirror / Atom feed
From: thomas.petazzoni@free-electrons.com (Thomas Petazzoni)
To: linux-arm-kernel@lists.infradead.org
Subject: Intel I350 mini-PCIe card (igb) on Mirabox (mvebu / Armada 370)
Date: Tue, 8 Apr 2014 23:50:17 +0200	[thread overview]
Message-ID: <20140408235017.2ae3816a@skate> (raw)
In-Reply-To: <2FF5F2FD-2DAC-4E51-87A0-90DBAFE6AA01@xyratex.com>

Dear Matthew Minter,

On Tue, 8 Apr 2014 21:17:47 +0100, Matthew Minter wrote:

> This is just a guess but perhaps the best we could do would be to round up but also add logic to fail completely if adding a PCI device would cause a window to overlap?

Rounding up does not work. Let me try to explain why.

The PCI core enumerates all the BARs into your device and decides that
it should be mapped from 0xe0000000 to 0xe0900000 (9 MB). The PCI core
writes those values into the BAR of the emulated bridge in the
pci-mvebu driver. The pci-mvebu driver captures this write, and figures
out that 9 MB is not good for a MBus window. So it rounds it up to 16
MB, from 0xe0000000 to 0xe1000000, and configures the MBus window. Your
device will work perfectly fine.

Now, the PCI core goes on and enumerate the next device. This device
needs 1 MB of memory space. From the PCI core point of view, only
0xe0000000 to 0xe0900000 is occupied, so for the next device, it may
well decide to use 0xe0900000 -> 0xe0a00000, write this to the BAR of
the emulated bridge, which will lead the emulated bridge to try to
create a window for this area... which isn't possible as it overlaps
the previous window.

Conclusion: rounding up the size of the BARs cannot be done without the
cooperation of the PCI core.

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

  reply	other threads:[~2014-04-08 21:50 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-25 20:07 Intel I350 mini-PCIe card (igb) on Mirabox (mvebu / Armada 370) Neil Greatorex
2014-03-25 20:20 ` Thomas Petazzoni
2014-03-25 21:03   ` Willy Tarreau
2014-03-25 20:22 ` Jason Gunthorpe
2014-03-25 20:36   ` Thomas Petazzoni
2014-03-25 21:12     ` Jason Gunthorpe
2014-03-25 21:23       ` Thomas Petazzoni
2014-03-25 22:03     ` Neil Greatorex
2014-03-25 22:24       ` Jason Gunthorpe
2014-03-25 22:35         ` Jason Gunthorpe
2014-03-26 19:31           ` Neil Greatorex
2014-03-26 20:12             ` Jason Gunthorpe
2014-03-26 20:34               ` Neil Greatorex
2014-03-26 21:42                 ` Jason Gunthorpe
2014-03-26 21:52                   ` Thomas Petazzoni
2014-03-27  0:29                   ` Neil Greatorex
2014-03-27  4:40                     ` Jason Gunthorpe
2014-03-28  1:03                       ` Neil Greatorex
2014-03-28  2:04                         ` Jason Gunthorpe
2014-04-04 13:19                         ` Neil Greatorex
2014-04-05 17:32                           ` Willy Tarreau
2014-04-05 17:34                           ` Thomas Petazzoni
2014-04-05 18:04                             ` Willy Tarreau
2014-04-05 18:55                               ` Neil Greatorex
2014-04-05 19:03                                 ` Willy Tarreau
2014-04-05 19:00                             ` Neil Greatorex
2014-04-06 15:34                               ` Neil Greatorex
2014-04-06 17:43                                 ` Willy Tarreau
2014-04-08 15:13                                 ` Thomas Petazzoni
2014-04-08 15:40                                   ` Thomas Petazzoni
2014-04-08 15:55                                     ` Thomas Petazzoni
2014-04-08 16:02                                       ` Matthew Minter
2014-04-08 17:14                                       ` Jason Gunthorpe
2014-04-08 17:53                                         ` Willy Tarreau
2014-04-08 18:08                                           ` Jason Gunthorpe
2014-04-08 18:15                                             ` Thomas Petazzoni
2014-04-08 18:40                                               ` Jason Gunthorpe
2014-04-08 19:15                                             ` Willy Tarreau
2014-04-08 19:21                                               ` Jason Gunthorpe
2014-04-08 20:17                                                 ` Matthew Minter
2014-04-08 21:50                                                   ` Thomas Petazzoni [this message]
2014-04-08 20:19                                                 ` Neil Greatorex
2014-04-08 20:43                                                 ` Willy Tarreau
2014-04-08 18:01                                         ` Thomas Petazzoni
2014-04-08 18:22                                           ` Jason Gunthorpe
2014-04-08 18:32                                             ` Thomas Petazzoni
2014-04-08 15:53                                   ` Willy Tarreau
2014-04-08 16:00                                     ` Thomas Petazzoni
2014-04-08 16:05                                       ` Willy Tarreau
2014-04-06 18:58                           ` Willy Tarreau
2014-04-06 19:11                             ` Thomas Petazzoni
2014-04-06 21:57                             ` Neil Greatorex
2014-04-06 22:04                               ` Willy Tarreau
2014-04-06 22:16                               ` Thomas Petazzoni
2014-04-07  0:50                                 ` Neil Greatorex
2014-04-07 17:41                               ` Jason Gunthorpe
2014-04-07 19:41                                 ` Neil Greatorex
2014-04-07 20:48                                   ` Jason Gunthorpe
2014-04-07 21:58                                     ` Neil Greatorex
2014-04-08  6:28                                       ` Willy Tarreau
2014-04-08  6:40                                       ` Willy Tarreau
2014-04-08 10:53                                         ` Matthew Minter
2014-04-08 12:31                                           ` Matthew Minter
2014-04-08 12:36                                             ` Willy Tarreau
2014-04-08 14:43                                               ` Thomas Petazzoni
2014-04-08 14:52                                                 ` Matthew Minter
2014-04-08 14:53                                                 ` Willy Tarreau
2014-04-08 15:25                                                   ` Thomas Petazzoni
2014-04-08 17:56                                             ` Willy Tarreau

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=20140408235017.2ae3816a@skate \
    --to=thomas.petazzoni@free-electrons.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.