diff mbox series

[v2] ARM: sa1100/assabet: move dmabounce hack to ohci driver

Message ID 20220203083658.559803-1-arnd@kernel.org
State New
Headers show
Series [v2] ARM: sa1100/assabet: move dmabounce hack to ohci driver | expand

Commit Message

Arnd Bergmann Feb. 3, 2022, 8:36 a.m. UTC
From: Arnd Bergmann <arnd@arndb.de>

The sa1111 platform is one of the two remaining users of the old Arm
specific "dmabounce" code, which is an earlier implementation of the
generic swiotlb.

Linus Walleij submitted a patch that removes dmabounce support from
the ixp4xx, and I had a look at the other user, which is the sa1111
companion chip.

Looking at how dmabounce is used, I could narrow it down to one driver
one three machines:

 - dmabounce is only initialized on assabet/neponset, jornada720 and
   badge4, which are the platforms that have an sa1111 and support
   DMA on it.

 - All three of these suffer from "erratum #7" that requires only
   doing DMA to half the memory sections based on one of the address
   lines, in addition, the neponset also can't DMA to the RAM that
   is connected to sa1111 itself.

 - the pxa lubbock machine also has sa1111, but does not support DMA
   on it and does not set dmabounce.

 - only the OHCI and audio devices on sa1111 support DMA, but as
   there is no audio driver for this hardware, only OHCI remains.

In the OHCI code, I noticed that two other platforms already have
a local bounce buffer support in the form of the "local_mem"
allocator. Specifically, TMIO and SM501 use this on a few other ARM
boards with 16KB or 128KB of local SRAM that can be accessed from the
OHCI and from the CPU.

While this is not the same problem as on sa1111, I could not find a
reason why we can't re-use the existing implementation but replace the
physical SRAM address mapping with a locally allocated DMA buffer.

There are two main downsides:

 - rather than using a dynamically sized pool, this buffer needs
   to be allocated at probe time using a fixed size. Without
   having any idea of what it should be, I picked a size of
   64KB, which is between what the other two OHCI front-ends use
   in their SRAM. If anyone has a better idea what that size
   is reasonable, this can be trivially changed.

 - Previously, only USB transfers to unaddressable memory needed
   to go through the bounce buffer, now all of them do, which may
   impact runtime performance for USB endpoints that do a lot of
   transfers.

On the upside, the local_mem support uses write-combining buffers,
which should be a bit faster for transfers to the device compared to
normal uncached coherent memory as used in dmabounce.

Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Laurentiu Tudor <laurentiu.tudor@nxp.com>
Cc: linux-usb@vger.kernel.org
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
Changes in v2:

 - drop check for assabet, as bounce buffers are required on
   all sa1100 machines
 - select CONFIG_ZONE_DMA again
 - update comments and changelog text based on discussion
---
 arch/arm/common/Kconfig        |  2 +-
 arch/arm/common/sa1111.c       | 64 ----------------------------------
 drivers/usb/core/hcd.c         | 17 +++++++--
 drivers/usb/host/ohci-sa1111.c | 25 +++++++++++++
 4 files changed, 40 insertions(+), 68 deletions(-)

Comments

Ard Biesheuvel Feb. 3, 2022, 8:47 a.m. UTC | #1
On Thu, 3 Feb 2022 at 09:38, Arnd Bergmann <arnd@kernel.org> wrote:
>
> From: Arnd Bergmann <arnd@arndb.de>
>
> The sa1111 platform is one of the two remaining users of the old Arm
> specific "dmabounce" code, which is an earlier implementation of the
> generic swiotlb.
>
> Linus Walleij submitted a patch that removes dmabounce support from
> the ixp4xx, and I had a look at the other user, which is the sa1111
> companion chip.
>
> Looking at how dmabounce is used, I could narrow it down to one driver
> one three machines:
>
>  - dmabounce is only initialized on assabet/neponset, jornada720 and
>    badge4, which are the platforms that have an sa1111 and support
>    DMA on it.
>
>  - All three of these suffer from "erratum #7" that requires only
>    doing DMA to half the memory sections based on one of the address
>    lines, in addition, the neponset also can't DMA to the RAM that
>    is connected to sa1111 itself.
>
>  - the pxa lubbock machine also has sa1111, but does not support DMA
>    on it and does not set dmabounce.
>
>  - only the OHCI and audio devices on sa1111 support DMA, but as
>    there is no audio driver for this hardware, only OHCI remains.
>
> In the OHCI code, I noticed that two other platforms already have
> a local bounce buffer support in the form of the "local_mem"
> allocator. Specifically, TMIO and SM501 use this on a few other ARM
> boards with 16KB or 128KB of local SRAM that can be accessed from the
> OHCI and from the CPU.
>
> While this is not the same problem as on sa1111, I could not find a
> reason why we can't re-use the existing implementation but replace the
> physical SRAM address mapping with a locally allocated DMA buffer.
>
> There are two main downsides:
>
>  - rather than using a dynamically sized pool, this buffer needs
>    to be allocated at probe time using a fixed size. Without
>    having any idea of what it should be, I picked a size of
>    64KB, which is between what the other two OHCI front-ends use
>    in their SRAM. If anyone has a better idea what that size
>    is reasonable, this can be trivially changed.
>

I suppose this is a problem if the driver falls back to ordinary DRAM
once the allocation runs out?

>  - Previously, only USB transfers to unaddressable memory needed
>    to go through the bounce buffer, now all of them do, which may
>    impact runtime performance for USB endpoints that do a lot of
>    transfers.
>
> On the upside, the local_mem support uses write-combining buffers,
> which should be a bit faster for transfers to the device compared to
> normal uncached coherent memory as used in dmabounce.
>

Talking from past experience using this trick on a NXP ARM9 SoC ~10
years ago, using on-chip SRAM for USB DMA likely results in a
significant performance boost, even without write combining, although
the exact scenario obviously matters.


> Cc: Linus Walleij <linus.walleij@linaro.org>
> Cc: Russell King <linux@armlinux.org.uk>
> Cc: Christoph Hellwig <hch@infradead.org>
> Cc: Laurentiu Tudor <laurentiu.tudor@nxp.com>
> Cc: linux-usb@vger.kernel.org
> Acked-by: Alan Stern <stern@rowland.harvard.edu>
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> ---
> Changes in v2:
>
>  - drop check for assabet, as bounce buffers are required on
>    all sa1100 machines
>  - select CONFIG_ZONE_DMA again
>  - update comments and changelog text based on discussion
> ---
>  arch/arm/common/Kconfig        |  2 +-
>  arch/arm/common/sa1111.c       | 64 ----------------------------------
>  drivers/usb/core/hcd.c         | 17 +++++++--
>  drivers/usb/host/ohci-sa1111.c | 25 +++++++++++++
>  4 files changed, 40 insertions(+), 68 deletions(-)
>
> diff --git a/arch/arm/common/Kconfig b/arch/arm/common/Kconfig
> index c8e198631d41..bc158fd227e1 100644
> --- a/arch/arm/common/Kconfig
> +++ b/arch/arm/common/Kconfig
> @@ -1,7 +1,7 @@
>  # SPDX-License-Identifier: GPL-2.0
>  config SA1111
>         bool
> -       select DMABOUNCE if !ARCH_PXA
> +       select ZONE_DMA if ARCH_SA1100
>
>  config DMABOUNCE
>         bool
> diff --git a/arch/arm/common/sa1111.c b/arch/arm/common/sa1111.c
> index 7df003b149c6..a00915883f78 100644
> --- a/arch/arm/common/sa1111.c
> +++ b/arch/arm/common/sa1111.c
> @@ -1391,70 +1391,9 @@ void sa1111_driver_unregister(struct sa1111_driver *driver)
>  }
>  EXPORT_SYMBOL(sa1111_driver_unregister);
>
> -#ifdef CONFIG_DMABOUNCE
> -/*
> - * According to the "Intel StrongARM SA-1111 Microprocessor Companion
> - * Chip Specification Update" (June 2000), erratum #7, there is a
> - * significant bug in the SA1111 SDRAM shared memory controller.  If
> - * an access to a region of memory above 1MB relative to the bank base,
> - * it is important that address bit 10 _NOT_ be asserted. Depending
> - * on the configuration of the RAM, bit 10 may correspond to one
> - * of several different (processor-relative) address bits.
> - *
> - * This routine only identifies whether or not a given DMA address
> - * is susceptible to the bug.
> - *
> - * This should only get called for sa1111_device types due to the
> - * way we configure our device dma_masks.
> - */
> -static int sa1111_needs_bounce(struct device *dev, dma_addr_t addr, size_t size)
> -{
> -       /*
> -        * Section 4.6 of the "Intel StrongARM SA-1111 Development Module
> -        * User's Guide" mentions that jumpers R51 and R52 control the
> -        * target of SA-1111 DMA (either SDRAM bank 0 on Assabet, or
> -        * SDRAM bank 1 on Neponset). The default configuration selects
> -        * Assabet, so any address in bank 1 is necessarily invalid.
> -        */
> -       return (machine_is_assabet() || machine_is_pfs168()) &&
> -               (addr >= 0xc8000000 || (addr + size) >= 0xc8000000);
> -}
> -
> -static int sa1111_notifier_call(struct notifier_block *n, unsigned long action,
> -       void *data)
> -{
> -       struct sa1111_dev *dev = to_sa1111_device(data);
> -
> -       switch (action) {
> -       case BUS_NOTIFY_ADD_DEVICE:
> -               if (dev->dev.dma_mask && dev->dma_mask < 0xffffffffUL) {
> -                       int ret = dmabounce_register_dev(&dev->dev, 1024, 4096,
> -                                       sa1111_needs_bounce);
> -                       if (ret)
> -                               dev_err(&dev->dev, "failed to register with dmabounce: %d\n", ret);
> -               }
> -               break;
> -
> -       case BUS_NOTIFY_DEL_DEVICE:
> -               if (dev->dev.dma_mask && dev->dma_mask < 0xffffffffUL)
> -                       dmabounce_unregister_dev(&dev->dev);
> -               break;
> -       }
> -       return NOTIFY_OK;
> -}
> -
> -static struct notifier_block sa1111_bus_notifier = {
> -       .notifier_call = sa1111_notifier_call,
> -};
> -#endif
> -
>  static int __init sa1111_init(void)
>  {
>         int ret = bus_register(&sa1111_bus_type);
> -#ifdef CONFIG_DMABOUNCE
> -       if (ret == 0)
> -               bus_register_notifier(&sa1111_bus_type, &sa1111_bus_notifier);
> -#endif
>         if (ret == 0)
>                 platform_driver_register(&sa1111_device_driver);
>         return ret;
> @@ -1463,9 +1402,6 @@ static int __init sa1111_init(void)
>  static void __exit sa1111_exit(void)
>  {
>         platform_driver_unregister(&sa1111_device_driver);
> -#ifdef CONFIG_DMABOUNCE
> -       bus_unregister_notifier(&sa1111_bus_type, &sa1111_bus_notifier);
> -#endif
>         bus_unregister(&sa1111_bus_type);
>  }
>
> diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c
> index 3c7c64ff3c0a..8417baedc89c 100644
> --- a/drivers/usb/core/hcd.c
> +++ b/drivers/usb/core/hcd.c
> @@ -1260,7 +1260,8 @@ void usb_hcd_unlink_urb_from_ep(struct usb_hcd *hcd, struct urb *urb)
>  EXPORT_SYMBOL_GPL(usb_hcd_unlink_urb_from_ep);
>
>  /*
> - * Some usb host controllers can only perform dma using a small SRAM area.
> + * Some usb host controllers can only perform dma using a small SRAM area,
> + * or have restrictions on addressable DRAM.
>   * The usb core itself is however optimized for host controllers that can dma
>   * using regular system memory - like pci devices doing bus mastering.
>   *
> @@ -3095,8 +3096,18 @@ int usb_hcd_setup_local_mem(struct usb_hcd *hcd, phys_addr_t phys_addr,
>         if (IS_ERR(hcd->localmem_pool))
>                 return PTR_ERR(hcd->localmem_pool);
>
> -       local_mem = devm_memremap(hcd->self.sysdev, phys_addr,
> -                                 size, MEMREMAP_WC);
> +       /*
> +        * if a physical SRAM address was passed, map it, otherwise
> +        * allocate system memory as a buffer.
> +        */
> +       if (phys_addr)
> +               local_mem = devm_memremap(hcd->self.sysdev, phys_addr,
> +                                         size, MEMREMAP_WC);
> +       else
> +               local_mem = dmam_alloc_attrs(hcd->self.sysdev, size, &dma,
> +                                            GFP_KERNEL,
> +                                            DMA_ATTR_WRITE_COMBINE);
> +
>         if (IS_ERR(local_mem))
>                 return PTR_ERR(local_mem);
>
> diff --git a/drivers/usb/host/ohci-sa1111.c b/drivers/usb/host/ohci-sa1111.c
> index 137f66f6977f..0da2badf0658 100644
> --- a/drivers/usb/host/ohci-sa1111.c
> +++ b/drivers/usb/host/ohci-sa1111.c
> @@ -206,6 +206,31 @@ static int ohci_hcd_sa1111_probe(struct sa1111_dev *dev)
>                 goto err1;
>         }
>
> +       /*
> +        * According to the "Intel StrongARM SA-1111 Microprocessor Companion
> +        * Chip Specification Update" (June 2000), erratum #7, there is a
> +        * significant bug in the SA1111 SDRAM shared memory controller.  If
> +        * an access to a region of memory above 1MB relative to the bank base,
> +        * it is important that address bit 10 _NOT_ be asserted. Depending
> +        * on the configuration of the RAM, bit 10 may correspond to one
> +        * of several different (processor-relative) address bits.
> +        *
> +        * Section 4.6 of the "Intel StrongARM SA-1111 Development Module
> +        * User's Guide" mentions that jumpers R51 and R52 control the
> +        * target of SA-1111 DMA (either SDRAM bank 0 on Assabet, or
> +        * SDRAM bank 1 on Neponset). The default configuration selects
> +        * Assabet, so any address in bank 1 is necessarily invalid.
> +        *
> +        * As a workaround, use a bounce buffer in addressable memory
> +        * as local_mem, relying on ZONE_DMA to provide an area that
> +        * fits within the above constraints.
> +        *
> +        * SZ_64K is an estimate for what size this might need.
> +        */
> +       ret = usb_hcd_setup_local_mem(hcd, 0, 0, SZ_64K);
> +       if (ret)
> +               goto err1;
> +
>         if (!request_mem_region(hcd->rsrc_start, hcd->rsrc_len, hcd_name)) {
>                 dev_dbg(&dev->dev, "request_mem_region failed\n");
>                 ret = -EBUSY;
> --
> 2.29.2
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Arnd Bergmann Feb. 8, 2022, 12:49 p.m. UTC | #2
On Thu, Feb 3, 2022 at 9:47 AM Ard Biesheuvel <ardb@kernel.org> wrote:
> On Thu, 3 Feb 2022 at 09:38, Arnd Bergmann <arnd@kernel.org> wrote:
>
> > There are two main downsides:
> >
> >  - rather than using a dynamically sized pool, this buffer needs
> >    to be allocated at probe time using a fixed size. Without
> >    having any idea of what it should be, I picked a size of
> >    64KB, which is between what the other two OHCI front-ends use
> >    in their SRAM. If anyone has a better idea what that size
> >    is reasonable, this can be trivially changed.
> >
>
> I suppose this is a problem if the driver falls back to ordinary DRAM
> once the allocation runs out?
Ard Biesheuvel Feb. 8, 2022, 1:35 p.m. UTC | #3
On Tue, 8 Feb 2022 at 13:49, Arnd Bergmann <arnd@kernel.org> wrote:
>
> On Thu, Feb 3, 2022 at 9:47 AM Ard Biesheuvel <ardb@kernel.org> wrote:
> > On Thu, 3 Feb 2022 at 09:38, Arnd Bergmann <arnd@kernel.org> wrote:
> >
> > > There are two main downsides:
> > >
> > >  - rather than using a dynamically sized pool, this buffer needs
> > >    to be allocated at probe time using a fixed size. Without
> > >    having any idea of what it should be, I picked a size of
> > >    64KB, which is between what the other two OHCI front-ends use
> > >    in their SRAM. If anyone has a better idea what that size
> > >    is reasonable, this can be trivially changed.
> > >
> >
> > I suppose this is a problem if the driver falls back to ordinary DRAM
> > once the allocation runs out?
>
> From what I can tell, there is no such fallback. If the localmem_pool
> runs out, the allocation fails, which may cause other problems, but
> it never falls back to the wrong DMA address.
>

OK that is the least bad outcome I suppose.

> > >  - Previously, only USB transfers to unaddressable memory needed
> > >    to go through the bounce buffer, now all of them do, which may
> > >    impact runtime performance for USB endpoints that do a lot of
> > >    transfers.
> > >
> > > On the upside, the local_mem support uses write-combining buffers,
> > > which should be a bit faster for transfers to the device compared to
> > > normal uncached coherent memory as used in dmabounce.
> > >
> >
> > Talking from past experience using this trick on a NXP ARM9 SoC ~10
> > years ago, using on-chip SRAM for USB DMA likely results in a
> > significant performance boost, even without write combining, although
> > the exact scenario obviously matters.
>
> Right, that makes sense, but it won't help here because there is
> no SRAM. One detail  I noticed is that the localmem pool normally
> gets mapped as WC, which is what I did in the new code as well, but
> dma_alloc_flags(..., DMA_ATTR_WRITE_COMBINE) does not always
> honor this flag. I think it will do it here because a GFP_KERNEL
> allocation should be served by the remap_allocator, while
> GFP_ATOMIC allocations would be served by pool_allocator_alloc(),
> which ignores the flag.
>

Ah yes, ignore me. For some reason, I thought this was about on-chip SRAM.
diff mbox series

Patch

diff --git a/arch/arm/common/Kconfig b/arch/arm/common/Kconfig
index c8e198631d41..bc158fd227e1 100644
--- a/arch/arm/common/Kconfig
+++ b/arch/arm/common/Kconfig
@@ -1,7 +1,7 @@ 
 # SPDX-License-Identifier: GPL-2.0
 config SA1111
 	bool
-	select DMABOUNCE if !ARCH_PXA
+	select ZONE_DMA if ARCH_SA1100
 
 config DMABOUNCE
 	bool
diff --git a/arch/arm/common/sa1111.c b/arch/arm/common/sa1111.c
index 7df003b149c6..a00915883f78 100644
--- a/arch/arm/common/sa1111.c
+++ b/arch/arm/common/sa1111.c
@@ -1391,70 +1391,9 @@  void sa1111_driver_unregister(struct sa1111_driver *driver)
 }
 EXPORT_SYMBOL(sa1111_driver_unregister);
 
-#ifdef CONFIG_DMABOUNCE
-/*
- * According to the "Intel StrongARM SA-1111 Microprocessor Companion
- * Chip Specification Update" (June 2000), erratum #7, there is a
- * significant bug in the SA1111 SDRAM shared memory controller.  If
- * an access to a region of memory above 1MB relative to the bank base,
- * it is important that address bit 10 _NOT_ be asserted. Depending
- * on the configuration of the RAM, bit 10 may correspond to one
- * of several different (processor-relative) address bits.
- *
- * This routine only identifies whether or not a given DMA address
- * is susceptible to the bug.
- *
- * This should only get called for sa1111_device types due to the
- * way we configure our device dma_masks.
- */
-static int sa1111_needs_bounce(struct device *dev, dma_addr_t addr, size_t size)
-{
-	/*
-	 * Section 4.6 of the "Intel StrongARM SA-1111 Development Module
-	 * User's Guide" mentions that jumpers R51 and R52 control the
-	 * target of SA-1111 DMA (either SDRAM bank 0 on Assabet, or
-	 * SDRAM bank 1 on Neponset). The default configuration selects
-	 * Assabet, so any address in bank 1 is necessarily invalid.
-	 */
-	return (machine_is_assabet() || machine_is_pfs168()) &&
-		(addr >= 0xc8000000 || (addr + size) >= 0xc8000000);
-}
-
-static int sa1111_notifier_call(struct notifier_block *n, unsigned long action,
-	void *data)
-{
-	struct sa1111_dev *dev = to_sa1111_device(data);
-
-	switch (action) {
-	case BUS_NOTIFY_ADD_DEVICE:
-		if (dev->dev.dma_mask && dev->dma_mask < 0xffffffffUL) {
-			int ret = dmabounce_register_dev(&dev->dev, 1024, 4096,
-					sa1111_needs_bounce);
-			if (ret)
-				dev_err(&dev->dev, "failed to register with dmabounce: %d\n", ret);
-		}
-		break;
-
-	case BUS_NOTIFY_DEL_DEVICE:
-		if (dev->dev.dma_mask && dev->dma_mask < 0xffffffffUL)
-			dmabounce_unregister_dev(&dev->dev);
-		break;
-	}
-	return NOTIFY_OK;
-}
-
-static struct notifier_block sa1111_bus_notifier = {
-	.notifier_call = sa1111_notifier_call,
-};
-#endif
-
 static int __init sa1111_init(void)
 {
 	int ret = bus_register(&sa1111_bus_type);
-#ifdef CONFIG_DMABOUNCE
-	if (ret == 0)
-		bus_register_notifier(&sa1111_bus_type, &sa1111_bus_notifier);
-#endif
 	if (ret == 0)
 		platform_driver_register(&sa1111_device_driver);
 	return ret;
@@ -1463,9 +1402,6 @@  static int __init sa1111_init(void)
 static void __exit sa1111_exit(void)
 {
 	platform_driver_unregister(&sa1111_device_driver);
-#ifdef CONFIG_DMABOUNCE
-	bus_unregister_notifier(&sa1111_bus_type, &sa1111_bus_notifier);
-#endif
 	bus_unregister(&sa1111_bus_type);
 }
 
diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c
index 3c7c64ff3c0a..8417baedc89c 100644
--- a/drivers/usb/core/hcd.c
+++ b/drivers/usb/core/hcd.c
@@ -1260,7 +1260,8 @@  void usb_hcd_unlink_urb_from_ep(struct usb_hcd *hcd, struct urb *urb)
 EXPORT_SYMBOL_GPL(usb_hcd_unlink_urb_from_ep);
 
 /*
- * Some usb host controllers can only perform dma using a small SRAM area.
+ * Some usb host controllers can only perform dma using a small SRAM area,
+ * or have restrictions on addressable DRAM.
  * The usb core itself is however optimized for host controllers that can dma
  * using regular system memory - like pci devices doing bus mastering.
  *
@@ -3095,8 +3096,18 @@  int usb_hcd_setup_local_mem(struct usb_hcd *hcd, phys_addr_t phys_addr,
 	if (IS_ERR(hcd->localmem_pool))
 		return PTR_ERR(hcd->localmem_pool);
 
-	local_mem = devm_memremap(hcd->self.sysdev, phys_addr,
-				  size, MEMREMAP_WC);
+	/*
+	 * if a physical SRAM address was passed, map it, otherwise
+	 * allocate system memory as a buffer.
+	 */
+	if (phys_addr)
+		local_mem = devm_memremap(hcd->self.sysdev, phys_addr,
+					  size, MEMREMAP_WC);
+	else
+		local_mem = dmam_alloc_attrs(hcd->self.sysdev, size, &dma,
+					     GFP_KERNEL,
+					     DMA_ATTR_WRITE_COMBINE);
+
 	if (IS_ERR(local_mem))
 		return PTR_ERR(local_mem);
 
diff --git a/drivers/usb/host/ohci-sa1111.c b/drivers/usb/host/ohci-sa1111.c
index 137f66f6977f..0da2badf0658 100644
--- a/drivers/usb/host/ohci-sa1111.c
+++ b/drivers/usb/host/ohci-sa1111.c
@@ -206,6 +206,31 @@  static int ohci_hcd_sa1111_probe(struct sa1111_dev *dev)
 		goto err1;
 	}
 
+	/*
+	 * According to the "Intel StrongARM SA-1111 Microprocessor Companion
+	 * Chip Specification Update" (June 2000), erratum #7, there is a
+	 * significant bug in the SA1111 SDRAM shared memory controller.  If
+	 * an access to a region of memory above 1MB relative to the bank base,
+	 * it is important that address bit 10 _NOT_ be asserted. Depending
+	 * on the configuration of the RAM, bit 10 may correspond to one
+	 * of several different (processor-relative) address bits.
+	 *
+	 * Section 4.6 of the "Intel StrongARM SA-1111 Development Module
+	 * User's Guide" mentions that jumpers R51 and R52 control the
+	 * target of SA-1111 DMA (either SDRAM bank 0 on Assabet, or
+	 * SDRAM bank 1 on Neponset). The default configuration selects
+	 * Assabet, so any address in bank 1 is necessarily invalid.
+	 *
+	 * As a workaround, use a bounce buffer in addressable memory
+	 * as local_mem, relying on ZONE_DMA to provide an area that
+	 * fits within the above constraints.
+	 *
+	 * SZ_64K is an estimate for what size this might need.
+	 */
+	ret = usb_hcd_setup_local_mem(hcd, 0, 0, SZ_64K);
+	if (ret)
+		goto err1;
+
 	if (!request_mem_region(hcd->rsrc_start, hcd->rsrc_len, hcd_name)) {
 		dev_dbg(&dev->dev, "request_mem_region failed\n");
 		ret = -EBUSY;