Skip to content

L
linux-davinci
Switch branch/tag
  1. 11 Nov, 2006 5 commits
  3. 10 Nov, 2006 6 commits
  5. 09 Nov, 2006 13 commits
    • Arnaud Patard's avatar
      dsp_fbexport: use kzalloc · becd02ee
      Arnaud Patard authored 
      The dsp_export function calls omapfb_register_client() right after
allocating the memory with kmalloc.
A commit added to omapfb_register_client() a check on the value of
plane_idx in the notified block structure. As we're only using kmalloc,
you may end-up comparing garbage/random values (It even leads to a oops
on my n770).

Using kzalloc() fixes the issue.
Signed-off-by: default avatarArnaud Patard <arnaud.patard@rtp-net.org>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      becd02ee
    • Arnaud Patard's avatar
      Fix some pt_regs users · d5c772f5
      Arnaud Patard authored 
      There are still some users of pt_regs in the -omap tree. This patch fixes them.
Signed-off-by: default avatarArnaud Patard <arnaud.patard@rtp-net.org>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      d5c772f5
    • David Brownell's avatar
      omap GP timer: HZ != 100 · 4b9fe9f9
      David Brownell authored 
      Teach OMAP2 gp timer that HZ isn't always 100.
Signed-off-by: default avatarDavid Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      4b9fe9f9
    • andrzej zaborowski's avatar
      ARM: OMAP: Convert gpio-switch to platform_driver · 6433b741
      andrzej zaborowski authored 
      Converts the gpio-switch driver to platform_driver to avoid bad cast
which occurs because the bus type is set to platform bus. Also
silences the following warnings:

arch/arm/plat-omap/gpio-switch.c:292: warning: ignoring return value
of 'device_create_file', declared with attribute warn_unused_result
arch/arm/plat-omap/gpio-switch.c:293: warning: ignoring return value
of 'device_create_file', declared with attribute warn_unused_result
arch/arm/plat-omap/gpio-switch.c:294: warning: ignoring return value
of 'device_create_file', declared with attribute warn_unused_result
Signed-off-by: default avatarAndrzej Zaborowski <balrog@zabor.org>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      6433b741
    • Dirk Behme's avatar
      ARM: OMAP: Fix warnings in ads7846 · d9e757b4
      Dirk Behme authored 
      Fix various warnings

warning: ignoring return value of 'device_create_file',
declared with attribute warn_unused_result

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      d9e757b4
    • Dirk Behme's avatar
      ARM: OMAP: Fix warnings in plat-omap/dsp · 15bc5db0
      Dirk Behme authored 
      Fix various warnings

warning: ignoring return value of 'xxx', declared with
attribute warn_unused_result

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      15bc5db0
    • Dirk Behme's avatar
      ARM: OMAP: Fix warning in mach-omap1 · 4da53389
      Dirk Behme authored 
      Fix warning

arch/arm/mach-omap1/pm.c: In function 'omap_pm_init':
arch/arm/mach-omap1/pm.c:765: warning: ignoring return value
of 'subsys_create_file', declared with attribute
warn_unused_result

Signed-off-by: Dirk Behme <dirk.behme_at_gmail.com>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      4da53389
    • David Brownell's avatar
      ARM: OMAP: abstract debug card setup (smc, leds) · ae0ef82b
      David Brownell authored 
      Additional cleanup for debug boards on H2/P2/H3/H4:  move the init
code that's not board-specific into a new file where it can be easily
shared between all the different boards (avoiding code duplication,
and making it easier to support more devices).  Make H4 use that.

This should be easy to drop in to the OMAP1 boards using these debug
cards; the only difference seems to be that the p2 does an extra reset
of the smc using the fpga (probably all boards could do that, if it's
necessary) and doesn't use the gpio mux or request APIs.
Signed-off-by: default avatarDavid Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      ae0ef82b
    • David Brownell's avatar
      ARM: OMAP: h4 must have blinky leds!! · e37758d2
      David Brownell authored 
      This adds generic support for the "debug board" LEDs used by most of
TI's OMAP reference boards, and board-specific support for the H4.

It's derived from the not-as-generic stuff used by OMAP1 H2/H3/P2.
Those should be able to switch easily to this version, and clean up
some of the omap1-specific code.

In addition to H4 support, one key improvement is supporting not just
the "old" ARM debug LED API (with timer and idle LEDs, plus four that
can be handy for kernel debugging), but it also supports the "new"
generic LED API (most useful for usermode stuff IMO).  Either or both
APIs can be enabled.
Signed-off-by: default avatarDavid Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: default avatarTony Lindgren <tony@atomide.com>
      e37758d2
    • David Brownell's avatar
      musb_hdrc: davinci · 636fa27e
      David Brownell authored 
      Update how DaVinci handles powering up VBUS, and make cable based role
switching work more or less as well as on TUSB6010.

  - Work around VBUS_ERROR IRQs triggering during enumeration, which
    seems to be a design botch from Mentor (TUSB 6010 has it too).

	This is particularly nasty on the EVM (30 second delay) because
	of differences between its RTL and the newer stuff in tusb6010.

  - Work around lack of ID change IRQ, a design botch in how the Mentor
    core was integrated into DaVinci.

	This builds on the polling mechanism used to cope with VBUS
	errors.  OTG devices will probe every few seconds until either
	the ID pin grounds, or the device leaves B_IDLE.

  - Rename vbus_power() to source_power(), better matching its role.
    (DaVinci doesn't sink VBUS e.g. to charge batteries -- yet.)

  - Provide more complete tracing of IRQ events.

  - Provide a symbol for the DRVVBUS irq.

Basically, with this driver's core behaving better (thanks to OTG/host
updates for TUSB), the DaVinci stuff doesn't need to fight them so much;
and it's more clear when problems are in the core vs its integration.
Signed-off-by: default avatarDavid Brownell <dbrownell@users.sourceforge.net>
      636fa27e
    • David Brownell's avatar
      musb_hdrc: host vbus and connect/disconnect fixes · 41416eec
      David Brownell authored 
      Teach host side VBUS powerup and connect logic about the problem whereby
the HDRC core wrongly reports VBUS_ERROR during initial powerup.

Remove previous non-working workaround for tusb powering VBUS:  don't
try handling CPEN by hand.

Rest of the fix for disconnection issues in A_SUSPEND.
Signed-off-by: default avatarDavid Brownell <dbrownell@users.sourceforge.net>
      41416eec
    • David Brownell's avatar
      musb_hdrc: small fixes, mostly suspend/resume · b659d87e
      David Brownell authored 
      A collection of mostly orthogonal and small fixes:

 - Update various corner cases in the side suspend/resume code,
   triggered mostly on the host side:
     * avoid spurious gadget driver resume callbacks in B_IDLE state
     * avoid spurious gadget driver suspend callbacks in A_HOST state
     * cope better with some spurious resume irqs
     * partial fix for disconnection issues in A_SUSPEND

 - Resolve some TUSB rmmod races by disabling IRQs at the proper moment,
   and killing the idle timer.

 - Only mark TUSB device as wakeup-capable if it's connected to an
   IRQ which allows that.

 - Don't idle the TUSB chip until khubd handles all pending port
   status change events; dump that status in procfs.
Signed-off-by: default avatarDavid Brownell <dbrownell@users.sourceforge.net>
      b659d87e
    • David Brownell's avatar
      musb_hdrc: remove broken Mentor OTG code · 8e84c387
      David Brownell authored 
      Remove most of the strange OTG-incompatible logic from Mentor, with
its CamelCasing and so forth.  It's better to trigger the transitions
explicitly with relevant operations, like suspend triggering HNP, than
trying to track a subset of the OTG triggers.

Plus some minor cleanups related to previous patches:  better fix
for the "use_dma=n" related compiler warning, pt_regs for omap2430.
Signed-off-by: default avatarDavid Brownell <dbrownell@users.sourceforge.net>
      8e84c387
  7. 01 Nov, 2006 16 commits