Commit Graph

159329 Commits

Author SHA1 Message Date
Will Deacon
797e410a39 ARM: 6387/1: errata: check primary part ID in proc-v7.S
Kconfig doesn't have any knowledge of specific v7 cores, so it is possible
to select errata workarounds that may cause inadvertent behaviour when
executed on a core other than those targetted by the fix.

This patch improves the variant and revision checking in proc-v7.S so
that the primary part number is also considered when applying errata
workarounds.

Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2011-01-29 09:23:15 -06:00
Russell King
48347800be ARM: udelay: prevent math rounding resulting in short udelays
We perform the microseconds to loops calculation using a number of
multiplies and shift rights.  Each shift right rounds down the
resulting value, which can result in delays shorter than requested.
Ensure that we always round up.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2011-01-29 09:21:06 -06:00
Matt Sealey
189a808937 siihdmi: binary edid property support (/sys/class/graphics/fb0/EDID) 2011-01-28 11:47:36 -06:00
Matt Sealey
3e0543205a cea861: fix hdmi version of vsdb (should match cea one, port configuration was not little endian ordered) 2011-01-27 11:02:09 -06:00
Matt Sealey
863319833d cea861: fix vendor specific data block (not an extension!) 2011-01-27 11:00:04 -06:00
Matt Sealey
6f9776c2ee siihdmi: fix hotplug, don't printk in an interrupt routine, clear ISR sooner 2011-01-27 10:57:58 -06:00
Matt Sealey
2fee5c50ff siihdmi: prefer 1280x720 if there is an exact video mode matching it 2011-01-27 10:56:51 -06:00
Matt Sealey
e4ab55fc35 siihdmi: add comment about vsync_len being too short (HDMI spec) 2011-01-27 10:55:26 -06:00
Matt Sealey
633174991e siihdmi: properly relinquish DDC bus 2011-01-27 10:54:26 -06:00
Matt Sealey
5ee41fa2a1 siihdmi: sink_present is now more chatty 2011-01-27 10:53:01 -06:00
Matt Sealey
e47e394903 siihdmi: remove instrumentation from bus timeouts 2011-01-27 10:52:09 -06:00
Matt Sealey
ef68e1e100 siihdmi: bus timeout back to 50, since jiffies_to_ms has a maximum resolution of 10ms 2011-01-27 10:51:15 -06:00
Matt Sealey
3b15ac9969 siihdmi: fix config for hotplug 2011-01-27 10:50:05 -06:00
Saleem Abdulrasool
59992d3a3d sbs: handle signed integer values correctly 2011-01-26 21:40:15 -08:00
Saleem Abdulrasool
dd755017f9 sbs: always report a value, 0 if the value is unavailable 2011-01-26 20:46:49 -08:00
Saleem Abdulrasool
5070e11542 sbs: rename platform insertion status functions 2011-01-26 20:42:18 -08:00
Saleem Abdulrasool
2b688d086c sbs: uniform names
current is a macro for the current process.  Use _current instead of current_now
as all the other cached fields are named the same as the register they
correspond to.  Also, add in a comment about the special relationship between
_serial_number and serial_number.
2011-01-26 19:34:30 -08:00
Saleem Abdulrasool
7dc6e014d1 sbs: reorganise function layout 2011-01-26 19:31:23 -08:00
Saleem Abdulrasool
1c11307698 sbs: clearer names for the battery/mains 2011-01-26 19:28:56 -08:00
Saleem Abdulrasool
7953c7ad6b sbs: dont leak memory
The power supply system does not actually expect allocated strings that it will
manage.  We were allocating the strings, expecting the power supply class driver
to clean them up, and as a result had a massive leak.  Fix this by storing the
string in the cache and returning a reference to that string.

This effectively overhauls the sbs_battery_field.  It is renamed
sbs_battery_register which is more accurate.  There is now string handling logic
for the string reading case.  The int/uint types are handled the same, so there
is a single read type for those two value types.
2011-01-26 19:08:34 -08:00
Saleem Abdulrasool
d07ef9f5f7 sbs: make mWh reporting correct 2011-01-26 19:08:34 -08:00
Matt Sealey
86e7bc45a7 mx: fixes for platform display code
* cleanup (useless definitions)
* fix IOMUX for reset pin definition (make it a straight boring GPIO)
* add IRQ pin IOMUX (can't believe I missed that)
* add DISP_EN lines from schematics (active low), and knock VGA on the head (might save a few uA)
2011-01-26 02:29:47 -06:00
Saleem Abdulrasool
7c7076df28 sbs: x^0 = 1, not 0 2011-01-25 19:48:45 -08:00
Saleem Abdulrasool
f75f50eefe sbs: current_now is a signed value 2011-01-25 19:48:45 -08:00
Saleem Abdulrasool
c1cb15151c sbs: slightly clearer return path 2011-01-25 19:48:45 -08:00
Saleem Abdulrasool
955db47146 power: fix manufacturer and serial number reporting in sysfs 2011-01-25 19:48:44 -08:00
Saleem Abdulrasool
53eca9c65d sbs: support mains/battery availability checks 2011-01-25 19:48:44 -08:00
Matt Sealey
9d1d18ad66 siihdmi: cull modes that have too short vsync_len. Not a nice way to specify to cull these modes in platform data yet. Yet another reason for a DRM connector and not this driver framework...
ipu: no longer platform-specific hack short vsync_len out, since we cull them in the siihdmi driver. Makes our ipu code == fsl code.
2011-01-25 16:01:29 -06:00
Saleem Abdulrasool
fffe14628f sbs: handle strings correctly 2011-01-25 00:04:48 -08:00
Matt Sealey
6f61305258 synaptics-usb 1.5rc8 for hacking around with the touchpad (no mods made for detection yet) 2011-01-24 11:56:07 -06:00
Saleem Abdulrasool
5e231689ba add new SBS driver 2011-01-24 11:37:35 -06:00
Matt Sealey
e5feb6097b sb: add brightness keys 2011-01-24 09:39:08 -06:00
Matt Sealey
564d3b01a9 siihdmi: clean up siihdmi reset, remove more cs8556 stuff, better description
of IPU errors and film modes.
2011-01-24 07:41:39 -06:00
Matt Sealey
84dc49b8fa Move devtmpfs configs to main defconfigs 2011-01-23 15:45:44 -06:00
Matt Sealey
bf90bf4d24 remove useless unupdated default configurations 2011-01-23 15:43:44 -06:00
Saleem Abdulrasool
48325687a2 Kconfig: siihdmi is not useful as a module 2011-01-23 12:01:06 -08:00
Saleem Abdulrasool
9b2cb056bd siihdmi: we no longer need to be tied to the machine details 2011-01-23 11:59:36 -08:00
Saleem Abdulrasool
b978e4e8ab siihdmi: rename mxcfb_siihdmi.c to siihdmi.c 2011-01-23 11:57:07 -08:00
Saleem Abdulrasool
c47513f811 sii9022: remove this driver entirely in favour of siihdmi 2011-01-23 00:12:07 -08:00
Saleem Abdulrasool
60efc2b1c1 mx51: delete unused display variables 2011-01-23 00:12:07 -08:00
Saleem Abdulrasool
b1660b0567 [arm] platform: remove unused edid code
This code is no longer used, just remove it.
2011-01-23 00:12:06 -08:00
Saleem Abdulrasool
f0b5078568 siihdmi: sii_hdmi.h -> siihdmi.h 2011-01-23 00:12:06 -08:00
Saleem Abdulrasool
b3237f5489 siihdmi: move video mode selection into its own function 2011-01-23 00:12:06 -08:00
Saleem Abdulrasool
8fc68cb0aa siihdmi: max pixclock is platform specific data 2011-01-23 00:12:06 -08:00
Saleem Abdulrasool
bdffde6c3f siihdmi: cleanup siihdmi_santize_modelist
Use list_for_each_entry_safe to make the code easier to follow, re-order the
checks for removal, and break once the remove flag is set in the list traversal.
2011-01-23 00:12:06 -08:00
Saleem Abdulrasool
7549523dfb siihdmi: improve siihdmi_dump_modelines
Make dump_modelines dump videomodes rather than monspecs so that we see the
actual supported video modelines.
2011-01-23 00:12:06 -08:00
Saleem Abdulrasool
d7d2fc3aae siihdmi: move res_equal into modedb
The comparitor function is not specific to SII9xxx, but a generic modedb
function.  Move it to a more appropriate home.
2011-01-23 00:12:05 -08:00
Saleem Abdulrasool
1407fa0f09 siihdmi: kill dead code 2011-01-23 00:12:05 -08:00
Saleem Abdulrasool
241488d963 siihdmi: better driver name for IRQ 2011-01-23 00:12:05 -08:00
Saleem Abdulrasool
eceb13787a siihdmi: reset if the reset method is provided 2011-01-23 00:12:05 -08:00