Abhilash Kesavan [Fri, 14 Dec 2012 06:02:19 +0000 (11:32 +0530)]
arm: exynos: Enable CHIP ID clock
The bootloader may gate the CLK_CHIPID_APBIF which will cause
an invalid value to be read for the Package ID in ASV. Explicitly
enable the clock in the ASV driver.
BUG=chrome-os-partner:16796
TEST=Boot 5 different snows to the login screen. Verify that the clock
is enabled and the appropriate speed group is being selected. Do a few
suspend/resume cycles with reboots.
Doug Anderson [Fri, 14 Dec 2012 19:21:12 +0000 (11:21 -0800)]
arm: exynos: Use MIF ASV group
The original code doesn't use the MIF ASV group. Add code to read it.
There are 2 ASV groups for the fused chips ASV0 at 1.05V and ASV1 at
1V. For the non-fused chips the MIF voltage remains at 1V.
BUG=chrome-os-partner:16820
TEST=Verified that we never change MIF voltage, so this doesn't affect
us right now.
TEST=Saw printout about MIF ASV group at bootup.
Vic Yang [Thu, 13 Dec 2012 02:09:51 +0000 (18:09 -0800)]
CHROMIUM: drm/exynos: dp: Handle bad hotplug IRQ
If the hotplug detect pin cannot be mapped to an IRQ number, we should
still be able to detect the display with polling loop. This CL prevents
driver initialization from failing in this case, and adds a call to
hotplug handling function in power on sequence.
Signed-off-by: Vic Yang <victoryang@chromium.org>
BUG=chrome-os-partner:16280
TEST=on Snow and Spring, boot and see display.
Change-Id: Idaa908ac38e85b8dd10c367a4c1f8ec47a4432b9
Reviewed-on: https://gerrit.chromium.org/gerrit/40363 Reviewed-by: Sean Paul <seanpaul@chromium.org>
Commit-Queue: Vic Yang <victoryang@chromium.org> Tested-by: Vic Yang <victoryang@chromium.org> Reviewed-by: Vincent Palatin <vpalatin@chromium.org>
The fw variable is initialized in request_ihex_firmware in the
non-failure case but the compiler is not smart enough to pick
this up so fixing this way.
Fixes the following warning:
drivers/usb/serial/keyspan.c:1298:18: warning: 'fw' may be used uninitialized in this function [-Wmaybe-uninitialized]
The bootcache header has 12-character fields for date and time.
However, the __DATE__ string is only 9 characters long incl. null
terminator. When comparing __DATE__ and __TIME__ to the header's
fields, we should use the lengths of __DATE__ and __TIME__, because they
are <= 12 (size of the header fields).
e.g.
strncmp(hdr->date, __DATE__, sizeof(hdr->date))
--> sizeof(hdr->date) is 12 but __DATE__ is only 9 characters long,
so will throw a smatch error.
strncmp(hdr->date, __DATE__, strlen(__DATE__) + 1)
--> strlen(__DATE__) + 1 == 9, so this will not result in comparing
a string that is shorter than the required length.
BUG=chromium-os:37656
TEST=Build kernel with USE=test, after this CL has been merged:
https://gerrit.chromium.org/gerrit/#/c/40054/
dm-bootcache.c should not throw any smatch errors about strncmp().
Change-Id: I88c7cbb96dc4da611b96609ef21b3787483c25db Signed-off-by: Simon Que <sque@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/40549 Reviewed-by: Paul Taysom <taysom@chromium.org>
BUG=chromium-os:37579
TEST=hciconfig shows device and hcitool scan works
Change-Id: I365fd3c8c7863fc634613ae6ad18a01e600572e2 Signed-off-by: Scott James Remnant <keybuk@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/40338 Reviewed-by: Paul Stewart <pstew@chromium.org>
Bing Zhao [Thu, 3 Jan 2013 01:45:19 +0000 (17:45 -0800)]
UPSTREAM: mwifiex: check wait_event_interruptible return value
wait_event_interruptible function returns -ERESTARTSYS if it's
interrupted by a signal. Driver should check the return value
and handle this case properly.
In mwifiex_wait_queue_complete() routine, as we are now checking
wait_event_interruptible return value, the condition check is not
required. Also, we have removed mwifiex_cancel_pending_ioctl()
call to avoid a chance of sending second command to FW by other path
as soon as we clear current command node. FW can not handle two
commands simultaneously.
Cc: "3.6+" <stable@vger.kernel.org> Signed-off-by: Bing Zhao <bzhao@marvell.com> Signed-off-by: Amitkumar Karwar <akarwar@marvell.com>
BUG=chrome-os-partner:16031,17015
TEST=Boot, associate, and run "rtcwake -s 10 -mmem" in a
"while/do/done" loop
Change-Id: I54deddb71864a5b638a559b9e51942bf1d46162d
Reviewed-on: https://gerrit.chromium.org/gerrit/40359 Reviewed-by: Paul Stewart <pstew@chromium.org>
Commit-Queue: Bing Zhao <bzhao@marvell.com> Tested-by: Bing Zhao <bzhao@marvell.com>
Daniel Kurtz [Thu, 3 Jan 2013 02:37:06 +0000 (10:37 +0800)]
CHROMIUM: drm/exynos: fbdev: Set visible dimensions correctly
When probing the fbdev, the drm_fb_helper may specify a visible area
(fb_width x fb_height) than is smaller than the total requested
framebuffer area (surface_width x surface_height).
Use these fb dimensions when setting the visible area in the "variable"
portion of the fb info. This is consistent with other drm fbdev drivers
(i915, radeon, nouveau).
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=none
TEST=splash screen displayed properly at boot w/ no HDMI
TEST=splash screen displayed properly at boot w/ HDMI
Change-Id: Id209ea395a7e5eb993632ae4cfe4263b5af18138
Reviewed-on: https://gerrit.chromium.org/gerrit/40443 Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Commit-Queue: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Tested-by: Daniel Kurtz <djkurtz@chromium.org>
Daniel Kurtz [Wed, 2 Jan 2013 10:18:22 +0000 (18:18 +0800)]
CHROMIUM: drm/exynos: gem: Remove gem_init_object
This function does nothing interesting, it is only called in the unused
drm_gem_object_alloc path, and the drm gem core ignores a NULL callback
anyway.
So, save a little space by just leaving it out.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chromium-os:37097
TEST=builds clean; boot/sanity check
Change-Id: Ief8ccf91c963ddc0a0cf4b1a1a973c6675437713
Reviewed-on: https://gerrit.chromium.org/gerrit/40440 Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Commit-Queue: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Tested-by: Daniel Kurtz <djkurtz@chromium.org>
Vic Yang [Thu, 3 Jan 2013 06:40:23 +0000 (14:40 +0800)]
CHROMIUM: drm/exynos: dp: Fix a bug that IRQ number should be signed
Negative IRQ value is used to represent error.
Signed-off-by: Vic Yang <victoryang@chromium.org>
BUG=chrome-os-partner:16280
TEST=Boot on spring and see bad IRQ translated into driver error.
Change-Id: I8777c0531dceab6c952a50988764be949729e598
Reviewed-on: https://gerrit.chromium.org/gerrit/40364 Reviewed-by: Sean Paul <seanpaul@chromium.org> Reviewed-by: Vincent Palatin <vpalatin@chromium.org> Reviewed-by: Stéphane Marchesin <marcheu@chromium.org> Tested-by: Vic Yang <victoryang@chromium.org>
Commit-Queue: Vic Yang <victoryang@chromium.org>
drm/i915: Mark the FB's bo as busy when we pageflip
Previously, the page flipping case wasn't accounted for, only front buffer
rendering. This change adds a hook from the page flip code to update the
busyness of the crtc in the page flip case.
BUG=chromium-os:37277
TEST=by hand
Change-Id: Iebf826e3dc073579ca1446d7b70bcbf06d7bba5d
Reviewed-on: https://gerrit.chromium.org/gerrit/40355 Tested-by: Jonathan Backer <backer@chromium.org> Reviewed-by: Chris Wolfe <cwolfe@chromium.org>
Commit-Queue: Stéphane Marchesin <marcheu@chromium.org> Tested-by: Stéphane Marchesin <marcheu@chromium.org>
CHROMIUM: Enable CONFIG_HID_MICROSOFT in base config.
This is needed for some MS mice and should be on all
systems. Specifically this was missing from ARM based devices.
Signed-off-by: Benson Leung <bleung@chromium.org>
BUG=chromium-os:36240
TEST=Tested Link and Daisy to make sure the mouse worked on both.
Change-Id: I610d2593cd6b3e5eb2f19d4831e7e01c582db93a
Reviewed-on: https://gerrit.chromium.org/gerrit/40335 Reviewed-by: Benson Leung <bleung@chromium.org> Tested-by: Benson Leung <bleung@chromium.org>
Commit-Queue: Andrew de los Reyes <adlr@chromium.org> Tested-by: Andrew de los Reyes <adlr@chromium.org>
Che-Liang Chiou [Wed, 7 Nov 2012 02:07:56 +0000 (18:07 -0800)]
CHROMIUM: add vbc_ec implementation
This allows kernel to read/write vboot context from EC's nvram.
BUG=chrome-os-partner:15609
TEST=manual, add debugging sysfs entry to call
chromeos_vbc_read/write, and test that the
driver can actually access vboot context on ec.
Che-Liang Chiou [Wed, 7 Nov 2012 00:49:37 +0000 (16:49 -0800)]
CHROMIUM: abstract vboot context access
As vboot context is not limited to nvram, vboot context access should be
made abstract from storage device.
BUG=chrome-os-partner:15609
TEST=manual, add debugging sysfs entry to call
chromeos_vbc_read/write, and test that the driver can actually
access vboot context on nvram on link, and disk on snow.
In the implementation of change I4df35f08 the same name
is used for the macro "pos" parameter and "head" parameter.
As consequence "head" gets corrupted. this fix introduces an
auxiliar pointer for "head" and rearrange the code to make it
a bit more readable
Daniel Kurtz [Wed, 19 Dec 2012 03:26:51 +0000 (11:26 +0800)]
CHROMIUM: Input: cyapa - fix compile warning
The string pointed to by fw_name is non-mutable, so make it const char *.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chromium-os:36911
TEST=builds without following warning:
drivers/input/mouse/cyapa.c: In function 'cyapa_update_fw_store':
drivers/input/mouse/cyapa.c:1351:10: warning: assignment discards 'const' qualifier from pointer target type [enabled by default]
Change-Id: I43a3fc57a9462cece94d0c1db8011ae3084f8744
Reviewed-on: https://gerrit.chromium.org/gerrit/39934 Reviewed-by: Benson Leung <bleung@chromium.org>
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Tested-by: Daniel Kurtz <djkurtz@chromium.org>
Thomas Pedersen [Wed, 12 Dec 2012 23:30:16 +0000 (15:30 -0800)]
CHROMIUM: mwifiex: reset card on cmd timeout
If mwifiex hits a commmand timeout the firmware likely
crashed, so trigger a card reset which will clear all
logical card state and reprobe the hardware. Implement
this for sdio which requires fondling of some extra gpio
pins.
Also linearize sdio work by putting all work on the same
work queue. This ensures no work is running when the sdio
reset takes place.
Based on work by Bing Zhao.
TEST=trigger reset by debugfs hook
BUG=chrome-os-partner:15389 Signed-off-by: Thomas Pedersen <thomas@cozybit.com>
Change-Id: I3cf12ac88e7598f6daa60f38acea5e0c32302b31
Reviewed-on: https://gerrit.chromium.org/gerrit/39635 Reviewed-by: Bing Zhao <bzhao@marvell.com> Reviewed-by: Paul Stewart <pstew@chromium.org>
Some adapters return DEFER indefinitely, which results in being stuck in the
kernel, which triggers the watchdog, which reboots. To avoid this, limit to
100 tries, after which we return an error and pring a message.
BUG=chrome-os-partner:15612
TEST=by hand, use the apple miniDP to dual-link adapter, plug a 30 inch
TEST=monitor, the machine shouldn't reboot spontaneously (the monitor still
TEST=won't work, but that is a separate bug).
Doug Anderson [Fri, 14 Dec 2012 19:21:12 +0000 (11:21 -0800)]
arm: exynos: Don't force ASV group 0 to ASV group 1
The old ASV code had a bug where it would never use group 0. That
makes sense for the "original" table, but not for the new table and
not for any fused values.
BUG=chrome-os-partner:16820
TEST=Test across a wide variety of devices
Change-Id: Ifc44ed51c9d6f2effee4f0134160f74c320f4335 Signed-off-by: Doug Anderson <dianders@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/39711 Reviewed-by: Abhilash Kesavan <a.kesavan@samsung.com> Tested-by: Abhilash Kesavan <a.kesavan@samsung.com> Reviewed-by: Jon Kliegman <kliegs@chromium.org>
Doug Anderson [Fri, 14 Dec 2012 21:19:04 +0000 (13:19 -0800)]
arm: exynos: Call ABB code when using a fused speed group
The previous checkin titled "Add Adaptive Body Bias Control for
exynos5" only added ABB when not using a fused speed group. Fix it to
also use ABB when using a non-fused speed group.
BUG=chrome-os-partner:16820
TEST=Test this plus the CHIP_ID patch on a wide variety of devices,
including those that had suspend/resume problems.
Benson Leung [Sun, 16 Dec 2012 00:29:51 +0000 (16:29 -0800)]
CHROMIUM: Input: cyapa - Allow filename to be changed in update_fw
Allow the name of the designated firmware to be passed as the
argument to update_fw from user space. This will allow user space
to specify which firmware to load.
Signed-off-by: Benson Leung <bleung@chromium.org>
BUG=chromium-os:36911,chromium-os:37150
TEST=Make sure you are using chromeos-touch-firmware-daisy-2.4-r2
which contains firmware 2.4 stable, but 3 trial firmwares linked
from /lib/firmware/cyapa.bin-trial[1-3]
Check by manually issuing the command from user space:
1. echo -n cyapa.bin-trial1 > /sys/bus/i2c/devices/1-0067/update_fw
2. cat /sys/bus/i2c/devices/1-0067/firmware_version
3. double check that that returns 2.7
4. echo -n cyapa.bin-trial3 > /sys/bus/i2c/devices/1-0067/update_fw
5. cat /sys/bus/i2c/devices/1-0067/firmware_version
6. double check that that returns 2.9
7. echo -n cyapa.bin > /sys/bus/i2c/devices/1-0067/update_fw
8. cat /sys/bus/i2c/devices/1-0067/firmware_version
9. Check that that returns 2.4
Sonny Rao [Fri, 14 Dec 2012 22:48:30 +0000 (14:48 -0800)]
CHROMIUM: cpufreq: cpu_thermal only allows lower max frequency, fix it
The cpu_thermal generic thermal management code has a bug where once
max cpu frequency has been lowered in sysfs (scaling_max_freq) it is
not possible to raise the max back up later. The bug is that the
notifer gets called by __cpufreq_set_policy() before the user policy
max is raised, and is incorrectly trying to enforce the max frequency
policy even when we are trying to change the policy.
agnescheng [Thu, 13 Dec 2012 08:44:05 +0000 (16:44 +0800)]
CHROMIUM: rts_pstor: fix formatting, writing, copying failures with
specific SDSDB cards.
Patch from Realtek to correct init option for specific SD cards -
Realtek found those cards cannot be successfully written (read is OK)
because current mount option causes writes(maybe to update the access time) when read.
TEST=insert SDSDB/SDSC card, format and cannot copy files to the card.
CHROMIUM: atkbd: Fix scancode handling on reconnect / resume from suspend.
On resume from suspend there is a possibility for multi-byte scancodes
to be handled incorrectly. atkbd_reconnect disables the processing of
scancodes in software by calling atkbd_disable, but the keyboard may
still be active because no disconnect command was sent. Later, software
handling is re-enabled. If a multi-byte scancode sent from the keyboard
straddles the re-enable, only the latter byte(s) will be handled.
In practice, this leads to cases where multi-byte break codes (ex. "e0
4d" - break code for right-arrow) are misread as make codes ("4d" - make
code for numeric 6), leading to one or more unwanted, untyped characters
being interpreted.
The solution implemented here involves sending command f5 (reset
disable) to the keyboard prior to disabling software handling of codes.
Later, the command to re-enable the keyboard is sent only after we are
prepared to handle scancodes.
TEST=spam right/left arrow keys during resume for many iterations,
verify that no phantom characters are seen.
BUG=chrome-os-partner:16605
Change-Id: Ieba1e42398c2f4f6a1623afa3ed82903dc393045 Signed-off-by: Shawn Nematbakhsh <shawnn@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/39690 Reviewed-by: Derek Basehore <dbasehore@chromium.org> Reviewed-by: Andrew de los Reyes <adlr@chromium.org>
Doug Anderson [Fri, 14 Dec 2012 16:56:34 +0000 (08:56 -0800)]
arm: exynos: Fix potential buffer overrun in printk with asv
The asv code stores 5 characters in in a 5-byte array. It properly
compares against this with strncmp, but then goes and passes the
character array to printk() which could read off the end.
Terminate the array properly, then go ahead and promote the printout
to print all the details we have every time.
BUG=chrome-os-partner:16820
TEST=Boot and see message printed out.
Ray Smith [Tue, 20 Nov 2012 12:53:23 +0000 (12:53 +0000)]
MALI: Allow kbase to cancel a lock before it is acquired
This is based on Ib42b8e5c8039fd5f900885f097418b87d9614225,
most of which has been integrated into the wk45 release but
this particular commit is to account for local changes to
KDS in this repo.
Michael Spang [Wed, 14 Nov 2012 22:51:39 +0000 (17:51 -0500)]
CHROMIUM: gobi: Take clients_lock in qc_deregister
It is not legal to access the clients list without taking
clients_lock. This change makes qc_deregister take the lock while
finding new clients to free.
This fixes a crash when qc_deregister races devqmi_release.
Sonny Rao [Mon, 10 Dec 2012 19:06:02 +0000 (11:06 -0800)]
CHROMIUM: sysrq: add ability for sysrq-x to signal chrome/X
Because we weren't getting reports of actual kernel issues, and
we've already trained some users to hit sysrq-x when the system
appears to freeze, we decided to have sysrq-x try a few other things
in addition to crashing the kernel. For the first keypress it will
try to find the Chrome browser process, by looking for the one which
has the session_manager as a parent, and send a SIGABRT to that. If
there is a second sysrq-x within 5 seconds, it will send a SIGABRT to
the X process, and if there's a third then it will crash the kernel as
before.
BUG=chromium-os:36901
TEST=manual (debugging feature):
hold down alt-volup then x - chrome should crash
repeat immediately afterward - session should restart
repeat and machine should reboot
Daniel Kurtz [Mon, 10 Dec 2012 07:34:24 +0000 (15:34 +0800)]
CHROMIUM: drm/exynos: fbdev: Offset is always zero
A freshly filled fb_info will always have .xoffset and .yoffset = 0.
So, computing an offset during during fbdev_update is not necessary,
since it is only called in the fbdev_probe/fbdev_create path.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chromium-os:37097
TEST=compile test; manual check fbdev created and shows chromeball at boot
Change-Id: I648e16200cf3840a69e59a8f8d4dea69541c9cf7
Reviewed-on: https://gerrit.chromium.org/gerrit/39479 Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Tested-by: Daniel Kurtz <djkurtz@chromium.org>
Daniel Kurtz [Mon, 10 Dec 2012 05:53:11 +0000 (13:53 +0800)]
CHROMIUM: drm/exynos: gem: Remove useless size field
exynos_drm_gem_obj, drm_gem_object, the dma_buf and the exynos_drm_gem_buf
all store a buffer's size. This seems a bit excessive.
Remove the exynos_drm_gem_obj copy, since it is written to but never read.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chromium-os:37097
TEST=compile test
Change-Id: I5bf057964351d08f40146f0e928e6eb8414f164f
Reviewed-on: https://gerrit.chromium.org/gerrit/39475
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Tested-by: Daniel Kurtz <djkurtz@chromium.org>
Daniel Kurtz [Mon, 10 Dec 2012 05:51:40 +0000 (13:51 +0800)]
CHROMIUM: drm/exynos: Remove vaddr from drm_overlay
Consumers of the drm_overlay vaddr object only care about the dma_addr,
so they can program DMA in a display controller. They never need to know
the corresponding vaddr.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chromium-os:37097
TEST=compile test; booted w/ debug enabled and verified debug messages
no longer include vaddr
Change-Id: Ia96ebea8f39d4a733bc6732b6cc9be85f1a8c0e3
Reviewed-on: https://gerrit.chromium.org/gerrit/39474
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Tested-by: Daniel Kurtz <djkurtz@chromium.org>
CHROMIUM: drm/exynos: don't warn about non-kds modeset
Now that the page_flip logic is shared with the mode_set code,
it's possible for us to call page_flip on a non-kds buffer.
Modify the warning to only check the non-modeset case.
Olof Johansson [Fri, 7 Dec 2012 19:42:19 +0000 (11:42 -0800)]
CHROMIUM: LSM: allow old style loads when locking disabled
This is needed when our kernel is used with other userspaces, i.e. Ubuntu
and other distros.
Enabling old-style module init when locking is explicitly disabled doesn't
open up any security exposure for users unless they have disabled
rootfs_verification AND module load restrictions. I.e. default users have
no added exposure.
BUG=chromium-os:37055
TEST=needs a new testcase (crosbug.com/37056)
This patch switches zcache and zram dependency to ZSMALLOC
rather than X86. There is no net change since ZSMALLOC
depends on X86, however, this prevent further changes to
these files as zsmalloc dependencies change.
Signed-off-by: Seth Jennings <sjenning@linux.vnet.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
BUG=chromium-os:36829
TEST=compile, verify that zram module is updated and /boot/config is correct
Change-Id: I91f6ed46549fe62a9fe3035063d98beb2f5d966a Signed-off-by: Luigi Semenzato <semenzato@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/39289 Tested-by: Luigi Semenzato <semenzato@google.com> Reviewed-by: Sonny Rao <sonnyrao@chromium.org>
Commit-Ready: Luigi Semenzato <semenzato@google.com>
John Sheu [Thu, 6 Dec 2012 01:23:26 +0000 (17:23 -0800)]
CHROMIUM: exynos: keep exynos_plane zpos on disable
Exynos DRM planes are implemented using overlay hardware. Presently we
reset the overlay zpos back to default every time a plane is disabled.
Don't do this anymore.
Signed-off-by: John Sheu <sheu@chromium.org>
BUG=chromium-os:36991
TEST=local build, run on exynos
Change-Id: I4c3490ef05d1ecdfeea6cbf26f020b9a19ebe764
Reviewed-on: https://gerrit.chromium.org/gerrit/39296 Tested-by: John Sheu <sheu@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Commit-Ready: John Sheu <sheu@chromium.org>
Michael Spang [Fri, 9 Nov 2012 19:33:47 +0000 (14:33 -0500)]
CHROMIUM: serial: exynos: Resume UART earlier with SAMSUNG_PM_DEBUG
Even with SAMSUNG_PM_DEBUG enabled, we're losing lots of console UART
output during resume. This patch makes SAMSUNG_PM_DEBUG correcly
resume the UART as early as possible, so that no output is lost.
BUG=chrome-os-partner:10932
TEST=powerd_suspend with no_console_suspend and SAMSUNG_PM_DEBUG
Change-Id: Ic892ed225c9075e8d72d851a4f2e22263ee6047d Signed-off-by: Michael Spang <spang@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/37734 Reviewed-by: Jon Kliegman <kliegs@chromium.org>
Dylan Reid [Mon, 19 Nov 2012 06:56:40 +0000 (22:56 -0800)]
BACKPORT: ALSA: hda - Fix Acer Aspire models with analog mics.
The Acer Aspire AO756 has an analog built-in mic on nid 0x1b and an
external mic on nid 0x19. The BIOS doesn't set this up.
The mic detect on this Acer Aspire netbook and Acer C7 ChromeBook is
only valid when the headphone is plugged. The detect circuit relies on
the tip detect switch being closed on the jack. Tell hda_jack to ignore
the mic sense unless the headphones are plugged.
Dylan Reid [Mon, 19 Nov 2012 18:48:07 +0000 (19:48 +0100)]
BACKPORT: ALSA: hda - Allow jack state to depend on another jack
Introduce the concept of a "gated" jack. The gated jack's pin sense
is
only valid when the "gating" jack is plugged. This requires checking
the gating jack when the gated jack changes and re-checking the gated
jack when the gating jack is plugged/unplugged.
This allows handling of devices where the mic jack detect floats when
the headphone jack is unplugged.
[Rewritten for fixing the possible snd_array reallocation, covering
the missing callback calls and jack sync operations, as well as some
code cleanups -- tiwai]
Michael Spang [Fri, 9 Nov 2012 01:06:06 +0000 (20:06 -0500)]
CHROMIUM: serial: samsung: Avoid waiting forever for TX ready
The no_console_suspend option allows the console to write to the UART
during early resume, before the serial port itself is resumed. Such
writes hang indefinitely waiting for TX ready.
This adds a check to the console putchar function to drop characters
instead of hanging if the port hasn't been initialized. That way, we
can use no_console_suspend without risk of hanging.
BUG=chrome-os-partner:10932
TEST=powerd_suspend with no_console_suspend but not SAMSUNG_PM_DEBUG
Change-Id: I64e8def33fe984b97776206451e6030d6c761355 Signed-off-by: Michael Spang <spang@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/37714
Vincent Palatin [Fri, 16 Nov 2012 17:32:27 +0000 (09:32 -0800)]
CHROMIUM: exynos: dts: update configuration for Spring proto-1
- we have now an HDMI output, re-activate it.
- disable the OHCI controller until the hardware is fixed :
This OHCI controller has no individual port power support,
so activating the controller always enables the port power on port 0
independantly of the EHCI state or the port power bit. This will then
turn on the VBUS regulator and might create an electrical conflict
with the USB boost.
- disable port 0 on EHCI to keep the VBUS regular off.
- tune the eDP bridge regulator configuration :
it needs to be on even if the bootloader is not using the screen.
- disable high-speed mode on Wifi SDIO until we find the issue.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BUG=chrome-os-partner:14489
TEST=boot on Spring.
Change-Id: Ibcfc2848d932026a711833471af908867659f74b
Reviewed-on: https://gerrit.chromium.org/gerrit/38223
Commit-Ready: Vincent Palatin <vpalatin@chromium.org> Tested-by: Vincent Palatin <vpalatin@chromium.org> Reviewed-by: Olof Johansson <olofj@chromium.org>
Vincent Palatin [Fri, 16 Nov 2012 20:22:44 +0000 (12:22 -0800)]
CHROMIUM: usb: allow to de-activate some ports on EHCI root hub
To optimize power consumption and sometimes avoid electrical conflicts
on the board, we might want to avoid activating some of the ports on
the root hub.
The platform specific part of the driver can optionally provide a bitmap
of the enabled ports. By default, all ports are enabled.
This might be reverted once we have another mechanism to control VBUS
regulator on the next hardware revision.
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BUG=chrome-os-partner:14490
TEST=boot on Spring, verify that port 0 regulator is not enabled and
port 1 is detecting a USB key correctly.
boot on Snow, verify that port 0 is detecting a USB key
Change-Id: Ib08d18a105e366b37f5117a6f7c0e1e56e04dce5
Reviewed-on: https://gerrit.chromium.org/gerrit/38222
Commit-Ready: Vincent Palatin <vpalatin@chromium.org> Tested-by: Vincent Palatin <vpalatin@chromium.org> Reviewed-by: Olof Johansson <olofj@chromium.org>
Benson Leung [Tue, 20 Nov 2012 07:16:54 +0000 (23:16 -0800)]
CHROMIUM: Input: cyapa - More reliable cyapa_get_state.
In cyapa_get_state, add an additional check for OP_STATUS_SRC bit
in REG_OP_STATUS. This bit should only be set when in operational
mode.
Under some conditions, the REG_OP_STATUS may return 0x89,
with the device status bits being 0b01, indicating the operational
mode firmware is busy executing a command from the host. Handle this
status and return state CYAPA_STATE_OP if we are not in bootloader based
on OP_STATUS_SRC.
Signed-off-by: Benson Leung <bleung@chromium.org>
BUG=chrome-os-partner:16217
TEST=Test Lumpy (Samsung Series 5 550) and Snow (Samsung Chromebook ARM)
using this driver.
1. Cold boot the system (power on the system from full on)
2. Warm reboot the system (reboot the system using command sudo reboot)
Check that the touchpad is detected in both cases and is functional.
3. Force a firmware update and check that it succeeds.
/opt/google/touchpad/firmware/chromeos_touch_firmwareupdate -d cyapa -f
Furthermore, see crosbug.com/p/16217 for detailed repro steps.
System with this patch should survive the suspend/resume test
with a grounded metal probe touching the touchpad generating activity.
suspend_stress_test --wake_min 0 --wake_max 0 --backup_rtc
Change-Id: Ie2d513a3a868c6914776355222fad21935af88a5
Reviewed-on: https://gerrit.chromium.org/gerrit/38364 Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Dudley Du <dudl@cypress.com>
Commit-Ready: Benson Leung <bleung@chromium.org> Tested-by: Benson Leung <bleung@chromium.org>
Bing Zhao [Mon, 3 Dec 2012 19:52:13 +0000 (11:52 -0800)]
Bluetooth: btmrvl: enable RESET_ON_CLOSE quirk
When user brings down the HCI interface without explicitly
disconnecting the HID device(s) or resetting the HCI interface,
SD8797 firmware won't get the indication of "interface down".
This results that the SD8797 Wi-Fi stays at 1x1 mode without
switching back to 2x2 mode.
Work it around by setting RESET_ON_CLOSE bit to have the stack
reset the HCI interface upon close.
BUG=chrome-os-partner:14937
TEST=WiFi associates to 2.4GHz AP close enough; Bluetooth connects
to HID mouse; "iw mlan0 link" shows WiFi link tx bitrates at MCS7.
Uncheck "Enable bluetooth" checkbox, or issue "hciconfig hci0 down"
command, "iw mlan0 link" shows tx bitrates at MCS15 or MCS14.
Change-Id: Ic1468c3f60b504aa2efe80cbc72a22872b256f04 Signed-off-by: Bing Zhao <bzhao@marvell.com>
Reviewed-on: https://gerrit.chromium.org/gerrit/39163 Reviewed-by: Paul Stewart <pstew@chromium.org> Reviewed-by: Scott James Remnant <keybuk@chromium.org>
Bing Zhao [Mon, 3 Dec 2012 19:38:10 +0000 (11:38 -0800)]
UPSTREAM: Bluetooth: Rename HCI_QUIRK_NO_RESET to HCI_QUIRK_RESET_ON_CLOSE
HCI_QUIRK_NO_RESET name is misleading - purpose of this quirk is to
reset device on close instead of init, not to not reset at all.
Rename it to HCI_QUIRK_RESET_ON_CLOSE to avoid confusion.
Signed-off-by: Szymon Janc <szymon.janc@tieto.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Bing Zhao <bzhao@marvell.com>
BUG=chrome-os-partner:14937
TEST=WiFi associates to 2.4GHz AP close enough; Bluetooth connects
to HID mouse; "iw mlan0 link" shows WiFi link tx bitrates at MCS7.
Uncheck "Enable bluetooth" checkbox, or issue "hciconfig hci0 down"
command, "iw mlan0 link" shows tx bitrates at MCS15 or MCS14.
Change-Id: Ib83bdea24c9820efe321c02f10ffce43c42368e5
Reviewed-on: https://gerrit.chromium.org/gerrit/39162 Reviewed-by: Paul Stewart <pstew@chromium.org>
Commit-Ready: Bing Zhao <bzhao@marvell.com> Tested-by: Bing Zhao <bzhao@marvell.com> Reviewed-by: Scott James Remnant <keybuk@chromium.org>
According to Creative there are some extreme circumstances where the
DSP can take > 1second to respond to a command. I have asked them to
investigate further and eliminate this possibility. Until the DSP
side can be fixed, lengthen the timeout to 2 seconds so the system
avoid resuming without audio.
BUG=chrome-os-partner:14465
TEST=suspend/resume for 4k cycles, never see DSP load failure.
Doug Anderson [Mon, 3 Dec 2012 19:00:29 +0000 (11:00 -0800)]
CHROMIUM: scripts: Make genconfig clean configs
Right now genconfig just dumps out the raw output of concatenating all
of the relevant configs. This makes it hard to diff. Make it easier
by running "oldconfig" over this and also providing a savedefconfig.
BUG=None
TEST=Run ./chromeos/scripts/kernelconfig genconfig and look at
CONFIGS directory; try this even when the configs aren't normalized.
Change-Id: I19da6d54d5e550682737f0a23440fd2cc05dbd16 Signed-off-by: Doug Anderson <dianders@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/39094 Reviewed-by: Mike Frysinger <vapier@chromium.org> Reviewed-by: Mandeep Singh Baines <msb@chromium.org>
CHROMIUM: UPSTREAM: staging: zsmalloc: remove x86 dependency
This patch replaces the page table assisted object mapping
method, which has x86 dependencies, with a arch-independent
method that does a simple copy into a temporary per-cpu
buffer.
While a copy seems like it would be worse than mapping the pages,
tests demonstrate the copying is always faster and, in the case of
running inside a KVM guest, roughly 4x faster.
Ben Hutchings [Wed, 20 Jun 2012 01:31:11 +0000 (02:31 +0100)]
CHROMIUM: BACKPORT: staging: zsmalloc: Finish conversion to a separate module
Need BACKPORT, rather than UPSTREAM, since it won't apply cleanly because
we had already fixed the MODULE_LICENSE problem.
---original comment:
ZSMALLOC is tristate, but the code has no MODULE_LICENSE and since it
depends on GPL-only symbols it cannot be loaded as a module. This in
turn breaks zram which now depends on it. I assume it's meant to be
Dual BSD/GPL like the other z-stuff.
There is also no module_exit, which will make it impossible to unload.
Add the appropriate module_init and module_exit declarations suggested
by comments.
Reported-by: Christian Ohm <chr.ohm@gmx.net>
References: http://bugs.debian.org/677273 Cc: stable@vger.kernel.org # v3.4 Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Reviewed-by: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Conflicts:
By chance, struct attribute's "attr" field is first, so its address is the
address of the struct attribute, so this typo has no functional effect, it
just suppresses this compiler warning:
drivers/input/touchscreen/atmel_mxt_ts.c:2059:2: warning: initialization from incompatible pointer type [enabled by default]
drivers/input/touchscreen/atmel_mxt_ts.c:2059:2: warning: (near initialization for 'mxt_power_attrs[0]') [enabled by default]
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=chromium-os:36877
TEST=builds clean
Change-Id: If79a410afb82984472893673b97aa784249a3ca5
Reviewed-on: https://gerrit.chromium.org/gerrit/39069 Reviewed-by: Yufeng Shen <miletus@chromium.org>
Commit-Ready: Daniel Kurtz <djkurtz@chromium.org> Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> Tested-by: Daniel Kurtz <djkurtz@chromium.org>
Benson Leung [Fri, 30 Nov 2012 23:26:49 +0000 (15:26 -0800)]
CHROMIUM: Input: atmel_mxt_ts - Prevent null dereferences when in bootloader
When we are in bootloader mode, the input_dev is not valid, and neither is
the object table. Several paths may result in something trying to dereference
these. This patch addresses the following :
1) mxt_resume calls input_unregister_device on input_dev.
2) mxt_suspend/resume reads and writes from the object table.
Signed-off-by: Benson Leung <bleung@chromium.org>
BUG=chrome-os-partner:16507
TEST=First, get the touch device into a bad state by doing the following:
1. Modify chromeos/config/base.config and set CONFIG_TOUCHSCREEN_ATMEL_MXT=m
2. Build, boot this kernel, and make sure that the touch device works.
3. /opt/google/touch/firmware/chromeos-touch-firmwareupdate.sh \
-d atmel_mxt_ts -n maxtouch-ts.fw -f
4. Before it can finish, CTRL-C to interrupt the firmware update.
This will ensure that the touch device is stuck in bootloader mode.
Now, test that mxt_remove is not vulnerable:
1. rmmod atmel_mxt_ts
2. check that the system does not reboot.
3. modprobe chromeos_mxt_ts
Test that suspend/resume does is not vulnerable:
1. Close the lid to suspend the system.
2. Open the lid to suspend the system.
3. Check that the system did not reboot.
CHROMIUM: sound / hda: Fix D0 power state checking.
Power state checking for HDA cards is broken. According to the HDA spec, the
lower four bits of the power state register define the "set" power state and
the next four bits define the "actual" state. Subsequent bits have different
meanings. When we check the power state of our card during transition to D0,
we read the entire register and compare to AC_PWRST_D0 (0). This leads to
timeout of the wait loop and ~500ms wasted during resume if some upper
status bits are set.
Solution is to mask off all upper bits and compare only get + set state. If
these states are 0, it means the transition has completed.
This is fixed differently in the mainstream kernel repo, but there are some
extra dependencies involved. Therefore, this patch should NOT be carried
forward through the next kernel rebase.
DO NOT CARRY FORWARD.
TEST=Run power_Resume test on HDA card that reports PS-ClkStopOk status.
Confirm that resume time is no longer ~500 ms.
BUG=chrome-os-partner:15729
Vincent Palatin [Fri, 30 Nov 2012 17:33:44 +0000 (09:33 -0800)]
BACKPORT: x86, fpu: avoid FPU lazy restore after suspend
When a cpu enters S3 state, the FPU state is lost.
After resuming for S3, if we try to lazy restore the FPU for a process running
on the same CPU, this will result in a corrupted FPU context.
Ensure that "fpu_owner_task" is properly invalided when (re-)initializing a CPU,
so nobody will try to lazy restore a state which doesn't exist in the hardware.
Cc: Duncan Laurie <dlaurie@chromium.org> Cc: Olof Johansson <olofj@chromium.org> Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
BUG=chrome-os-partner:16295
TEST=On Link, run thousands of suspend/resume cycles with 4 processes doing FPU
operations running as described in the bug.
Without the patch, a process is killed after a few hundreds cycles by a SIGFPE.
Change-Id: I33ee884dea6cad048cbe1aa60c64cbb5141d3f66
Reviewed-on: https://gerrit.chromium.org/gerrit/39006
Commit-Ready: Vincent Palatin <vpalatin@chromium.org> Tested-by: Vincent Palatin <vpalatin@chromium.org> Reviewed-by: Olof Johansson <olofj@chromium.org>
Shan Wei [Wed, 28 Nov 2012 02:54:31 +0000 (10:54 +0800)]
UPSTREAM: staging: gdm72xx: use %pI4 format to print IPv4 address and remove last usage of NIP6
commit cf4ca4874fc45 removed the definition of NIPQUAD and NIPQUAD_FMT,
and NIP6 also is out of date.
commit 2874762b31d8d replace deprecated NIPQUAD marco to C code, but we can use %pI4 to
print IPv4 address more simply. And remove constant condition judge.
Because DEBUG_SDU is not defined in gdm_wimax.h, no error message when compiling.
Ben Chan [Tue, 27 Nov 2012 04:18:45 +0000 (20:18 -0800)]
UPSTREAM: staging: gdm72xx: fix unused variable warning in gdm_usb_send
This patch fixes an unused variable warning in gdm_usb_send
(when CONFIG_WIMAX_GDM72XX_K_MODE=n), which was introduced in
commit 1a276b80466bbd195cf94ec7178f68f2ab351467 (staging:
gdm72xx: protect access of rx / tx structs).
Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Ben Chan <benchan@chromium.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
(cherry picked from commit a7e4a982fe2a9375952798294ebfb38dc0ed6782)
Change-Id: Ifcc7f122a8342f51d697a48b55d604f71a582f52 Signed-off-by: Ben Chan <benchan@chromium.org>
Reviewed-on: https://gerrit.chromium.org/gerrit/38962 Reviewed-by: Olof Johansson <olofj@chromium.org>
YAMANE Toshiaki [Mon, 29 Oct 2012 11:05:57 +0000 (20:05 +0900)]
UPSTREAM: staging/gdm72xx: Use dev_ printks in usb_boot.c
fixed below checkpatch warnings.
- WARNING: Prefer netdev_err(netdev, ... then dev_err(dev, ... then pr_err(... to printk(KERN_ERR ...
- WARNING: Prefer netdev_info(netdev, ... then dev_info(dev, ... then pr_info(... to printk(KERN_INFO ...
YAMANE Toshiaki [Mon, 29 Oct 2012 11:05:43 +0000 (20:05 +0900)]
UPSTREAM: staging/gdm72xx: Use dev_ printks in gdm_usb.c
fixed below checkpatch warnings.
- WARNING: Prefer netdev_err(netdev, ... then dev_err(dev, ... then pr_err(... to printk(KERN_ERR ...
- WARNING: Prefer netdev_info(netdev, ... then dev_info(dev, ... then pr_info(... to printk(KERN_INFO ...
YAMANE Toshiaki [Mon, 29 Oct 2012 11:05:30 +0000 (20:05 +0900)]
UPSTREAM: staging/gdm72xx: Use dev_ printks in sdio_boot.c
fixed below checkpatch warnings.
- WARNING: Prefer netdev_err(netdev, ... then dev_err(dev, ... then pr_err(... to printk(KERN_ERR ...
- WARNING: Prefer netdev_info(netdev, ... then dev_info(dev, ... then pr_info(... to printk(KERN_INFO ...
YAMANE Toshiaki [Mon, 29 Oct 2012 11:05:16 +0000 (20:05 +0900)]
UPSTREAM: staging/gdm72xx: Use dev_ printks in gdm_sdio.c
fixed below checkpatch warnings.
- WARNING: Prefer netdev_err(netdev, ... then dev_err(dev, ... then pr_err(... to printk(KERN_ERR ...
- WARNING: Prefer netdev_info(netdev, ... then dev_info(dev, ... then pr_info(... to printk(KERN_INFO ...
YAMANE Toshiaki [Mon, 29 Oct 2012 11:04:42 +0000 (20:04 +0900)]
UPSTREAM: staging/gdm72xx: Use netdev_ or pr_ printks in gdm_qos.c
fixed below checkpatch warnings.
- WARNING: Prefer netdev_err(netdev, ... then dev_err(dev, ... then pr_err(... to printk(KERN_ERR ...
- WARNING: Prefer netdev_info(netdev, ... then dev_info(dev, ... then pr_info(... to printk(KERN_INFO ...
- WARNING: Prefer netdev_dbg(netdev, ... then dev_dbg(dev, ... then pr_debug(... to printk(KERN_DEBUG ...
projects / people / aperard / linux-chromebook.git / log