Commit Graph

15 Commits

Author SHA1 Message Date
Peter Chen e48aa1eb44
usb: chipidea: udc: add software sg list support
The chipidea controller doesn't support short transfer for sg list,
so we still keep setting IOC per TD, otherwise, there will be no interrupt
for short transfer. Each TD has five entries for data buffer, each data
buffer could be non-countinuous 4KB buffer, so it could handle
up to 5 sg buffers one time. The benefit of this patch is avoiding
OOM for low memory system(eg, 256MB) during large USB transfers, see
below for detail. The non-sg handling has not changed.

ufb: page allocation failure: order:4, mode:0x40cc0(GFP_KERNEL|__GFP_COMP),
nodemask=(null),cpuset=/,mems_allowed=0
CPU: 2 PID: 370 Comm: ufb Not tainted 5.4.3-1.1.0+g54b3750d61fd #1
Hardware name: NXP i.MX8MNano DDR4 EVK board (DT)
Call trace:
 dump_backtrace+0x0/0x140
 show_stack+0x14/0x20
 dump_stack+0xb4/0xf8
 warn_alloc+0xec/0x158
 __alloc_pages_slowpath+0x9cc/0x9f8
 __alloc_pages_nodemask+0x21c/0x280
 alloc_pages_current+0x7c/0xe8
 kmalloc_order+0x1c/0x88
 __kmalloc+0x25c/0x298
 ffs_epfile_io.isra.0+0x20c/0x7d0
 ffs_epfile_read_iter+0xa8/0x188
 new_sync_read+0xe4/0x170
 __vfs_read+0x2c/0x40
 vfs_read+0xc8/0x1a0
 ksys_read+0x68/0xf0
 __arm64_sys_read+0x18/0x20
 el0_svc_common.constprop.0+0x68/0x160
 el0_svc_handler+0x20/0x80
 el0_svc+0x8/0xc
Mem-Info:
active_anon:2856 inactive_anon:5269 isolated_anon:12
 active_file:5238 inactive_file:18803 isolated_file:0
 unevictable:0 dirty:22 writeback:416 unstable:0
 slab_reclaimable:4073 slab_unreclaimable:3408
 mapped:727 shmem:7393 pagetables:37 bounce:0
 free:4104 free_pcp:118 free_cma:0
Node 0 active_anon:11436kB inactive_anon:21076kB active_file:20988kB inactive_file:75216kB unevictable:0kB isolated(ano
Node 0 DMA32 free:16820kB min:1808kB low:2260kB high:2712kB active_anon:11436kB inactive_anon:21076kB active_file:2098B
lowmem_reserve[]: 0 0 0
Node 0 DMA32: 508*4kB (UME) 242*8kB (UME) 730*16kB (UM) 21*32kB (UME) 5*64kB (UME) 2*128kB (M) 0*256kB 0*512kB 0*1024kB
Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=1048576kB
Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=32768kB
Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB
Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=64kB
31455 total pagecache pages
0 pages in swap cache
Swap cache stats: add 0, delete 0, find 0/0
Free swap  = 0kB
Total swap = 0kB
65536 pages RAM
0 pages HighMem/MovableOnly
10766 pages reserved
0 pages cma reserved
0 pages hwpoisoned

Reviewed-by: Jun Li <jun.li@nxp.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
2020-05-06 09:42:35 +08:00
Peter Chen 5de131d213
usb: chipidea: udc: fix the kernel doc for udc.h
The kernel doc for td_node is outdated, update it.

Signed-off-by: Peter Chen <peter.chen@nxp.com>
2020-03-30 14:42:34 +08:00
Nishad Kamdar d9958306d4
USB: chipidea: Use the correct style for SPDX License Identifier
This patch corrects the SPDX License Identifier style in
header files related to ChipIdea Highspeed Dual Role Controller.
For C header files Documentation/process/license-rules.rst
mandates C-like comments (opposed to C source files where
C++ style should be used).

Changes made by using a script provided by Joe Perches here:
https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flkml.org%2Flkml%2F2019%2F2%2F7%2F46&amp;data=02%7C01%7CPeter.Chen%40nxp.com%7Cbea69ff84b574ca6b48e08d7c8cf58cf%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0%7C637198665199494622&amp;sdata=bk1n4%2BvnrfRS6ZDrps%2BuXiImdzaxKZ00YskBg6pjtn4%3D&amp;reserved=0.

Suggested-by: Joe Perches <joe@perches.com>
Signed-off-by: Nishad Kamdar <nishadkamdar@gmail.com>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
2020-03-16 11:13:47 +08:00
Greg Kroah-Hartman 51b751f112 USB: chipidea: Remove redundant license text
Now that the SPDX tag is in all USB files, that identifies the license
in a specific and legally-defined manner.  So the extra GPL text wording
can be removed as it is no longer needed at all.

This is done on a quest to remove the 700+ different ways that files in
the kernel describe the GPL license text.  And there's unneeded stuff
like the address (sometimes incorrect) for the FSF which is never
needed.

No copyright headers or other non-license-description text was removed.

Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: Jonathan Hunter <jonathanh@nvidia.com>
Acked-by: Peter Chen <peter.chen@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-07 15:45:01 +01:00
Greg Kroah-Hartman 5fd54ace47 USB: add SPDX identifiers to all remaining files in drivers/usb/
It's good to have SPDX identifiers in all files to make it easier to
audit the kernel tree for correct licenses.

Update the drivers/usb/ and include/linux/usb* files with the correct
SPDX license identifier based on the license text in the file itself.
The SPDX identifier is a legally binding shorthand, which can be used
instead of the full boiler plate text.

This work is based on a script and data from Thomas Gleixner, Philippe
Ombredanne, and Kate Stewart.

Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Kate Stewart <kstewart@linuxfoundation.org>
Cc: Philippe Ombredanne <pombredanne@nexb.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Acked-by: Felipe Balbi <felipe.balbi@linux.intel.com>
Acked-by: Johan Hovold <johan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-04 11:48:02 +01:00
Stephen Boyd 34445fb433 usb: chipidea: Properly mark little endian descriptors
The DMA descriptors are little endian, and we do a pretty good
job of handling them with the proper le32_to_cpu() markings, but
we don't actually mark them as __le32. This means checkers like
sparse can't easily find new bugs. Let's mark the members of
structures properly and fix the few places where we're missing
conversions.

Cc: Peter Chen <peter.chen@nxp.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Stephen Boyd <stephen.boyd@linaro.org>
Signed-off-by: Peter Chen <peter.chen@nxp.com>
2016-11-14 10:03:40 +08:00
Peter Chen 3f124d233e usb: chipidea: add role init and destroy APIs
- The role's init will be called at probe procedure.
- The role's destroy will be called at fail patch
at probe and driver's removal.
- The role's start/stop will be called when specific
role has started.

Tested-by: Marek Vasut <marex@denx.de>
Signed-off-by: Peter Chen <peter.chen@freescale.com>
Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2013-08-14 12:37:19 -07:00
Alexander Shishkin 8e22978c57 usb: chipidea: drop "13xxx" infix
"ci13xxx" is bad for at least the following reasons:
  * people often mistype it
  * it doesn't add any informational value to the names it's used in
  * it needlessly attracts mail filters

This patch replaces it with "ci_hdrc", "ci_udc" or "ci_hw", depending
on the situation. Modules with ci13xxx prefix are also renamed accordingly
and aliases are added for compatibility. Otherwise, no functional changes.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2013-06-24 16:16:55 -07:00
Michael Grzeschik cc9e6c495b usb: chipidea: udc: manage dynamic amount of tds with a linked list
Instead of having a limited number of usable tds in the udc we use a
linked list to support dynamic amount of needed tds for all special
gadget types. This improves throughput.

Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2013-06-17 13:45:47 -07:00
Michael Grzeschik e4ce4ecd96 usb: chipidea: udc: configure iso endpoints
This patch adds iso endpoint support to the device controller.
It makes use of the multiplication bits in the maxpacket field
of the endpoint and calculates the multiplier bits for each
transfer description on every request.

Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
Reviewed-by: Peter Chen <peter.chen@freescale.com>
Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2013-06-17 13:45:46 -07:00
Michael Grzeschik a9c174302b usb: chipidea: udc: fix memory access of shared memory on armv5 machines
The udc uses an shared dma memory space between hard and software. This
memory layout is described in ci13xxx_qh and ci13xxx_td which are marked
with the attribute ((packed)).

The compiler currently does not know about the alignment of the memory
layout, and will create strb and ldrb operations.

The Datasheet of the synopsys core describes, that some operations on
the mapped memory need to be atomic double word operations. I.e. the
next pointer addressing in the qhead, as otherwise the hardware will
read wrong data and totally stuck.

This is also possible while working with the current active td queue,
and preparing the td->ptr.next in software while the hardware is still
working with the current active td which is supposed to be changed:

writeb(0xde, &td->ptr.next + 0x0); /* strb */
writeb(0xad, &td->ptr.next + 0x1); /* strb */

<----- hardware reads value of td->ptr.next and get stuck!

writeb(0xbe, &td->ptr.next + 0x2); /* strb */
writeb(0xef, &td->ptr.next + 0x3); /* strb */

This appeares on armv5 machines where the hardware does not support
unaligned 32bit operations.

This patch adds the attribute ((aligned(4))) to the structures to tell
the compiler to use 32bit operations. It also adds an wmb() for the
prepared TD data before it gets enqueued into the qhead.

Cc: stable <stable@vger.kernel.org> # v3.5
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2013-04-05 13:43:14 -07:00
Alexander Shishkin 5e0aa49ec6 usb: chipidea: use generic map/unmap routines
We're one of the remaining drivers to map/unmap requests by hand. Switch
to using generic gadget routines for that instead.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-05-11 16:59:35 -07:00
Alexander Shishkin 551a8ac64e usb: chipidea: brush up structure definitions
Get rid of trailing comments in the structure definitions in favor of
kernel-doc.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-05-11 16:54:11 -07:00
Alexander Shishkin 5f36e231e9 usb: chipidea: add support for roles
Add some generic code for roles and implement simple role switching
based on ID pin state and/or a sysfs file. At this, we also rename
the device to ci_hdrc, which is what it is.

The "manual" switch is made into a sysfs file and not debugfs, because
it might be useful even in non-debug context. For some boards, like
sheevaplug, it seems to be the only way to switch roles without modifying
the hardware, since the ID pin is always grounded.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-05-11 16:52:10 -07:00
Alexander Shishkin e443b33362 usb: chipidea: split the driver code into units
Split the driver into the following parts:
  * core  -- resources, register access, capabilities, etc;
  * udc   -- device controller functionality;
  * debug -- logging events.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-05-11 16:49:51 -07:00