iommu.lists.linux-foundation.org archive mirror
 help / color / mirror / Atom feed
From: Alex Williamson <alex.williamson-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
To: linux-pci-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org
Cc: bhelgaas-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	linux-gpGsJRJZ3PBBDgjK7y7TUQ@public.gmane.org
Subject: [PATCH v3 02/15] PCI: quirk pci_for_each_dma_alias()
Date: Sat, 10 May 2014 09:02:35 -0600	[thread overview]
Message-ID: <20140510150235.2997.88101.stgit@bling.home> (raw)
In-Reply-To: <20140510145619.2997.429.stgit-xdHQ/5r00wBBDLzU/O5InQ@public.gmane.org>

There are a few broken devices that use the requester ID of a different
function in the slot for their DMA.  To handle these, add a bitmap to
struct pci_dev (using an alignment gap) that quirks can populate.  As
we iterate over the device and bus DMA aliases, also iterate over any
bits in the map.

Signed-off-by: Alex Williamson <alex.williamson-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
---
 drivers/pci/search.c |   21 +++++++++++++++++++++
 include/linux/pci.h  |    2 ++
 2 files changed, 23 insertions(+)

diff --git a/drivers/pci/search.c b/drivers/pci/search.c
index 5601cdb..ad698b2 100644
--- a/drivers/pci/search.c
+++ b/drivers/pci/search.c
@@ -37,6 +37,27 @@ int pci_for_each_dma_alias(struct pci_dev *pdev,
 	if (ret)
 		return ret;
 
+	/*
+	 * dma_func_alias provides a bitmap of other function numbers on
+	 * this same PCI slot to use as DMA aliases.
+	 */
+	if (unlikely(pdev->dma_func_alias)) {
+		u8 map = pdev->dma_func_alias & ~(1 << PCI_FUNC(pdev->devfn));
+		int func;
+
+		for (func = 0; map && func < 8; func++, map >>= 1) {
+			if (!(map & 1))
+				continue;
+
+			ret = fn(pdev,
+				 PCI_DEVID(pdev->bus->number,
+					   PCI_DEVFN(PCI_SLOT(pdev->devfn),
+					   func)), data);
+			if (ret)
+				return ret;
+		}
+	}
+
 	for (bus = pdev->bus; !pci_is_root_bus(bus); bus = bus->parent) {
 		struct pci_dev *tmp;
 
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 14b074b..b4c97d2 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -268,6 +268,8 @@ struct pci_dev {
 	u8		rom_base_reg;	/* which config register controls the ROM */
 	u8		pin;		/* which interrupt pin this device uses */
 	u16		pcie_flags_reg;	/* cached PCIe Capabilities Register */
+	u8		dma_func_alias;	/* bitmap of functions used as DMA
+					   aliases for this device */
 
 	struct pci_driver *driver;	/* which driver has allocated this device */
 	u64		dma_mask;	/* Mask of the bits of bus address this

  parent reply	other threads:[~2014-05-10 15:02 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-05-10 15:02 [PATCH v3 00/15] PCI/iommu: Fix DMA alias problems Alex Williamson
     [not found] ` <20140510145619.2997.429.stgit-xdHQ/5r00wBBDLzU/O5InQ@public.gmane.org>
2014-05-10 15:02   ` [PATCH v3 01/15] PCI: Add DMA alias iterator Alex Williamson
2014-05-10 15:02   ` Alex Williamson [this message]
2014-05-10 15:02   ` [PATCH v3 03/15] PCI: quirk dma_func_alias for Ricoh devices Alex Williamson
2014-05-10 15:02   ` [PATCH v3 04/15] PCI: quirk dma_func_alias for Marvell devices Alex Williamson
2014-05-10 15:02   ` [PATCH v3 05/15] PCI: Quirk pci_for_each_dma_alias() for bridges Alex Williamson
2014-05-10 15:02   ` [PATCH v3 06/15] PCI: Add quirks for ASMedia and Tundra bridges Alex Williamson
2014-05-10 15:03   ` [PATCH v3 07/15] PCI: Consolidate isolation domain code Alex Williamson
2014-05-10 15:03   ` [PATCH v3 08/15] iommu/amd: Use pci_find_dma_isolation_root() for IOMMU groups Alex Williamson
2014-05-14 10:34     ` Joerg Roedel
     [not found]       ` <20140514103449.GE6026-zLv9SwRftAIdnm+yROfE0A@public.gmane.org>
2014-05-14 14:27         ` Alex Williamson
2014-05-10 15:03   ` [PATCH v3 09/15] iommu/amd: Update to use PCI DMA aliases Alex Williamson
2014-05-10 15:03   ` [PATCH v3 10/15] iommu/intel: Use pci_find_dma_isolation_root() for IOMMU groups Alex Williamson
2014-05-10 15:03   ` [PATCH v3 11/15] iommu/intel: Update to use PCI DMA aliases Alex Williamson
2014-05-10 15:03   ` [PATCH v3 12/15] iommu/fsl: Use pci_find_dma_isolation_root() for IOMMU groups Alex Williamson
2014-05-10 15:03   ` [PATCH v3 13/15] iommu: Remove pci.h Alex Williamson
2014-05-10 15:03   ` [PATCH v3 14/15] PCI: Remove pci_find_upstream_pcie_bridge() Alex Williamson
2014-05-10 15:03   ` [PATCH v3 15/15] PCI: Remove pci_get_dma_source() Alex Williamson
2014-05-10 16:51   ` [PATCH v3 00/15] PCI/iommu: Fix DMA alias problems George Spelvin
     [not found] ` <1400020520427-859747.post@n7.nabble.com>
     [not found]   ` <1400020520427-859747.post-hbTxXoYVu7Vl57MIdRCFDg@public.gmane.org>
2014-05-14 17:18     ` Alex Williamson

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=20140510150235.2997.88101.stgit@bling.home \
    --to=alex.williamson-h+wxahxf7alqt0dzr+alfa@public.gmane.org \
    --cc=bhelgaas-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org \
    --cc=iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org \
    --cc=linux-gpGsJRJZ3PBBDgjK7y7TUQ@public.gmane.org \
    --cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=linux-pci-u79uwXL29TY76Z2rM5mHXA@public.gmane.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 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).