| Message ID | 1471435517-7840-2-git-send-email-rogerq@ti.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Aug 17, 2016 at 03:05:17PM +0300, Roger Quadros wrote: > Since commit 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation"), > dma_to_pfn() already returns the PFN with the physical memory start offset > so we don't need to add it again. > > This fixes USB mass storage lock-up problem on systems that can't do DMA > over the entire physical memory range (e.g.) Keystone 2 systems with 4GB RAM > can only do DMA over the first 2GB. [K2E-EVM]. > > What happens there is that without this patch SCSI layer sets a wrong > bounce buffer limit in scsi_calculate_bounce_limit() for the USB mass > storage device. dma_max_pfn() evaluates to 0x8fffff and bounce_limit > is set to 0x8fffff000 whereas maximum DMA'ble physical memory on Keystone 2 > is 0x87fffffff. This results in non DMA'ble pages being given to the > USB controller and hence the lock-up. > > NOTE: in the above case, USB-SCSI-device's dma_pfn_offset was showing as 0. > This should have really been 0x780000 as on K2e, LOWMEM_START is 0x80000000 > and HIGHMEM_START is 0x800000000. DMA zone is 2GB so dma_max_pfn should be > 0x87ffff. The incorrect dma_pfn_offset for the USB storage device is because > USB devices are not correctly inheriting the dma_pfn_offset from the > USB host controller. This will be fixed by a separate patch. I'd like to hear from Santosh, as the author of the original change. The original commit doesn't mention which platform it was intended for or what the problem was, which would've been helpful. > > Fixes: 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation") > Cc: stable@vger.kernel.org > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> > Cc: Russell King <linux@arm.linux.org.uk> > Cc: Arnd Bergmann <arnd@arndb.de> > Cc: Olof Johansson <olof@lixom.net> > Cc: Catalin Marinas <catalin.marinas@arm.com> > Cc: Linus Walleij <linus.walleij@linaro.org> > Reported-by: Grygorii Strashko <grygorii.strashko@ti.com> > Signed-off-by: Roger Quadros <rogerq@ti.com> > --- > arch/arm/include/asm/dma-mapping.h | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/arch/arm/include/asm/dma-mapping.h b/arch/arm/include/asm/dma-mapping.h > index d009f79..bf02dbd 100644 > --- a/arch/arm/include/asm/dma-mapping.h > +++ b/arch/arm/include/asm/dma-mapping.h > @@ -111,7 +111,7 @@ static inline dma_addr_t virt_to_dma(struct device *dev, void *addr) > /* The ARM override for dma_max_pfn() */ > static inline unsigned long dma_max_pfn(struct device *dev) > { > - return PHYS_PFN_OFFSET + dma_to_pfn(dev, *dev->dma_mask); > + return dma_to_pfn(dev, *dev->dma_mask); > } > #define dma_max_pfn(dev) dma_max_pfn(dev) > > -- > 2.7.4 >
Hi Russell, On 8/18/2016 7:24 AM, Russell King - ARM Linux wrote: > On Wed, Aug 17, 2016 at 03:05:17PM +0300, Roger Quadros wrote: >> Since commit 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation"), >> dma_to_pfn() already returns the PFN with the physical memory start offset >> so we don't need to add it again. >> >> This fixes USB mass storage lock-up problem on systems that can't do DMA >> over the entire physical memory range (e.g.) Keystone 2 systems with 4GB RAM >> can only do DMA over the first 2GB. [K2E-EVM]. >> >> What happens there is that without this patch SCSI layer sets a wrong >> bounce buffer limit in scsi_calculate_bounce_limit() for the USB mass >> storage device. dma_max_pfn() evaluates to 0x8fffff and bounce_limit >> is set to 0x8fffff000 whereas maximum DMA'ble physical memory on Keystone 2 >> is 0x87fffffff. This results in non DMA'ble pages being given to the >> USB controller and hence the lock-up. >> >> NOTE: in the above case, USB-SCSI-device's dma_pfn_offset was showing as 0. >> This should have really been 0x780000 as on K2e, LOWMEM_START is 0x80000000 >> and HIGHMEM_START is 0x800000000. DMA zone is 2GB so dma_max_pfn should be >> 0x87ffff. The incorrect dma_pfn_offset for the USB storage device is because >> USB devices are not correctly inheriting the dma_pfn_offset from the >> USB host controller. This will be fixed by a separate patch. > > I'd like to hear from Santosh, as the author of the original change. > The original commit doesn't mention which platform it was intended for > or what the problem was, which would've been helpful. > From what I recollect, we did these changes to make the max pfn behave same on ARM arch as other archs. This patch was evolved as part of fixing the max*pfn assumption. commit 26ba47b18318abe7dadbe9294a611c0e932651d8 Author: Santosh Shilimkar <santosh.shilimkar@ti.com> Date: Thu Aug 1 03:12:01 2013 +0100 ARM: 7805/1: mm: change max*pfn to include the physical offset of memory Most of the kernel code assumes that max*pfn is maximum pfns because the physical start of memory is expected to be PFN0. Since this assumption is not true on ARM architectures, the meaning of max*pfn is number of memory pages. This is done to keep drivers happy which are making use of of these variable to calculate the dma bounce limit using dma_mask. Now since we have a architecture override possibility for DMAable maximum pfns, lets make meaning of max*pfns as maximum pnfs on ARM as well. >> >> Fixes: 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation") >> Cc: stable@vger.kernel.org >> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> >> Cc: Russell King <linux@arm.linux.org.uk> >> Cc: Arnd Bergmann <arnd@arndb.de> >> Cc: Olof Johansson <olof@lixom.net> >> Cc: Catalin Marinas <catalin.marinas@arm.com> >> Cc: Linus Walleij <linus.walleij@linaro.org> >> Reported-by: Grygorii Strashko <grygorii.strashko@ti.com> >> Signed-off-by: Roger Quadros <rogerq@ti.com> >> --- >> arch/arm/include/asm/dma-mapping.h | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/arch/arm/include/asm/dma-mapping.h b/arch/arm/include/asm/dma-mapping.h >> index d009f79..bf02dbd 100644 >> --- a/arch/arm/include/asm/dma-mapping.h >> +++ b/arch/arm/include/asm/dma-mapping.h >> @@ -111,7 +111,7 @@ static inline dma_addr_t virt_to_dma(struct device *dev, void *addr) >> /* The ARM override for dma_max_pfn() */ >> static inline unsigned long dma_max_pfn(struct device *dev) >> { >> - return PHYS_PFN_OFFSET + dma_to_pfn(dev, *dev->dma_mask); >> + return dma_to_pfn(dev, *dev->dma_mask); >> } >> #define dma_max_pfn(dev) dma_max_pfn(dev) By doing this change I hope we don't break other drivers on Keystone so am not sure about the change.
On Thu, Aug 18, 2016 at 09:55:55AM -0700, Santosh Shilimkar wrote: > Hi Russell, > > On 8/18/2016 7:24 AM, Russell King - ARM Linux wrote: > >On Wed, Aug 17, 2016 at 03:05:17PM +0300, Roger Quadros wrote: > >>Since commit 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation"), > >>dma_to_pfn() already returns the PFN with the physical memory start offset > >>so we don't need to add it again. > >> > >>This fixes USB mass storage lock-up problem on systems that can't do DMA > >>over the entire physical memory range (e.g.) Keystone 2 systems with 4GB RAM > >>can only do DMA over the first 2GB. [K2E-EVM]. > >> > >>What happens there is that without this patch SCSI layer sets a wrong > >>bounce buffer limit in scsi_calculate_bounce_limit() for the USB mass > >>storage device. dma_max_pfn() evaluates to 0x8fffff and bounce_limit > >>is set to 0x8fffff000 whereas maximum DMA'ble physical memory on Keystone 2 > >>is 0x87fffffff. This results in non DMA'ble pages being given to the > >>USB controller and hence the lock-up. > >> > >>NOTE: in the above case, USB-SCSI-device's dma_pfn_offset was showing as 0. > >>This should have really been 0x780000 as on K2e, LOWMEM_START is 0x80000000 > >>and HIGHMEM_START is 0x800000000. DMA zone is 2GB so dma_max_pfn should be > >>0x87ffff. The incorrect dma_pfn_offset for the USB storage device is because > >>USB devices are not correctly inheriting the dma_pfn_offset from the > >>USB host controller. This will be fixed by a separate patch. > > > >I'd like to hear from Santosh, as the author of the original change. > >The original commit doesn't mention which platform it was intended for > >or what the problem was, which would've been helpful. > > > From what I recollect, we did these changes to make the max pfn behave > same on ARM arch as other archs. This patch was evolved as part of > fixing the max*pfn assumption. To me, the proposed patch _looks_ correct, because... > >>diff --git a/arch/arm/include/asm/dma-mapping.h b/arch/arm/include/asm/dma-mapping.h > >>index d009f79..bf02dbd 100644 > >>--- a/arch/arm/include/asm/dma-mapping.h > >>+++ b/arch/arm/include/asm/dma-mapping.h > >>@@ -111,7 +111,7 @@ static inline dma_addr_t virt_to_dma(struct device *dev, void *addr) > >> /* The ARM override for dma_max_pfn() */ > >> static inline unsigned long dma_max_pfn(struct device *dev) > >> { > >>- return PHYS_PFN_OFFSET + dma_to_pfn(dev, *dev->dma_mask); > >>+ return dma_to_pfn(dev, *dev->dma_mask); > >> } > >> #define dma_max_pfn(dev) dma_max_pfn(dev) > By doing this change I hope we don't break other drivers on Keystone so > am not sure about the change. dma_to_pfn() returns the page frame number referenced from physical address zero - the default implementation of dma_to_pfn() is bus_to_pfn(), which is __phys_to_pfn(x), which is just x >> PAGE_SHIFT. The other thing about dma_to_pfn() is that it should return a zero-referenced PFN number, where PFN 0 = physical address 0. If there is some offset for keystone2, that should be taken care of via "dev->dma_pfn_offset", and not offsetting the return value from dma_to_pfn(). So I'm 99.9% convinced that the proposed change is correct.
On 8/18/2016 4:07 PM, Russell King - ARM Linux wrote: > On Thu, Aug 18, 2016 at 09:55:55AM -0700, Santosh Shilimkar wrote: >> Hi Russell, >> >> On 8/18/2016 7:24 AM, Russell King - ARM Linux wrote: >>> On Wed, Aug 17, 2016 at 03:05:17PM +0300, Roger Quadros wrote: >>>> Since commit 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation"), >>>> dma_to_pfn() already returns the PFN with the physical memory start offset >>>> so we don't need to add it again. >>>> >>>> This fixes USB mass storage lock-up problem on systems that can't do DMA >>>> over the entire physical memory range (e.g.) Keystone 2 systems with 4GB RAM >>>> can only do DMA over the first 2GB. [K2E-EVM]. >>>> >>>> What happens there is that without this patch SCSI layer sets a wrong >>>> bounce buffer limit in scsi_calculate_bounce_limit() for the USB mass >>>> storage device. dma_max_pfn() evaluates to 0x8fffff and bounce_limit >>>> is set to 0x8fffff000 whereas maximum DMA'ble physical memory on Keystone 2 >>>> is 0x87fffffff. This results in non DMA'ble pages being given to the >>>> USB controller and hence the lock-up. >>>> >>>> NOTE: in the above case, USB-SCSI-device's dma_pfn_offset was showing as 0. >>>> This should have really been 0x780000 as on K2e, LOWMEM_START is 0x80000000 >>>> and HIGHMEM_START is 0x800000000. DMA zone is 2GB so dma_max_pfn should be >>>> 0x87ffff. The incorrect dma_pfn_offset for the USB storage device is because >>>> USB devices are not correctly inheriting the dma_pfn_offset from the >>>> USB host controller. This will be fixed by a separate patch. >>> >>> I'd like to hear from Santosh, as the author of the original change. >>> The original commit doesn't mention which platform it was intended for >>> or what the problem was, which would've been helpful. >>> >> From what I recollect, we did these changes to make the max pfn behave >> same on ARM arch as other archs. This patch was evolved as part of >> fixing the max*pfn assumption. > > To me, the proposed patch _looks_ correct, because... > >>>> diff --git a/arch/arm/include/asm/dma-mapping.h b/arch/arm/include/asm/dma-mapping.h >>>> index d009f79..bf02dbd 100644 >>>> --- a/arch/arm/include/asm/dma-mapping.h >>>> +++ b/arch/arm/include/asm/dma-mapping.h >>>> @@ -111,7 +111,7 @@ static inline dma_addr_t virt_to_dma(struct device *dev, void *addr) >>>> /* The ARM override for dma_max_pfn() */ >>>> static inline unsigned long dma_max_pfn(struct device *dev) >>>> { >>>> - return PHYS_PFN_OFFSET + dma_to_pfn(dev, *dev->dma_mask); >>>> + return dma_to_pfn(dev, *dev->dma_mask); >>>> } >>>> #define dma_max_pfn(dev) dma_max_pfn(dev) >> By doing this change I hope we don't break other drivers on Keystone so >> am not sure about the change. > > dma_to_pfn() returns the page frame number referenced from physical > address zero - the default implementation of dma_to_pfn() is > bus_to_pfn(), which is __phys_to_pfn(x), which is just x >> PAGE_SHIFT. > The other thing about dma_to_pfn() is that it should return a > zero-referenced PFN number, where PFN 0 = physical address 0. > > If there is some offset for keystone2, that should be taken care of > via "dev->dma_pfn_offset", and not offsetting the return value from > dma_to_pfn(). > > So I'm 99.9% convinced that the proposed change is correct. > I will got with that then :-) and take my objection back. Just saying that if there other breakages which I can't recollect now, those drivers needs to be patched as well.
Hi Santosh, On 19/08/16 05:01, Santosh Shilimkar wrote: > On 8/18/2016 4:07 PM, Russell King - ARM Linux wrote: >> On Thu, Aug 18, 2016 at 09:55:55AM -0700, Santosh Shilimkar wrote: >>> Hi Russell, >>> >>> On 8/18/2016 7:24 AM, Russell King - ARM Linux wrote: >>>> On Wed, Aug 17, 2016 at 03:05:17PM +0300, Roger Quadros wrote: >>>>> Since commit 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation"), >>>>> dma_to_pfn() already returns the PFN with the physical memory start offset >>>>> so we don't need to add it again. >>>>> >>>>> This fixes USB mass storage lock-up problem on systems that can't do DMA >>>>> over the entire physical memory range (e.g.) Keystone 2 systems with 4GB RAM >>>>> can only do DMA over the first 2GB. [K2E-EVM]. >>>>> >>>>> What happens there is that without this patch SCSI layer sets a wrong >>>>> bounce buffer limit in scsi_calculate_bounce_limit() for the USB mass >>>>> storage device. dma_max_pfn() evaluates to 0x8fffff and bounce_limit >>>>> is set to 0x8fffff000 whereas maximum DMA'ble physical memory on Keystone 2 >>>>> is 0x87fffffff. This results in non DMA'ble pages being given to the >>>>> USB controller and hence the lock-up. >>>>> >>>>> NOTE: in the above case, USB-SCSI-device's dma_pfn_offset was showing as 0. >>>>> This should have really been 0x780000 as on K2e, LOWMEM_START is 0x80000000 >>>>> and HIGHMEM_START is 0x800000000. DMA zone is 2GB so dma_max_pfn should be >>>>> 0x87ffff. The incorrect dma_pfn_offset for the USB storage device is because >>>>> USB devices are not correctly inheriting the dma_pfn_offset from the >>>>> USB host controller. This will be fixed by a separate patch. >>>> >>>> I'd like to hear from Santosh, as the author of the original change. >>>> The original commit doesn't mention which platform it was intended for >>>> or what the problem was, which would've been helpful. >>>> >>> From what I recollect, we did these changes to make the max pfn behave >>> same on ARM arch as other archs. This patch was evolved as part of >>> fixing the max*pfn assumption. >> >> To me, the proposed patch _looks_ correct, because... >> >>>>> diff --git a/arch/arm/include/asm/dma-mapping.h b/arch/arm/include/asm/dma-mapping.h >>>>> index d009f79..bf02dbd 100644 >>>>> --- a/arch/arm/include/asm/dma-mapping.h >>>>> +++ b/arch/arm/include/asm/dma-mapping.h >>>>> @@ -111,7 +111,7 @@ static inline dma_addr_t virt_to_dma(struct device *dev, void *addr) >>>>> /* The ARM override for dma_max_pfn() */ >>>>> static inline unsigned long dma_max_pfn(struct device *dev) >>>>> { >>>>> - return PHYS_PFN_OFFSET + dma_to_pfn(dev, *dev->dma_mask); >>>>> + return dma_to_pfn(dev, *dev->dma_mask); >>>>> } >>>>> #define dma_max_pfn(dev) dma_max_pfn(dev) >>> By doing this change I hope we don't break other drivers on Keystone so >>> am not sure about the change. >> >> dma_to_pfn() returns the page frame number referenced from physical >> address zero - the default implementation of dma_to_pfn() is >> bus_to_pfn(), which is __phys_to_pfn(x), which is just x >> PAGE_SHIFT. >> The other thing about dma_to_pfn() is that it should return a >> zero-referenced PFN number, where PFN 0 = physical address 0. >> >> If there is some offset for keystone2, that should be taken care of >> via "dev->dma_pfn_offset", and not offsetting the return value from >> dma_to_pfn(). >> >> So I'm 99.9% convinced that the proposed change is correct. >> > I will got with that then :-) and take my objection back. Just > saying that if there other breakages which I can't recollect now, > those drivers needs to be patched as well. > I was able to boot the Keystone2 Edison EVM over NFS with the $subject patch. Boot log is below. Do you see anything suspicious? --- U-Boot SPL 2016.05-00100-g052f3bf-dirty (Aug 10 2016 - 12:28:40) Trying to boot from SPI U-Boot 2016.05-00100-g052f3bf-dirty (Aug 10 2016 - 12:28:40 +0300) CPU: 66AK2Ex SR1.0 I2C: ready DRAM: DDR3A Speed will be configured for 1600 Operation. Detected SO-DIMM [18KSF51272HZ-1G6K2] DDR3 speed 1600 DRAM: 4 GiB Clear entire DDR3 memory to enable ECC 2 GiB NAND: 512 MiB Net: eth0: netcp@24000000 Hit any key to stop autoboot: 0 netcp@24000000 Waiting for SGMII auto negotiation to complete. done Using netcp@24000000 device TFTP from server 192.168.1.36; our IP address is 192.168.1.90 Filename '/tftpboot/k2-fw-initrd.cpio.gz'. Load address: 0x88080000 Loading: ########## 958 KiB/s done Bytes transferred = 48117 (bbf5 hex) netcp@24000000 Waiting for SGMII auto negotiation to complete. done Using netcp@24000000 device TFTP from server 192.168.1.36; our IP address is 192.168.1.90 Filename 'keystone-k2e-evm.dtb'. Load address: 0x88000000 Loading: ##### 884.8 KiB/s done Bytes transferred = 22660 (5884 hex) netcp@24000000 Waiting for SGMII auto negotiation to complete. done Using netcp@24000000 device TFTP from server 192.168.1.36; our IP address is 192.168.1.90 Filename '/tftpboot/skern-k2e.bin'. Load address: 0xc140000 Loading: ######### 956.1 KiB/s done Bytes transferred = 45056 (b000 hex) netcp@24000000 Waiting for SGMII auto negotiation to complete. done Using netcp@24000000 device TFTP from server 192.168.1.36; our IP address is 192.168.1.90 Filename 'zImage'. Load address: 0x82000000 Loading: ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# #################### 1 MiB/s done Bytes transferred = 4427600 (438f50 hex) K2_BM_15.07-39-g035329c SoC:k2e built:10:59:38, Jul 28 2016 ## installed monitor, freq [208333333], status 0 Kernel image @ 0x82000000 [ 0x000000 - 0x438f50 ] ## Flattened Device Tree blob at 88000000 Booting using the fdt blob at 0x88000000 Using Device Tree in place at 88000000, end 88008883 80000000 800000000 800000000:80000000 880000000:80000000 Starting kernel ... [ 0.000000] Booting Linux on physical CPU 0x0 [ 0.000000] Linux version 4.8.0-rc2-00003-ga0a0676 (roger@lta0400828d) (gcc version 4.7.3 20130226 (prerelease) (crosstool-NG linaro-1.13.1-4.7-2013.03-20130313 - Linaro GCC 2013.03) ) #974 SMP PREEMPT Fri A6 [ 0.000000] CPU: ARMv7 Processor [412fc0f4] revision 4 (ARMv7), cr=30c5387d [ 0.000000] CPU: div instructions available: patching division code [ 0.000000] CPU: PIPT / VIPT nonaliasing data cache, PIPT instruction cache [ 0.000000] OF: fdt:Machine model: Texas Instruments Keystone 2 Edison EVM [ 0.000000] Switching physical address space to 0x800000000 [ 0.000000] cma: Reserved 16 MiB at 0x000000087ec00000 [ 0.000000] Forcing write-allocate cache policy for SMP [ 0.000000] Memory policy: Data cache writealloc [ 0.000000] psci: probing for conduit method from DT. [ 0.000000] psci: Using PSCI v0.1 Function IDs from DT [ 0.000000] percpu: Embedded 14 pages/cpu @edf9f000 s27648 r8192 d21504 u57344 [ 0.000000] Built 1 zonelists in Zone order, mobility grouping on. Total pages: 1047040 [ 0.000000] Kernel command line: console=ttyS0,115200n8 rootwait=1 rootfstype=nfs root=/dev/nfs rw nfsroot=192.168.1.36:/srv/nfsroot ip=dhcp [ 0.000000] log_buf_len individual max cpu contribution: 4096 bytes [ 0.000000] log_buf_len total cpu_extra contributions: 12288 bytes [ 0.000000] log_buf_len min size: 16384 bytes [ 0.000000] log_buf_len: 32768 bytes [ 0.000000] early log buf free: 14444(88%) [ 0.000000] PID hash table entries: 4096 (order: 2, 16384 bytes) [ 0.000000] Dentry cache hash table entries: 131072 (order: 7, 524288 bytes) [ 0.000000] Inode-cache hash table entries: 65536 (order: 6, 262144 bytes) [ 0.000000] Memory: 4131012K/4194304K available (6144K kernel code, 366K rwdata, 2008K rodata, 2048K init, 202K bss, 46908K reserved, 16384K cma-reserved, 3391488K highmem) [ 0.000000] Virtual kernel memory layout: [ 0.000000] vector : 0xffff0000 - 0xffff1000 ( 4 kB) [ 0.000000] fixmap : 0xffc00000 - 0xfff00000 (3072 kB) [ 0.000000] vmalloc : 0xf0800000 - 0xff800000 ( 240 MB) [ 0.000000] lowmem : 0xc0000000 - 0xf0000000 ( 768 MB) [ 0.000000] pkmap : 0xbfe00000 - 0xc0000000 ( 2 MB) [ 0.000000] modules : 0xbf000000 - 0xbfe00000 ( 14 MB) [ 0.000000] .text : 0xc0008000 - 0xc0800000 (8160 kB) [ 0.000000] .init : 0xc0c00000 - 0xc0e00000 (2048 kB) [ 0.000000] .data : 0xc0e00000 - 0xc0e5bbf0 ( 367 kB) [ 0.000000] .bss : 0xc0e5d000 - 0xc0e8f9e0 ( 203 kB) [ 0.000000] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=4, Nodes=1 [ 0.000000] Preemptible hierarchical RCU implementation. [ 0.000000] Build-time adjustment of leaf fanout to 32. [ 0.000000] NR_IRQS:16 nr_irqs:16 16 [ 0.000000] arm_arch_timer: Architected cp15 timer(s) running at 208.33MHz (virt). [ 0.000000] clocksource: arch_sys_counter: mask: 0xffffffffffffff max_cycles: 0x300c4cdf1c, max_idle_ns: 440795202901 ns [ 0.000003] sched_clock: 56 bits at 208MHz, resolution 4ns, wraps every 4398046511101ns [ 0.000011] Switching to timer-based delay loop, resolution 4ns [ 0.000129] keystone timer clock @208333333 Hz [ 0.000238] Console: colour dummy device 80x30 [ 0.000257] Calibrating delay loop (skipped), value calculated using timer frequency.. 416.66 BogoMIPS (lpj=2083333) [ 0.000267] pid_max: default: 4096 minimum: 301 [ 0.000328] Mount-cache hash table entries: 2048 (order: 1, 8192 bytes) [ 0.000336] Mountpoint-cache hash table entries: 2048 (order: 1, 8192 bytes) [ 0.000761] CPU: Testing write buffer coherency: ok [ 0.000918] /cpus/cpu@0 missing clock-frequency property [ 0.000934] /cpus/cpu@1 missing clock-frequency property [ 0.000950] /cpus/cpu@2 missing clock-frequency property [ 0.000967] /cpus/cpu@3 missing clock-frequency property [ 0.000975] CPU0: thread -1, cpu 0, socket 0, mpidr 80000000 [ 0.000987] Setting up static identity map for 0x80200000 - 0x80200130 [ 0.114753] CPU1: thread -1, cpu 1, socket 0, mpidr 80000001 [ 0.154800] CPU2: thread -1, cpu 2, socket 0, mpidr 80000002 [ 0.194844] CPU3: thread -1, cpu 3, socket 0, mpidr 80000003 [ 0.194930] Brought up 4 CPUs [ 0.194946] SMP: Total of 4 processors activated (1666.66 BogoMIPS). [ 0.194952] CPU: All CPU(s) started in SVC mode. [ 0.195471] devtmpfs: initialized [ 0.200375] VFP support v0.3: implementor 41 architecture 4 part 30 variant f rev 0 [ 0.200642] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns [ 0.200897] pinctrl core: initialized pinctrl subsystem [ 0.201164] NET: Registered protocol family 16 [ 0.201903] DMA: preallocated 256 KiB pool for atomic coherent allocations [ 0.202737] hw-breakpoint: found 5 (+1 reserved) breakpoint and 4 watchpoint registers. [ 0.202744] hw-breakpoint: maximum watchpoint size is 8 bytes. [ 0.218300] vgaarb: loaded [ 0.218565] SCSI subsystem initialized [ 0.218936] usbcore: registered new interface driver usbfs [ 0.219019] usbcore: registered new interface driver hub [ 0.219112] usbcore: registered new device driver usb [ 0.220983] clocksource: Switched to clocksource arch_sys_counter [ 0.249890] NET: Registered protocol family 2 [ 0.250312] TCP established hash table entries: 8192 (order: 3, 32768 bytes) [ 0.250369] TCP bind hash table entries: 8192 (order: 4, 65536 bytes) [ 0.250474] TCP: Hash tables configured (established 8192 bind 8192) [ 0.250514] UDP hash table entries: 512 (order: 2, 16384 bytes) [ 0.250539] UDP-Lite hash table entries: 512 (order: 2, 16384 bytes) [ 0.250661] NET: Registered protocol family 1 [ 0.250872] RPC: Registered named UNIX socket transport module. [ 0.250879] RPC: Registered udp transport module. [ 0.250885] RPC: Registered tcp transport module. [ 0.250890] RPC: Registered tcp NFSv4.1 backchannel transport module. [ 0.251053] Unpacking initramfs... [ 0.252978] Initramfs unpacking failed: read error [ 0.253002] Freeing initrd memory: 32K (c8080000 - c8088000) [ 0.253146] hw perfevents: enabled with armv7_cortex_a15 PMU driver, 7 counters available [ 0.253718] platform alarmtimer: set dma_pfn_offset00780000 [ 0.254334] futex hash table entries: 16 (order: -2, 1024 bytes) [ 0.255083] workingset: timestamp_bits=30 max_order=20 bucket_order=0 [ 0.266086] Installing knfsd (copyright (C) 1996 okir@monad.swb.de). [ 0.266169] ntfs: driver 2.1.32 [Flags: R/O]. [ 0.266501] jffs2: version 2.2. (NAND) �© 2001-2006 Red Hat, Inc. [ 0.269063] NET: Registered protocol family 38 [ 0.269102] bounce: pool size: 64 pages [ 0.269315] Block layer SCSI generic (bsg) driver version 0.4 loaded (major 252) [ 0.269324] io scheduler noop registered [ 0.269330] io scheduler deadline registered [ 0.269497] io scheduler cfq registered (default) [ 0.269642] keystone_irq soc:keystone_irq@26202a0: irqchip registered, nr_irqs 28 [ 0.270665] keystone-navigator-qmss soc:qmss@2a40000: qmgr start queue 0, number of queues 8192 [ 0.270771] keystone-navigator-qmss soc:qmss@2a40000: added qmgr start queue 0, num of queues 8192, reg_peek f0940000, reg_status f0938000, reg_config f093a000, reg_region f093c000, reg_push f0980000, reg_po0 [ 0.270891] keystone-navigator-qmss soc:qmss@2a40000: loading /lib/firmware/ks2_qmss_pdsp_acc48.bin failed with error -26 [ 0.270900] keystone-navigator-qmss soc:qmss@2a40000: failed to get firmware for pdsp [ 0.271650] keystone-navigator-qmss soc:qmss@2a40000: pdsp id 0 not started for range acc-low-0 [ 0.272395] keystone-navigator-dma soc:knav_dmas@0: DMA dma_gbe registered 149 logical channels, flows 128, tx chans: 21, rx chans: 91 [ 0.334684] Serial: 8250/16550 driver, 4 ports, IRQ sharing disabled [ 0.334763] platform serial8250: set dma_pfn_offset00780000 [ 0.336304] console [ttyS0] disabled [ 0.336362] 2530c00.serial: ttyS0 at MMIO 0x2530c00 (irq = 26, base_baud = 13020833) is a 16550A [ 1.054186] console [ttyS0] enabled [ 1.058371] 2531000.serial: ttyS1 at MMIO 0x2531000 (irq = 27, base_baud = 13020833) is a 16550A [ 1.072642] loop: module loaded [ 1.076059] at24 0-0050: 131072 byte 24c1024 EEPROM, writable, 1 bytes/write [ 1.084006] spi_davinci 21000400.spi: DMA is not supported (-19) [ 1.090629] m25p80 spi0.0: n25q128a11 (16384 Kbytes) [ 1.095618] 2 ofpart partitions found on MTD device spi0.0 [ 1.101129] Creating 2 MTD partitions on "spi0.0": [ 1.105913] 0x000000000000-0x000000080000 : "u-boot-spl" [ 1.112168] 0x000000080000-0x000001000000 : "misc" [ 1.117833] spi_davinci 21000400.spi: Controller at 0xf097a400 [ 1.123826] spi_davinci 21000600.spi: DMA is not supported (-19) [ 1.130167] spi_davinci 21000600.spi: Controller at 0xf097c600 [ 1.136138] spi_davinci 21000800.spi: DMA is not supported (-19) [ 1.142504] spi_davinci 21000800.spi: Controller at 0xf097e800 [ 1.149201] platform Fixed MDIO bus.0: set dma_pfn_offset00780000 [ 1.155461] libphy: Fixed MDIO Bus: probed [ 1.211007] davinci_mdio 24200f00.mdio: davinci mdio revision 1.7 [ 1.217093] libphy: 24200f00.mdio: probed [ 1.221761] davinci_mdio 24200f00.mdio: phy[0]: device 24200f00.mdio:00, driver Marvell 88E1510 [ 1.230451] davinci_mdio 24200f00.mdio: phy[1]: device 24200f00.mdio:01, driver Marvell 88E1510 [ 4.252103] netcp-1.0 2620110.netcp: module(netcp-xgbe) not used for device [ 4.259980] platform xhci-hcd.0.auto: set dma_pfn_offset00780000 [ 4.266566] platform xhci-hcd.1.auto: set dma_pfn_offset00780000 [ 4.273224] xhci-hcd xhci-hcd.0.auto: xHCI Host Controller [ 4.278865] xhci-hcd xhci-hcd.0.auto: new USB bus registered, assigned bus number 1 [ 4.286767] xhci-hcd xhci-hcd.0.auto: hcc params 0x0298f06d hci version 0x100 quirks 0x00010010 [ 4.295481] xhci-hcd xhci-hcd.0.auto: irq 34, io mem 0x02690000 [ 4.301549] usb usb1: New USB device found, idVendor=1d6b, idProduct=0002 [ 4.308329] usb usb1: New USB device strings: Mfr=3, Product=2, SerialNumber=1 [ 4.315557] usb usb1: Product: xHCI Host Controller [ 4.320428] usb usb1: Manufacturer: Linux 4.8.0-rc2-00003-ga0a0676 xhci-hcd [ 4.327409] usb usb1: SerialNumber: xhci-hcd.0.auto [ 4.332755] hub 1-0:1.0: USB hub found [ 4.336521] hub 1-0:1.0: 1 port detected [ 4.340702] xhci-hcd xhci-hcd.0.auto: xHCI Host Controller [ 4.346376] xhci-hcd xhci-hcd.0.auto: new USB bus registered, assigned bus number 2 [ 4.354094] usb usb2: We don't know the algorithms for LPM for this host, disabling LPM. [ 4.362297] usb usb2: New USB device found, idVendor=1d6b, idProduct=0003 [ 4.369077] usb usb2: New USB device strings: Mfr=3, Product=2, SerialNumber=1 [ 4.376303] usb usb2: Product: xHCI Host Controller [ 4.381199] usb usb2: Manufacturer: Linux 4.8.0-rc2-00003-ga0a0676 xhci-hcd [ 4.388151] usb usb2: SerialNumber: xhci-hcd.0.auto [ 4.393481] hub 2-0:1.0: USB hub found [ 4.397244] hub 2-0:1.0: 1 port detected [ 4.401485] xhci-hcd xhci-hcd.1.auto: xHCI Host Controller [ 4.407122] xhci-hcd xhci-hcd.1.auto: new USB bus registered, assigned bus number 3 [ 4.415007] xhci-hcd xhci-hcd.1.auto: hcc params 0x0298f06d hci version 0x100 quirks 0x00010010 [ 4.423738] xhci-hcd xhci-hcd.1.auto: irq 68, io mem 0x25010000 [ 4.429766] usb usb3: New USB device found, idVendor=1d6b, idProduct=0002 [ 4.436561] usb usb3: New USB device strings: Mfr=3, Product=2, SerialNumber=1 [ 4.443871] usb usb3: Product: xHCI Host Controller [ 4.448741] usb usb3: Manufacturer: Linux 4.8.0-rc2-00003-ga0a0676 xhci-hcd [ 4.455706] usb usb3: SerialNumber: xhci-hcd.1.auto [ 4.461032] hub 3-0:1.0: USB hub found [ 4.464796] hub 3-0:1.0: 1 port detected [ 4.468965] xhci-hcd xhci-hcd.1.auto: xHCI Host Controller [ 4.474617] xhci-hcd xhci-hcd.1.auto: new USB bus registered, assigned bus number 4 [ 4.482347] usb usb4: We don't know the algorithms for LPM for this host, disabling LPM. [ 4.490522] usb usb4: New USB device found, idVendor=1d6b, idProduct=0003 [ 4.497317] usb usb4: New USB device strings: Mfr=3, Product=2, SerialNumber=1 [ 4.504555] usb usb4: Product: xHCI Host Controller [ 4.509424] usb usb4: Manufacturer: Linux 4.8.0-rc2-00003-ga0a0676 xhci-hcd [ 4.516480] usb usb4: SerialNumber: xhci-hcd.1.auto [ 4.521814] hub 4-0:1.0: USB hub found [ 4.525577] hub 4-0:1.0: 1 port detected [ 4.529927] usbcore: registered new interface driver usb-storage [ 4.536231] mousedev: PS/2 mouse device common for all mice [ 4.542036] i2c /dev entries driver [ 4.546204] davinci-wdt 22f0080.wdt: heartbeat 60 sec [ 4.551956] usbcore: registered new interface driver usbhid [ 4.557520] usbhid: USB HID core driver [ 4.562252] nand: device found, Manufacturer ID: 0x2c, Chip ID: 0xac [ 4.568597] nand: Micron MT29F4G08ABBDAHC [ 4.572702] nand: 512 MiB, SLC, erase size: 128 KiB, page size: 2048, OOB size: 64 [ 4.580549] Bad block table found at page 262080, version 0x01 [ 4.586896] Bad block table found at page 262016, version 0x01 [ 4.593059] 3 ofpart partitions found on MTD device 30000000.nand [ 4.599143] Creating 3 MTD partitions on "30000000.nand": [ 4.604550] 0x000000000000-0x000000100000 : "u-boot" [ 4.610415] 0x000000100000-0x000000180000 : "params" [ 4.616256] 0x000000180000-0x000020000000 : "ubifs" [ 4.625578] davinci_nand 30000000.nand: controller rev. 2.5 [ 4.631478] platform oprofile-perf.0: set dma_pfn_offset00780000 [ 4.637710] oprofile: using timer interrupt. [ 4.642025] Netfilter messages via NETLINK v0.30. [ 4.646933] nf_conntrack version 0.5.0 (16384 buckets, 65536 max) [ 4.653307] ctnetlink v0.93: registering with nfnetlink. [ 4.658856] ipip: IPv4 and MPLS over IPv4 tunneling driver [ 4.664820] gre: GRE over IPv4 demultiplexor driver [ 4.669690] ip_gre: GRE over IPv4 tunneling driver [ 4.675526] ip_tables: (C) 2000-2006 Netfilter Core Team [ 4.680898] ipt_CLUSTERIP: ClusterIP Version 0.8 loaded successfully [ 4.687272] arp_tables: arp_tables: (C) 2002 David S. Miller [ 4.692979] Initializing XFRM netlink socket [ 4.698180] NET: Registered protocol family 10 [ 4.703401] NET: Registered protocol family 17 [ 4.707848] NET: Registered protocol family 15 [ 4.712316] 8021q: 802.1Q VLAN Support v1.8 [ 4.716535] sctp: Hash tables configured (bind 512/512) [ 4.722069] Registering SWP/SWPB emulation handler [ 4.832242] netcp-1.0 2620110.netcp eth0: Link is Down [ 4.839687] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready [ 4.845527] 8021q: adding VLAN 0 to HW filter on device eth0 [ 4.972238] netcp-1.0 2620110.netcp eth1: Link is Down [ 4.978292] IPv6: ADDRCONF(NETDEV_UP): eth1: link is not ready [ 4.984131] 8021q: adding VLAN 0 to HW filter on device eth1 [ 4.989824] IP-Config: Failed to open gretap0 [ 5.020995] Sending DHCP requests . [ 7.991361] netcp-1.0 2620110.netcp eth0: Link is Up - 100Mbps/Full - flow control off [ 7.999451] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready [ 8.520991] ., OK [ 8.841123] IP-Config: Got DHCP answer from 192.168.1.1, my address is 192.168.1.38 [ 9.081896] IP-Config: Complete: [ 9.085120] device=eth0, hwaddr=76:34:13:00:7c:66, ipaddr=192.168.1.38, mask=255.255.255.0, gw=192.168.1.1 [ 9.095273] host=192.168.1.38, domain=Home, nis-domain=(none) [ 9.101459] bootserver=0.0.0.0, rootserver=192.168.1.36, rootpath= nameserver0=192.168.1.1 [ 9.160413] VFS: Mounted root (nfs filesystem) on device 0:14. [ 9.166843] devtmpfs: mounted [ 9.170868] Freeing unused kernel memory: 2048K (c0c00000 - c0e00000) [ 9.202908] random: fast init done INIT: version 2.88 booting [info] Using makefile-style concurrent boot in runlevel S. [....] Starting the hotplug events dispatcher: udevd[ 10.314986] udevd[1309]: starting version 175 . ok [....] Synthesizing the initial hotplug events...done. [....] Waiting for /dev to be fully populated...done. [....] Setting parameters of disc: (none). ok [....] Activating swap...done. [....] Cleaning up temporary files... /tmp. ok [ 12.160599] random: crng init done [....] Activating lvm and md swap...done. [....] Checking file systems...fsck from util-linux 2.20.1 done. [....] Mounting local filesystems...done. [....] Activating swapfile swap...done. [....] Cleaning up temporary files.... ok [....] Setting kernel variables ...done. [....] Configuring network interfaces...done. [....] Cleaning up temporary files.... ok [....] Setting up ALSA...done (none loaded). [....] Setting up X socket directories... /tmp/.X11-unix /tmp/.ICE-unix. ok INIT: Entering runlevel: 2 [info] Using makefile-style concurrent boot in runlevel 2. [....] Starting advanced IEEE 802.11 management: hostapdioctl(SIOCGIFINDEX): No such device failed! [....] Starting enhanced syslogd: rsyslogd. ok [....] Starting periodic command scheduler: cron. ok [....] Starting system message bus: dbus. ok [....] Starting OpenBSD Secure Shell server: sshd. ok [....] Starting Avahi mDNS/DNS-SD Daemon: avahi-daemon. ok Last login: Thu Jan 1 00:00:17 UTC 1970 on ttyS0 Linux rockdesk 4.8.0-rc2-00003-ga0a0676 #974 SMP PREEMPT Fri Aug 19 10:26:47 EEST 2016 armv7l The programs included with the Debian GNU/Linux system are free software; the exact distribution terms for each program are described in the individual files in /usr/share/doc/*/copyright. Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent permitted by applicable law. root@rockdesk:~# [ 21.941001] usb 1-1: new high-speed USB device number 2 using xhci-hcd [ 22.111469] usb 1-1: New USB device found, idVendor=0781, idProduct=5539 [ 22.118166] usb 1-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3 [ 22.125359] usb 1-1: Product: Cruzer Micro [ 22.129449] usb 1-1: Manufacturer: SanDisk [ 22.133606] usb 1-1: SerialNumber: 173733160C524A8E [ 22.139581] usb-storage 1-1:1.0: USB Mass Storage device detected [ 22.145949] scsi host0: usb-storage 1-1:1.0 [ 22.150210] scsi host0: runtime PM trying to activate child device host0 but parent (1-1:1.0) is not active [ 23.191374] scsi 0:0:0:0: Direct-Access SanDisk Cruzer Micro 8.02 PQ: 0 ANSI: 0 CCS [ 23.200473] sd 0:0:0:0: [sda] 15695871 512-byte logical blocks: (8.04 GB/7.48 GiB) [ 23.208341] sd 0:0:0:0: [sda] Write Protect is off [ 23.213490] sd 0:0:0:0: [sda] No Caching mode page found [ 23.218993] sd 0:0:0:0: [sda] Assuming drive cache: write through [ 23.229000] sda: [ 23.232085] sd 0:0:0:0: [sda] Attached SCSI removable disk root@rockdesk:~# ./usbmsctest.sh [ 29.678847] FAT-fs (sda): Volume was not properly unmounted. Some data may be corrupt. Please run fsck. Mounted /org/freedesktop/UDisks/devices/sda at /media/PENDRIVE copying file syncing removing file speed test 10+0 records in 10+0 records out 10485760 bytes (10 MB) copied, 2.7477 s, 3.8 MB/s done root@rockdesk:~#
On 8/19/2016 12:30 AM, Roger Quadros wrote: > Hi Santosh, > >>> So I'm 99.9% convinced that the proposed change is correct. >>> >> I will got with that then :-) and take my objection back. Just >> saying that if there other breakages which I can't recollect now, >> those drivers needs to be patched as well. >> > I was able to boot the Keystone2 Edison EVM over NFS with the $subject patch. > Boot log is below. Do you see anything suspicious? > Logs looks ok to me. Probably do some tests where DMA and bounce buffers etc gets tested. Running it through your internal regression suit will be good idea as well if thats possible. Regards, Santosh
Hi Santosh & Russell, On 19/08/16 19:38, Santosh Shilimkar wrote: > > On 8/19/2016 12:30 AM, Roger Quadros wrote: >> Hi Santosh, >> > >>>> So I'm 99.9% convinced that the proposed change is correct. >>>> >>> I will got with that then :-) and take my objection back. Just >>> saying that if there other breakages which I can't recollect now, >>> those drivers needs to be patched as well. >>> >> I was able to boot the Keystone2 Edison EVM over NFS with the $subject patch. >> Boot log is below. Do you see anything suspicious? >> > Logs looks ok to me. Probably do some tests where DMA and bounce buffers etc gets tested. Running it through your internal regression > suit will be good idea as well if thats possible. > This has been running in our internal test suite for a week on various TI platforms. There haven't been any surprises. Is it a good idea to at least put this in -next for a wider test audience? cheers, -roger
Hi, On 12/09/16 14:38, Roger Quadros wrote: > Hi Santosh & Russell, > > On 19/08/16 19:38, Santosh Shilimkar wrote: >> >> On 8/19/2016 12:30 AM, Roger Quadros wrote: >>> Hi Santosh, >>> >> >>>>> So I'm 99.9% convinced that the proposed change is correct. >>>>> >>>> I will got with that then :-) and take my objection back. Just >>>> saying that if there other breakages which I can't recollect now, >>>> those drivers needs to be patched as well. >>>> >>> I was able to boot the Keystone2 Edison EVM over NFS with the $subject patch. >>> Boot log is below. Do you see anything suspicious? >>> >> Logs looks ok to me. Probably do some tests where DMA and bounce buffers etc gets tested. Running it through your internal regression >> suit will be good idea as well if thats possible. >> > > This has been running in our internal test suite for a week on various TI > platforms. There haven't been any surprises. > > Is it a good idea to at least put this in -next for a wider test audience? Gentle reminder. regards, -roger
On Wed, Sep 28, 2016 at 10:53:49AM +0300, Roger Quadros wrote: > Hi, > > On 12/09/16 14:38, Roger Quadros wrote: > > Hi Santosh & Russell, > > > > On 19/08/16 19:38, Santosh Shilimkar wrote: > >> > >> On 8/19/2016 12:30 AM, Roger Quadros wrote: > >>> Hi Santosh, > >>> > >> > >>>>> So I'm 99.9% convinced that the proposed change is correct. > >>>>> > >>>> I will got with that then :-) and take my objection back. Just > >>>> saying that if there other breakages which I can't recollect now, > >>>> those drivers needs to be patched as well. > >>>> > >>> I was able to boot the Keystone2 Edison EVM over NFS with the $subject patch. > >>> Boot log is below. Do you see anything suspicious? > >>> > >> Logs looks ok to me. Probably do some tests where DMA and bounce buffers etc gets tested. Running it through your internal regression > >> suit will be good idea as well if thats possible. > >> > > > > This has been running in our internal test suite for a week on various TI > > platforms. There haven't been any surprises. > > > > Is it a good idea to at least put this in -next for a wider test audience? > > Gentle reminder. Please put it in the patch system, thanks.
diff --git a/arch/arm/include/asm/dma-mapping.h b/arch/arm/include/asm/dma-mapping.h index d009f79..bf02dbd 100644 --- a/arch/arm/include/asm/dma-mapping.h +++ b/arch/arm/include/asm/dma-mapping.h @@ -111,7 +111,7 @@ static inline dma_addr_t virt_to_dma(struct device *dev, void *addr) /* The ARM override for dma_max_pfn() */ static inline unsigned long dma_max_pfn(struct device *dev) { - return PHYS_PFN_OFFSET + dma_to_pfn(dev, *dev->dma_mask); + return dma_to_pfn(dev, *dev->dma_mask); } #define dma_max_pfn(dev) dma_max_pfn(dev)
Since commit 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation"), dma_to_pfn() already returns the PFN with the physical memory start offset so we don't need to add it again. This fixes USB mass storage lock-up problem on systems that can't do DMA over the entire physical memory range (e.g.) Keystone 2 systems with 4GB RAM can only do DMA over the first 2GB. [K2E-EVM]. What happens there is that without this patch SCSI layer sets a wrong bounce buffer limit in scsi_calculate_bounce_limit() for the USB mass storage device. dma_max_pfn() evaluates to 0x8fffff and bounce_limit is set to 0x8fffff000 whereas maximum DMA'ble physical memory on Keystone 2 is 0x87fffffff. This results in non DMA'ble pages being given to the USB controller and hence the lock-up. NOTE: in the above case, USB-SCSI-device's dma_pfn_offset was showing as 0. This should have really been 0x780000 as on K2e, LOWMEM_START is 0x80000000 and HIGHMEM_START is 0x800000000. DMA zone is 2GB so dma_max_pfn should be 0x87ffff. The incorrect dma_pfn_offset for the USB storage device is because USB devices are not correctly inheriting the dma_pfn_offset from the USB host controller. This will be fixed by a separate patch. Fixes: 6ce0d2001692 ("ARM: dma: Use dma_pfn_offset for dma address translation") Cc: stable@vger.kernel.org Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Russell King <linux@arm.linux.org.uk> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Olof Johansson <olof@lixom.net> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Linus Walleij <linus.walleij@linaro.org> Reported-by: Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by: Roger Quadros <rogerq@ti.com> --- arch/arm/include/asm/dma-mapping.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)