| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Signed-off-by: Nobuhiro Iwamatsu <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Signed-off-by: Nobuhiro Iwamatsu <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The 88f6282 has one more TWSI(TWSI1). This add the information to enable
pinctl of TWSI1.
Signed-off-by: Nobuhiro Iwamatsu <[email protected]>
Acked-by: Andrew Lunn <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
This commit adds a pinmux option, pmx_sdio, to enable the muxing of
the SDIO interface on the 88F6282 SoC from Marvell.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Now that the mvsdio driver has a Device Tree binding, and the SDIO
controller is declared in kirkwood.dtsi, migrate the mplcec4 board to
use the Device Tree to probe the SDIO controller and to mux the pins
of the SDIO interface correctly.
This patch has not been tested, it remains to be tested by a person
having access to the hardware.
Signed-off-by: Thomas Petazzoni <[email protected]>
Cc: Stefan Peter <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Now that the mvsdio driver has a Device Tree binding, and the SDIO
controller is declared in kirkwood.dtsi, migrate the dreamplug board
to use the Device Tree to probe the SDIO controller and to mux this
interface properly.
Signed-off-by: Thomas Petazzoni <[email protected]>
Cc: Jason Cooper <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Now that the SDIO controller has a Device Tree binding, let's use it
in kirkwood.dtsi.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The Globalscale Mirabox uses the SDIO interface of the Armada 370 to
connect to a Wifi/Bluetooth SD8787 chip, so we enable the SDIO
interface of this board in its Device Tree file.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The Armada XP DB evaluation board has one SD card slot, directly
connected to the SDIO IP of the SoC, so we add a device tree
description for it.
However, in the default configuration of the board, the SD card slot
is not usable: the connector plugged into CON40 must be changed
against a different one, provided with the board by the
manufacturer. Since such a manual modification of the hardware is
needed, we did not enable the SDIO interface by default, and left it
to the board user to modify the Device Tree if needed. Since this
board is really only an evaluation board for developers and not a
final product, it is not too bad.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The Armada XP DB evaluation board has one SD card slot, directly
connected to the SDIO IP of the SoC, so we enable this
IP. Unfortunately, there are no GPIOs for card-detect and
write-protect.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The SDIO interface is only available on pins MPP30/31/32/33/34/35 on
the various Armada XP variants, so we provide a pin muxing option for
this in the Armada XP .dtsi files.
Even though those muxing options are the same for MV78230, MV78260 and
MV78460, we keep them in each .dtsi file, because the number of pins,
and therefore the declaration of the pinctrl node, is different for
each SoC variant.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The SDIO interface is available either on pins MPP9/11/12/13/14/15 or
MPP47/48/49/50/51/52 on the Armada 370. Even though all combinations
are potentially possible, those two muxing options are the most
probable ones, so we provide those at the SoC level .dtsi file.
In practice, in turns out the Armada 370 DB board uses the former,
while the Armada 370 Mirabox uses the latter.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Now that the mvsdio MMC driver has a Device Tree binding, we add the
Device Tree informations to describe the SDIO interface available in
the Armada 370/XP SoCs.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The Armada 370 and Armada XP Socs have the same controller that the
one used in the orion platforms. This patch updates the device tree
for these SoCs.
Signed-off-by: Gregory CLEMENT <[email protected]>
Acked-by: Andrew Lunn <[email protected]>
Tested-by: Florian Fainelli <[email protected]>
Acked-by: Arnd Bergmann <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Some of the mvebu boards (mainly the development board) come with
plug-in RAM modules. This patch allows to let the bootloaders which
have no support for DTS to give the real amount of memory available on
the board.
Signed-off-by: Gregory CLEMENT <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The Armada XP DB-MV784MP-GP board has an SPI flash device.
These options allow to access that device over MTD.
Signed-off-by: Ezequiel Garcia <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Cc: Thomas Petazzoni <[email protected]>
Cc: Lior Amsalem <[email protected]>
Acked-by: Gregory Clement <[email protected]>
Signed-off-by: Ezequiel Garcia <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
This is the new Armada XP evaluation board from Marvell. It comes with
a RS232 port over USB, a SATA link, an internal SSD, 4 Ethernet
Gigabit links.
Support for USB (Host and device), SDIO, PCIe will be added as drivers
when they become available for Armada XP in mainline.
Tested-by: Simon Guinot <[email protected]>
Tested-by: Florian Fainelli <[email protected]>
Signed-off-by: Gregory CLEMENT <[email protected]>
Tested-by: Ezequiel Garcia <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Now that we have support for local timers, enable it by default
Signed-off-by: Gregory CLEMENT <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
On the SOCs Armada 370 and Armada XP, each CPU comes with two private
timers. This patch use the timer 0 of each CPU as local timer for the
clockevent if CONFIG_LOCAL_TIMER is selected. In the other case, use
only the private Timer 0 of CPU 0.
Signed-off-by: Gregory CLEMENT <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
MPIC allows the use of private interrupt for each CPUs. The 28th first
interrupts are per-cpu. This patch adds support to use them.
Signed-off-by: Gregory CLEMENT <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Cc: Lior Amsalem <[email protected]>
Cc: Thomas Petazzoni <[email protected]>
Cc: Gregory CLEMENT <[email protected]>
Tested-by: Nobuhiro Iwamatsu <[email protected]>
Tested-by: Florian Fainelli <[email protected]>
Signed-off-by: Ezequiel Garcia <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Add a device tree entry for the Guruplug Server Plus board. This port
was based both on the work done on the dreamplug and the dockstar.
It builds, boots and works on my Guruplug Server Plus.
Signed-off-by: Willy Tarreau <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
This patch adds the DTS file to support the Marvell RD-A370-A1
(Reference Design board) also known as RD-88F6710 board. It is almost
entirely similar to the DB-A370 board except that the first Ethernet PHY
is SGMII-wired and the second is a switch which is RGMII-wired.
Signed-off-by: Florian Fainelli <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
This patch makes the interrupt controller driver more SMP aware for
the Armada XP SoCs. It adds the support for the per-CPU irq. It also
adds the implementation for the set_affinity hook.
Patch initialy wrote by Yehuda Yitschak and reworked by Gregory
CLEMENT.
Signed-off-by: Yehuda Yitschak <[email protected]>
Signed-off-by: Gregory CLEMENT <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
In the beginning of DT for Dove it was reasonable to have it close to
non-DT code. With improved DT support, it became more and more difficult
to not break non-DT while changing DT code.
This patch splits up DT board setup and introduces a DOVE_LEGACY config
to allow to remove legacy code for DT-only kernels.
Signed-off-by: Sebastian Hesselbarth <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
This refreshes the dove_defconfig, and adds:
PRINTK_TIME
DEVTMPFS
EXT4
They're quite useful, and allows booting a cubox ubuntu rootfs on SD card,
since that by default uses ext4.
The rest of the churn is due to options and defaults moving around, no
functional difference.
Signed-off-by: Olof Johansson <[email protected]>
Cc: Sebastian Hesselbarth <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Now that SDIO is instantiated via DT, and the SDIO DT node has a clocks
property, we no longer need a C coded clock alias. Remove it.
Signed-off-by: Andrew Lunn <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Now that USB is instantiated via DT, and the USB DT node has a clocks
property, we no longer need a C coded clock alias. Remove it.
Signed-off-by: Andrew Lunn <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The OpenBlocks AX3-4 platform has several LEDs, so it sounds wise to
enable LED support in mvebu_defconfig. We anticipate that more
platforms using Marvell EBU SoCs will have LEDs in the future.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The Globalscale Mirabox platform, based on the Armada 370 from
Marvell, has a SD8787 Wireless/Bluetooth chip connected on the SDIO
interface. Now that the mvsdio has a Device Tree binding, and the
necessary Device Tree informations have been added at the SoC and
board level, let's enable the btmrvl driver for the Bluetooth part of
the SD8787 chip.
For now, the driver gets probed correctly, detects the device but
apparently fails to push the firmware to the device:
Bluetooth: vendor=0x2df, device=0x911a, class=255, fn=2
Bluetooth: FW failed to be active in time!
Bluetooth: Downloading firmware failed!
Bluetooth: vendor=0x2df, device=0x911b, class=255, fn=3
Bluetooth: FW failed to be active in time!
Bluetooth: Downloading firmware failed!
This will have to be investigated separately.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The Globalscale Mirabox platform, based on the Armada 370 from
Marvell, has a SD8787 Wireless chip connected on the SDIO
interface. Now that the mvsdio has a Device Tree binding, and the
necessary Device Tree informations have been added at the SoC and
board level, let's enable the mwifiex driver for the Wireless part of
the SD8787 chip.
For now, the driver gets probed correctly, detects a device and shows
the network interfaces. However, scanning Wifi networks doesn't work
for now, with a 'CMD_RESP: cmd 0x6 error, result=0x1' message. This
will have to be investigated separately.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Now that the mvsdio driver has gained Device Tree support and the
necessary Device Tree informations has been added for Armada 370 and
Armada XP platforms, we enable the MMC subsystem and the mvsdio driver
in mvebu_defconfig.
Signed-off-by: Thomas Petazzoni <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
The RTC class driver is already part of the mvebu_defconfig but the
Marvell internal RTC not yet. Now that its support is added for mvebu
let's update the config file.
Signed-off-by: Gregory CLEMENT <[email protected]>
Tested-by: Florian Fainelli <[email protected]>
Acked-by: Arnd Bergmann <[email protected]>
Signed-off-by: Jason Cooper <[email protected]>
|
|
Basing the mvebu patches on top of the timer cleanup
avoids some nasty merges.
Signed-off-by: Arnd Bergmann <[email protected]>
|
|
These patches from the mmc tree were merged into v3.9 already
and the later mvebu patches depend on them.
Signed-off-by: Arnd Bergmann <[email protected]>
|
|
It doesn't appear that anyone actually needs to connect asynchronously.
Also, using a workqueue for the connect means we lose the namespace
information from the original process. This is a problem since there's
no way to explicitly pass in a filesystem namespace for resolution of an
AF_LOCAL address.
Acked-by: Trond Myklebust <[email protected]>
Signed-off-by: J. Bruce Fields <[email protected]>
|
|
Pages get the PG_writeback flag set before cifs sends its
request to SMB server in cifs_writepages(), if the SMB service
goes down, cifs may try to recommit the writing requests in
cifs_writev_requeue(). However, it does not clean its PG_writeback
flag and relaimed the pages even if it fails again in
cifs_writev_requeue(), which may lead to the hanging of the
processes accessing the cifs directory. This patch just cleans
the PG_writeback flags and reclaims the pages under that circumstances.
Steps to reproduce the bug(trying serveral times may trigger the issue):
1.Write from cifs client continuously.(e.g dd if=/dev/zero of=<cifs file>)
2.Stop SMB service from server.(e.g service smb stop)
3.Wait for two minutes， and then start SMB service from
server.(e.g service smb start)
4.The processes which are accessing cifs directory may hang up.
Signed-off-by: Ouyang Maochun <[email protected]>
Signed-off-by: Jiang Yong <[email protected]>
Tested-by: Zhang Xianwei <[email protected]>
Reviewed-by: Wang Liang <[email protected]>
Reviewed-by: Cai Qu <[email protected]>
Reviewed-by: Jiang Biao <[email protected]>
Reviewed-by: Jeff Layton <[email protected]>
Reviewed-by: Pavel Shilovsky <[email protected]>
Signed-off-by: Steve French <[email protected]>
|
|
The git commit 8eaffa67b43e99ae581622c5133e20b0f48bcef1
(xen/pat: Disable PAT support for now) explains in details why
we want to disable PAT for right now. However that
change was not enough and we should have also disabled
the pat_enabled value. Otherwise we end up with:
mmap-example:3481 map pfn expected mapping type write-back for
[mem 0x00010000-0x00010fff], got uncached-minus
------------[ cut here ]------------
WARNING: at /build/buildd/linux-3.8.0/arch/x86/mm/pat.c:774 untrack_pfn+0xb8/0xd0()
mem 0x00010000-0x00010fff], got uncached-minus
------------[ cut here ]------------
WARNING: at /build/buildd/linux-3.8.0/arch/x86/mm/pat.c:774
untrack_pfn+0xb8/0xd0()
...
Pid: 3481, comm: mmap-example Tainted: GF 3.8.0-6-generic #13-Ubuntu
Call Trace:
[<ffffffff8105879f>] warn_slowpath_common+0x7f/0xc0
[<ffffffff810587fa>] warn_slowpath_null+0x1a/0x20
[<ffffffff8104bcc8>] untrack_pfn+0xb8/0xd0
[<ffffffff81156c1c>] unmap_single_vma+0xac/0x100
[<ffffffff81157459>] unmap_vmas+0x49/0x90
[<ffffffff8115f808>] exit_mmap+0x98/0x170
[<ffffffff810559a4>] mmput+0x64/0x100
[<ffffffff810560f5>] dup_mm+0x445/0x660
[<ffffffff81056d9f>] copy_process.part.22+0xa5f/0x1510
[<ffffffff81057931>] do_fork+0x91/0x350
[<ffffffff81057c76>] sys_clone+0x16/0x20
[<ffffffff816ccbf9>] stub_clone+0x69/0x90
[<ffffffff816cc89d>] ? system_call_fastpath+0x1a/0x1f
---[ end trace 4918cdd0a4c9fea4 ]---
(a similar message shows up if you end up launching 'mcelog')
The call chain is (as analyzed by Liu, Jinsong):
do_fork
--> copy_process
--> dup_mm
--> dup_mmap
--> copy_page_range
--> track_pfn_copy
--> reserve_pfn_range
--> line 624: flags != want_flags
It comes from different memory types of page table (_PAGE_CACHE_WB) and MTRR
(_PAGE_CACHE_UC_MINUS).
Stefan Bader dug in this deep and found out that:
"That makes it clearer as this will do
reserve_memtype(...)
--> pat_x_mtrr_type
--> mtrr_type_lookup
--> __mtrr_type_lookup
And that can return -1/0xff in case of MTRR not being enabled/initialized. Which
is not the case (given there are no messages for it in dmesg). This is not equal
to MTRR_TYPE_WRBACK and thus becomes _PAGE_CACHE_UC_MINUS.
It looks like the problem starts early in reserve_memtype:
if (!pat_enabled) {
/* This is identical to page table setting without PAT */
if (new_type) {
if (req_type == _PAGE_CACHE_WC)
*new_type = _PAGE_CACHE_UC_MINUS;
else
*new_type = req_type & _PAGE_CACHE_MASK;
}
return 0;
}
This would be what we want, that is clearing the PWT and PCD flags from the
supported flags - if pat_enabled is disabled."
This patch does that - disabling PAT.
CC: [email protected] # 3.3 and further
Reported-by: Sander Eikelenboom <[email protected]>
Reported-and-Tested-by: Konrad Rzeszutek Wilk <[email protected]>
Reported-and-Tested-by: Stefan Bader <[email protected]>
Signed-off-by: Konrad Rzeszutek Wilk <[email protected]>
|
|
Get rid of this one (false positive):
arch/s390/kernel/module.c: In function ‘apply_relocate_add’:
arch/s390/kernel/module.c:404:5: warning: ‘rc’ may be used uninitialized
in this function [-Wmaybe-uninitialized]
arch/s390/kernel/module.c:225:6: note: ‘rc’ was declared here
Play safe and preinitialize rc with an error value, so we see an error
if new users indeed don't initialize it.
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
Remove unused function parameters.
Signed-off-by: Stefan Raspl <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
When the kernel resides in home space and the mvcos instruction is not
available uaccesses for kernel ds happen via simple strnlen() or memcpy()
calls.
This however can break badly, since uaccesses in kernel space may fail as
well, especially if CONFIG_DEBUG_PAGEALLOC is turned on.
To fix this implement strnlen_kernel() and copy_in_kernel() functions
which can only be used by the page table uaccess functions. These two
functions detect invalid memory accesses and return the correct length
of processed data.. Both functions are more or less a copy of the std
variants without sacf calls.
Fixes ipl crashes on 31 bit machines as well on 64 bit machines without
mvcos. Caused by changing the default address space of the kernel being
home space.
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
The "standard" and page table walk variants of strncpy_from_user() first
check the length of the to be copied string in userspace.
The string is then copied to kernel space and the length returned to the
caller.
However userspace can modify the string at any time while the kernel
checks for the length of the string or copies the string. In result the
returned length of the string is not necessarily correct.
Fix this by copying in a loop which mimics the mvcos variant of
strncpy_from_user(), which handles this correctly.
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
Just add s390 to the list of architectures that don't want this driver.
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
We are using sizeof operator for an array given as function argument,
which is incorrect.
Signed-off-by: Syam Sidhardhan <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
access_ok() always returns 'true' on s390. Therefore all calls
are quite pointless and can be removed.
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
If the maximum length specified for the to be accessed string for
strncpy_from_user() and strnlen_user() is zero the following incorrect
values would be returned or incorrect memory accesses would happen:
strnlen_user_std() and strnlen_user_pt() incorrectly return "1"
strncpy_from_user_pt() would incorrectly access "dst[maxlen - 1]"
strncpy_from_user_mvcos() would incorrectly return "-EFAULT"
Fix all these oddities by adding early checks.
Reviewed-by: Gerald Schaefer <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
Always stay within page boundaries when copying from user within
strlen_user_mvcos()/strncpy_from_user_mvcos(). This allows to
shorten the code a bit and may prevent unnecessary faults, since
we copy quite large amounts of memory to kernel space.
Also directly call the mvcos variants of copy_from_user() to
avoid indirect branches.
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
Failfast bit was set incorrectly.
Use set_bit to enable failfast.
Reviewed-by: Stefan Weinhuber <[email protected]>
Signed-off-by: Stefan Haberland <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|
|
Add hint to the page tables that we don't care about the change bit
in storage keys that belong to vmemmap pages.
Signed-off-by: Heiko Carstens <[email protected]>
Signed-off-by: Martin Schwidefsky <[email protected]>
|