fix for a longstanding watchdog timer integration problem.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)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=gBFx
-----END PGP SIGNATURE-----
Merge tag 'omap-devel-c-for-3.5' of git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending into devel-hwmod-data
Some OMAP IP block data additions for 3.5, along with a
fix for a longstanding watchdog timer integration problem.
Without runtime PM enabled, hwmod needs to leave all IP blocks in an
enabled state by default so any driver access to the HW will succeed.
This is accomplished by seting the postsetup_state to enabled for all
hwmods during init when runtime PM is disabled.
Currently, we have a special case for WDT in that its postsetup_state
is always set to disabled. This is done so that the WDT is disabled
and the timer is disarmed at boot in case there is no WDT driver.
This also means that when runtime PM is disabled, if a WDT driver *is*
built in the kernel, the kernel will crash on the first access to the
WDT hardware.
We can't simply leave the WDT module enabled, because the timer is
armed by default after reset. That means that if there is no WDT
driver initialzed or loaded before the timer expires, the kernel will
reboot.
To fix this, a custom reset method is added to the watchdog class of
omap_hwmod. This method will *always* disarm the timer after hwmod
reset. The WDT timer then will only be rearmed when/if the driver is
loaded for the WDT. With the timer disarmed by default, we no longer
need a special-case for the postsetup_state of WDT during init, so it
is removed.
Any platforms wishing to ensure the watchdog remains armed across the
entire boot boot can simply disable the reset-on-init feature of the
watchdog hwmod using omap_hwmod_no_setup_reset().
Tested on 3530/Overo, 4430/Panda.
NOTE: on 4430, the hwmod OCP reset does not seem to rearm the timer as
documented in the TRM (and what happens on OMAP3.) I noticed this
because testing the HWMOD_INIT_NO_RESET feature with no driver loaded,
I expected a reboot part way through the boot, but did not see a
reboot. Adding some debug to read the counter, I verified that right
after OCP softreset, the counter is not firing. After writing the
magic start sequence, the timer starts counting. This means that the
timer disarm sequence added here does not seem to be needed for 4430,
but is technically the correct way to ensure the timer is disarmed, so
it is left in for OMAP4.
Special thanks to Paul Walmsley for helping brainstorm ideas to fix
this problem.
Cc: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>
Cc: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
[paul@pwsan.com: updated the omap2_wd_timer_reset() function in the
wake of commit 3c55c1baff ("ARM:
OMAP2+: hwmod: Revert "ARM: OMAP2+: hwmod: Make omap_hwmod_softreset
wait for reset status""); added kerneldoc; rolled in warning fix from Kevin]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Use 'common' as name for the common irq number in hwmod data for the McBSP
ports. The same name already in use for OMAP2430, and OMAP3.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Restore of context is not done for OMAP4. This patch
adds the OMAP_I2C_FLAG_RESET_REGS_POSTIDLE in the OMAP4
hwmod data which activates the restore for OMAP4.
Currently the OMAP4 does not hit device off still the
driver may have support for it.
Cc: Benoit Cousson <b-cousson@ti.com>
Cc: Paul Wamsley <paul@pwsan.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Use 'common' as name for the common irq number in hwmod data for the McBSP
ports. The same name already in use for OMAP2430, and the OMAP4 hwmod data
will be using the same name.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the HDQ1W hwmod for all OMAP2xxx devices.
Assume that OMAP2xxx chips have the same HDQ idle handling bug
as OMAP3:
http://www.spinics.net/lists/linux-omap/msg63576.html
and set the OCPIF_SWSUP_IDLE flag accordingly on the HDQ's OCP interface.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the HDQ1W hwmod for OMAP34xx, OMAP36xx, and AM3505/3517 devices.
According to the respective TRMs, it doesn't appear to be available for the
816x/814x or the AM335x.
The OCPIF_SWSUP_IDLE flag is added to work around an apparent hardware
bug: the hardware is not taking the CM_FCLKEN*_CORE.EN_HDQ bit into
account when considering whether to go idle:
http://www.spinics.net/lists/linux-omap/msg63576.html
This causes HDQ transfers to fail or become corrupt. Thanks to
NeilBrown for his help diagnosing and testing fixes for this problem.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: NeilBrown <neilb@suse.de>
Tested-by: NeilBrown <neilb@suse.de>
Much of the HDQ1W integration data is common between multiple generations
of OMAP SoCs, so rather than make several copies, we add it once into
files which are compiled for multiple SoCs.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: NeilBrown <neilb@suse.de>
Tested-by: NeilBrown <neilb@suse.de>
Implement a custom reset function for the HDQ1W IP block. This is
because the HDQ1W IP block, like I2C, has an internal clock gating bit
that needs to be toggled after setting the SOFTRESET bit to allow the
reset to propagate.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Avinash.H.M <avinashhm@ti.com>
Tested-by: NeilBrown <neilb@suse.de>
Add MMC for 2420 so we can pass the DMA request lines the same
way as we already do on omap2430 and later.
Cc: Benoit Cousson <b-cousson@ti.com>
Cc: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
[paul@pwsan.com: updated to apply on top of the 3.5 hwmod cleanup;
changed mmc hwmod name/class to "msdi" as documented in the 2420 TRM Rev X;
added sysconfig register information; added 16 bit register width flag;
added MSDI custom reset code]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
powerdomain, PRM, and CM changes.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
iQIcBAABAgAGBQJPqVA5AAoJEMePsQ0LvSpL4QYQALJgM0Qm6i7cmgpECPe+pOOr
WQTxCYZAn8xqZiKBHhveFLpw7qyXmJy1D5QohcBB1Ax5oEY69SLgHC2usUreeJ5L
jP+Za4N5aBWFnU14sacESIK9ahyF8hMS6SxO1iwrjgrMygFcjsqL7lk2B665Wiqn
6OS6kAdAemMoKTCsDFY8R18WiTuGGF5Oo8hEhxs0RzgX9LjG8iDaq19wtvmefo11
U6246BSDh7xrCl5hiiFMQT942oQ5VA3zUQT1ClwPStYGJmwkLwdl3f4yE3vHC0Q/
zNdS2bQM9xSB/GCikWsfbjf8dVO0t3uaqHva0LF/DTQHiMKptm2f+NWAjdZCDlYz
FbE3SRkX86Rr6P+sw69ztpRroKMbn8WNfP48fXbMEtiegpqjRc5wkGr4cIcl8yMg
TyBdF/o5QJgxoF4hFw5BQBrBvpL5+ralqrHA7A4zeCNhkPtnaXSzd1OHgqMawFmk
otEg66W13h+bOnNLmSbWhWyVmYwsD30BjmNLyOTXa16aTjo+X4johuVGlZMKWyD6
nQMdwnxdSmNMQ8z1Qczc4JzT/ixjfeEXmQUNZLmhkMK5cHRw6TYj9eEbvMk7Xh73
XCK38RCb86ECqn1A3qEWo6TV8CDdhhh/XwoHAzCfdkRdqe3d5cT5cC2vy/VmRrf1
8Il2JKJYveKPH+vc9rEE
=Q4kv
-----END PGP SIGNATURE-----
Merge tag 'omap-devel-b-for-3.5' of git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending into devel-prcm
Some OMAP PRCM updates for 3.5. Includes some clock, clockdomain,
powerdomain, PRM, and CM changes.
Clean up clockdomains3xxx_data.c a bit by removing the superfluous
commas in gfx_sgx_3xxx_wkdeps[].
Signed-off-by: Mark A. Greer <mgreer@animalcreek.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
With patch 'ARM: OMAP2+: powerdomain: Wait for powerdomain transition
in pwrdm_state_switch()', the pwrdm_clkdm_state_switch() API becomes
duplicate of pwrdm_state_switch().
Get rid off duplicate pwrdm_clkdm_state_switch() and update the
users of it with pwrdm_state_switch()
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Cc: Rajendra Nayak <rnayak@ti.com>
Cc: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the correct clockdomain for the HDQ functional clock. This is needed
for the clock and hwmod PM code to work correctly.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: NeilBrown <neilb@suse.de>
The current DPLL code enables and disables autoidle features
without checking whether the autoidle register is available.
Fix this by putting a check for the existence of the autoidle
register in the DPLL data.
With such a check in place, for DPLLs which do not support this
feature, simply skipping the autoidle_reg entry in the DPLL data
is sufficient.
Signed-off-by: Vaibhav Bedia <vaibhav.bedia@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
We do not use iclk anywhere in the dmtimer driver and so removing it.
Hence removing the timer iclk entries from OMAP4 clkdev table as well.
Signed-off-by: Tarun Kanti DebBarma <tarun.kanti@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Instead of statically defining seperate arrays for every OMAP4+ archs,
have a generic init function to populate the arrays. This avoids the
need for creating new array for every arch added in the future that
reuses the prm and cm registers read/write code.
Cc: Paul Walmsley <paul@pwsan.com>
Signed-off-by: R Sricharan <r.sricharan@ti.com>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add missing idle_st bit for 32k-sync timer into the prcm-common
header file, required for hwmod data.
Signed-off-by: Vaibhav Hiremath <hvaibhav@ti.com>
Reviewed-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Cc: Felipe Balbi <balbi@ti.com>
Cc: Benoit Cousson <b-cousson@ti.com>
Cc: Tony Lindgren <tony@atomide.com>
Cc: Paul Walmsley <paul@pwsan.com>
Cc: Kevin Hilman <khilman@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The register bits for MPU_CLK_SRC and IVA2_CLK_SRC in CM_CLKSEL1_PLL
register are 3 bits wide. Fix the MASK definition accordingly.
Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
To improve the clarity of the code, replace the CK_3517 flag used in
the clock data with CK_AM35XX. The CK_3505 flag can also be
removed, since it is now unused.
Acked-by: Vaibhav Hiremath <hvaibhav@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The init for 3505/3517 specific clocks depends on the ordering of
cpu_is checks, is error prone and confusing (there are 2 separate
checks for cpu_is_omap3505()).
Remove the 3505-specific checking since CK_3505 flag is not used, and
treat all AM35x clocks the same.
This means that the SGX clock (the only AM35x clkdev not currently
flagged for 3505) will now be registered on 3505, but that is
harmless. That can be cleaned up when the clkdev nodes are removed in
favor of them being registered by hwmod.
Acked-by: Vaibhav Hiremath <hvaibhav@ti.com>
Tested-by: Vaibhav Hiremath <hvaibhav@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The AM35x UART4 is common to all AM35x devices, so use CK_AM35XX instead
of (CK_3505 | CK_3517), which is equivalent.
Acked-by: Vaibhav Hiremath <hvaibhav@ti.com>
Tested-by: Vaibhav Hiremath <hvaibhav@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
OMAP44xx family of SoCs. There still seem to be a few missing, such
as those for the MMU IP blocks, but this seems to cover the bulk of
the remainder.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)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=ToTm
-----END PGP SIGNATURE-----
Merge tag 'omap-devel-a-for-3.5' of git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending into devel-hwmod
Add in most of the remaining hwmods (IP block descriptions) for the
OMAP44xx family of SoCs. There still seem to be a few missing, such
as those for the MMU IP blocks, but this seems to cover the bulk of
the remainder.
control code for OMAP SoCs. In particular, this series results in
a considerable diffstat savings by changing the way that IP block
interconnections are defined.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)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=NqUz
-----END PGP SIGNATURE-----
Merge tag 'omap-cleanup-b-for-3.5' of git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending into cleanup-hwmod
Clean up various aspects of the OMAP hwmod code, which is the IP block
control code for OMAP SoCs. In particular, this series results in
a considerable diffstat savings by changing the way that IP block
interconnections are defined.
Add a skeleton hwmod for the DEBUGSS and associated interconnect data.
This is a basic set of data that will need further additions as
further DEBUGSS information becomes available.
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the PRCM, CM, PRM, and related hwmod and associated interconnect
data. These IP blocks handle most of the on-chip power, reset, and clock
control.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Add the System Control Module hwmod and associated interconnect data.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Add the OCP-WP hwmod and associated interconnect data. The OCP-WP,
or OCP watchpoint, can be used to collect interconnect data and
transmit it via the STM port.
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the OCM RAM IP block and interconnect data. This is an oh-chip
block of SRAM connected directly to the L3 bus.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Add the OCP2SCP IP block and interconnect data. The OCP2SCP can be
used in conjunction with the on-chip embedded USB PHY, associated with
the OTG controller.
Add the on-chip full-speed USB host controller IP block and
interconnect data.
Cc: Felipe Balbi <balbi@ti.com>
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the SL2 interface IP block and interconnect data. The SL2 is related
to the IVA-HD subsystem.
Add IP block and interconnect data for the C2C ("Chip-to-chip")
interconnect. This can provide a direct system interconnect link to
other devices stacked on the OMAP package.
Add the ELM IP block and interconnect data. The ELM can be used
to locate errors in NAND flash connected to the GPMC.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Add the McASP hwmod and associated interconnect data. The McASP is a
general-purpose audio serial port.
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the Slimbus hwmods and associated interconnect data. The Slimbus
IP blocks implement a two-wire serial interface.
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the GPU hwmod and associated interconnect data. The GPU is a
graphics accelerator.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Add the EMIF1 and 2 hwmods and associated interconnect data. The EMIFs
are SDRAM interface IP blocks.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Add the GPMC hwmod and associated interconnect data. The GPMC is a
programmable parallel-bus memory controller.
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Add the HDQ/1-wire hwmod and associated interconnect data. The
HDQ/1-wire IP block is a low-speed serial interconnect.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Benoît Cousson <b-cousson@ti.com>
Add hwmod data for the OMAP4 FDIF IP block.
This patch also includes a change (originally from Fernando Guzman
Lugo <fernando.lugo@ti.com>) to set a softreset delay for the FDIF IP
block:
http://www.spinics.net/lists/arm-kernel/msg161874.html
Signed-off-by: Ming Lei <ming.lei@canonical.com>
Acked-by: Benoît Cousson <b-cousson@ti.com>
Cc: Fernando Guzman Lugo <fernando.lugo@ti.com>
[paul@pwsan.com: rearranged to match script output; fixed FDIF end address to
match script data; wrote trivial changelog; combined the FDIF portion of
Fernando's srst_udelay patch]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The PRM and CM implicit clockdomains will soon be used by OMAP44xx.
So, make them common to OMAP2+ and modify the OMAP4 clockdomains code
so use of these clockdomains doesn't crash the system.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Rajendra Nayak <rnayak@ti.com>
N800 logs this message on boot:
[ 0.182281] omap_hwmod: iva: cannot be enabled for reset (3)
Fix by creating basic IVA1 and DSP hwmods for OMAP2420, and a basic IVA2
hwmod for OMAP2430. There is still more information to be added, but
this should resolve the immediate issue.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The IVA hwmod data is missing some fields that cause the following
warning on boot:
[ 0.118011] omap_hwmod: iva: cannot be enabled for reset (3)
Fix by encoding the IP block's main functional clock, reset lines, and
clockdomain.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
The OMAP3 hwmod data listed iva2_ck as an interface clock between the
IVA and L3. This is incorrect. iva2_ck is not an interface clock.
Since it cannot auto-idle, specifying it here prevents the IVA and at
least one of the CORE clockdomains from going idle, which causes PM
problems such as these upon system suspend:
[ 70.626129] Powerdomain (iva2_pwrdm) didn't enter target state 1
[ 70.626190] Powerdomain (core_pwrdm) didn't enter target state 1
Fix by specifying the actual interface clock in the hwmod data.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Several struct omap_hwmod_ocp_if records can be shared between OMAP2420
and OMAP2430. Move these shared records out of the chip-specific files
into mach-omap2/omap_hwmod_2xxx_interconnect_data.c. This should save some
memory and source lines, at the cost of readability.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Benoît Cousson <b-cousson@ti.com>
After the link registration conversion, it's much easier to share some
hwmod data between OMAP2420 and 2430. Move the shareable data into a
common file. This should save some memory and lines of source, at the
cost of readability.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Benoît Cousson <b-cousson@ti.com>
Reorganize the hwmod data to declare the IP blocks first and the
interconnects second. This allows us to remove the forward
declarations, which this patch also does. Saves some lines of source
data. While here, take the opportunity to synchronize the order of
the OMAP44xx hwmod data with the autogenerator output -- it's slightly
different due to past mismerges -- and fix a few minor typos and
whitespace problems in the files.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Benoît Cousson <b-cousson@ti.com>
Now that the data has been converted to use interface registration, we
can remove the (now unused) direct hwmod registration code.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Benoît Cousson <b-cousson@ti.com>
Register interconnect links between IP blocks, rather than the IP
blocks themselves. (The IP blocks will be registered as a side-effect
of registering the links.)
The objective is to reduce the number of lines of static data and
facilitate the sharing of IP block data between different SoCs. These
objectives come at the penalty of increased boot time due to increased
computation.
While here, fix a few whitespace problems and inaccurate variable names.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Benoît Cousson <b-cousson@ti.com>
Add support for direct IP block interconnect ("link") registration to
the hwmod code via a new function, omap_hwmod_register_links(). This
will replace direct registration of hwmods, and a subsequent patch
will remove omap_hwmod_register().
This change will allow a subsequent patch to remove the hwmod data
link arrays. This will reduce the size of the hwmod static data and
also make it easier to generate the data files. It will also make it
possible to share some of the struct omap_hwmod records across
multiple SoCs, since the link array pointers will be removed from the
struct omap_hwmod.
The downside is that boot time will increase. Minimizing boot time
was the reason why the link arrays were originally introduced.
Removing them will require extra computation during boot to allocate
memory and associate IP blocks with their interconnects. However,
since the current kernel development focus is on reducing the number
of lines in arch/arm/mach-omap2/, boot time impact is now seemingly
considered a lower priority.
This patch contains additional complexity to reduce the number of
memory allocations required for this change. This reduces the boot
time impact: total hwmod link registration time was ~ 2655
microseconds with a simple allocation strategy, but is now ~ 549
microseconds[1] with the approach taken by this patch.
1. Measured on a BeagleBoard 35xx @ 500MHz MPU/333 MHz CORE, average
of 7 samples. Total uncertainty is +/- 61 microseconds.
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Benoît Cousson <b-cousson@ti.com>
