Provide sdhci host ops of enhanced strobe to notify
sdhci-msm on enabling/disabling cmdq. This is needed
because of following:
Before running CMDQ transfers in HS400 Enhanced Strobe mode,
SW should write 3 to
HC_VENDOR_SPECIFIC_FUNC3.CMDEN_HS400_INPUT_MASK_CNT register.
Default reset value of this register is 2.
Change-Id: I987605cd21f137dea49ddf3e8db3f1f41b5b501f
Signed-off-by: Ritesh Harjani <riteshh@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Provide cmdq_host ops of enhanced strobe to notify
sdhci on enabling/disabling cmdq. This is needed
because of following:
Before running CMDQ transfers in HS400 Enhanced Strobe mode,
SW should write 3 to
HC_VENDOR_SPECIFIC_FUNC3.CMDEN_HS400_INPUT_MASK_CNT register.
Default reset value of this register is 2.
Change-Id: I36ead91ca8c9aeed967f120f8bdc3d2180af7746
Signed-off-by: Ritesh Harjani <riteshh@codeaurora.org>
After halting CQE, if other commands are sent in legacy mode,
the timeout would be modified as per the requirements of the
command. Upon unhalting, this timeout would still persist for
CQE too, which in some cases may lead to timeout.
Hence, change the timeout to 0xf i.e. max during unhalt.
Change-Id: Ifb0a1f4b9508c5884d381401216ae6c1373bb7de
Signed-off-by: Asutosh Das <asutoshd@codeaurora.org>
This change adds usage for runtime PM for devfreq
context.
In case clock scaling is done from devfreq context its
best to vote for runtime PM to avoid race between
scaling and PM.
Change-Id: I6fc7e2236b44b1f3a322da36552a7f9237c4dac6
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
This change removes the error value (-EAGAIN) that
is returned in case of invalid state for clock scaling.
Invalid state for clock scaling doesn't mean its an error,
it merely means that at the current time we won't scale.
Devfreq will invoke scaling in next interval.
Change-Id: Idbd785da83e6ed00ede2b1b09529f7c81714ccf8
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
This change disables clock scaling in PM notification
for "prepare for system suspend".
This is needed because devfreq creates a dependency
between it and mmc which causes an issue for system suspend.
In this change we break this dependency in earlier stage.
Change-Id: I86dad94c77607b4e8f8fa67035323716f5eb197d
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
If eMMC is not a primary bootdevice, there isn't any point of probing
eMMC device hence disable the probing in such case.
Change-Id: I92fa8c2ef373fd8a9140dbfb41356684aaa28e4e
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Treat bus speed switch failures as error and don't fall back on
default speed. The card status could indicate that the current
function is busy and cannot be switched. Attempt to reinit the
bus speed switch again instead of falling back to default speed.
Silently falling back to lower speed could cause performance issues.
CRs-Fixed: 849567
Change-Id: I4fcd51b82e41746620a68a0a0eb5a18d630ccbea
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
This reverts commit 9cc95fde7c2c ("mmc: sdhci-msm: enable
clock scaling capability").
Clock scaling is breaking XO shutdown, keep the clock scaling
disabled till that issue is addressed.
Change-Id: Id4c530f8e2a8f565f9b957f46e5086078f808c96
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
As per eMMC specification, the PON (Power Off Notification)
must be sent by host to the card before turning off the power.
This will allow card to prepare itself for the power off and
may even reduce the initialization of eMMC upon next boot-up.
Send long PON during system power off and send short PON during
system reboot to reduce the reboot latency.
Change-Id: If4188b8b80aaa0e6c4e00e1807aa9589d5e7efdb
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: Krishna Konda <kkonda@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Add support for handling both manual and auto
bkops when command queuing is running.
Change-Id: Ib967ca3c0420f4e54b3e93c497eb538d7347199a
Signed-off-by: Dov Levenglick <dovl@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Following eMMC JEDEC JESD84-B51 standard, an ehannced form of
rpmb is supported. What this enhanced mode supports is in
addition to be able to write one rpmb or two rpmb frames at a
time, 32 frames can be written at a time.
Expose this information present in ext csd field so that the
user space application that wants to make use of this can do
so.
Change-Id: I53fd962fd7e04b5d2d7804c289d7865c2c5618d5
Signed-off-by: Krishna Konda <kkonda@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Timeout handler executed on softirq context. cmdq dumpstate() api resumes
platform device and it can't be executed on softirq context. Request
completion callback schedules error handler work in case of timeout error.
This change moves CQE registers dump to the error handler callback.
Change-Id: Iea26ca5240f6031218dcf374cafcf2708df1f125
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
Change ID I1333e1d4330393e446ba48a9474c83295744c050
added a call to the bus specific runtime_idle
callback. This callback is only implemented for mmc
thereby causing a crash when other types of memory
(e.g. sd card) are used.
This patch only calls the callback if it is implemented.
Change-Id: Ie26b1b52fdcc0f4d5e3638801e28e965fae0cbeb
Signed-off-by: Dov Levenglick <dovl@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
It is needed to initialize auto bkops before enabling
command queueing since the initialization uses CMD6 which
may not be called when cmdq is enabled.
Change-Id: Idecff761754fe394c01425e7fd4dd5f6a083135d
Signed-off-by: Dov Levenglick <dovl@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Qualcomm SDCC doesn't need NOP descriptor for End of
transfer (EOT).
This change insure that no NOP descriptor will be sent
in order to notify EOT, instead, the EOT will be marked for
the last descriptor.
Change-Id: Ia670bc901af2196a415c2aa54357704949a7b112
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
During halted CMDQ mode, HW expects descriptors sizes
same as used in CMDQ mode.
Our SW uses the legacy path (SDHCI) in case CMDQ is halted,
thus, we need to update legacy path of the fact that we
are in Halted CMDQ state so it can adjust the descriptors
size accordingly.
This change adds a notification mechanism for that purpose.
Change-Id: Iabb473696fb11827dfcce9b137c26b2c5a5a879d
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts & compilation
errors]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
This change adds clock scaling ability to command-queueing
mode, it does so by adding next logic:
* Statistics collection for CMDQ data path
* Empty the queue and Halt it before frequency change
* Scale from data path in case host claiming is not possible
Change-Id: I53a323b55df4d7c27e3ee3426ee4e856e533522c
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
The BKOPS statistics are used for BKOPS unit tests and APT tests
to determine test success or failure.
The BKOPS statstics provide the following information:
The number of times BKOPS were issued according to its severity level
The number of times manual BKOPS was started/stopped (HPI)
The number of times auto BKOPS was enabled/disabled
In order to enable and reset the statistics:
echo 1 > /sys/kernel/debug/mmc0/mmc0:0001/bkops_stats
In order to disable the statistics:
echo 0 > /sys/kernel/debug/mmc0/mmc0:0001/bkops_stats
In order to view the statistics:
cat /sys/kernel/debug/mmc0/mmc0:0001/bkops_stats
Change-Id: Ib84319aedfb49dc022bc27efbda842a5db38c7e9
Signed-off-by: Yaniv Gardi <ygardi@codeaurora.org>
Signed-off-by: Dov Levenglick <dovl@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
During system suspend CQE halted by queue suspend even before mmc host
suspend. When mmc host suspend calls halt helper, CQE is already halted.
This change ignores error value of halt helper.
Change-Id: I2d1456333f1d04f0e7a77485443b80a0ccaa3b4e
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
Compatible list should be terminated with NULL entry, fix it.
Change-Id: Id9be4fa95fd1cbbbe28a3c63b63bbd6c8c29fa85
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
Add support for manual and auto bkops for
legacy (not command-queue) mode.
Change-Id: I1333e1d4330393e446ba48a9474c83295744c050
Signed-off-by: Dov Levenglick <dovl@codeaurora.org>
[subhashj@codeaurora.org: fixed merge conflicts and compilation
errors]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
This commit changes the spinlock used by clock scaling
logic to use the bh flavor instead of irqsave.
This is done in order to avoid the unnecessary irq
disable while going into clock scaling critical sections.
Change-Id: Id660db89dc531336621f40908fd3a4ad4777a12d
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
Much of the sequence to scale up to HS400 is not required when
enhanced strobe mode for HS400 is supported. Handle scale up
to HS400 appropriately in this case.
Change-Id: I733f98bdb99e54d6cd3a074fcece2c89aaaee12f
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
The 8-bit bus width needs to be set first before switching to
DDR bus width when entering HS400 in enhanced strobe mode.
Also use the mmc_select_bus_width() for doing this instead of
rewriting portion of that code.
Change-Id: If9bec799de77714d7183c812a0ba04a9a4ac48f5
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Some hosts may need additional steps to get HS400 functional in
enhanced strobe mode. Add host ops to facilitate that.
Change-Id: I9663830e7ccedf8bf7970d0724a4c7ce212073fd
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Enhance Strobe is defined in v5.1 eMMC spec. This commit
is to implement it.
Normal Strobe signal for HS400 is only provided during
Data Out and CRC Response. While Enhance Strobe is enabled,
Strobe signal is provided during Data Out, CRC Response and
CMD Response.
While enabling Enhance Strobe, the initialization of HS400
does not need enabling HS200 and executing tuning anymore.
This simplifies the HS400 initialization process much.
Per spec, there is a STROBE_SUPPORT added in EXT_CSD register
to indicate that card supports Enhance Strobe or not. If it is
supported, host can enable this feature by enabling the most
significant bit of BUS_WIDTH before set HS_TIMING to HS400.
Change-Id: I1003352cb31dfaec01fde352da7b879a13c94b3f
Signed-off-by: Yi Sun <yi.y.sun@intel.com>
[venkatg@codeaurora.org: Fix minor conflicts & compilation failure]
Patch-mainline: linux-mmc @ 06/04/15, 19:50
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
The vendor specific func register doesn't get reset when using the
software reset register. The various bootloader's could leave this
in an unknown state, hence reset this register to it's power on reset
value during probe.
Change-Id: I3258eedf87b1bc945f43b85627948e6c92b74686
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
mmc_host_clk_hold/release pair should be added for relevant
cmdq_ops. Since it enables host->clock which is needed
for register access as well (apart from
MCLK).
Change-Id: I6d9d15a421225c5b4179cb19e467a17d01ad176f
Signed-off-by: Ritesh Harjani <riteshh@codeaurora.org>
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
Tracepoint measures time that takes runtime suspend, suspending crypto
engine time included into measurement.
Change-Id: I6108a9dc5b188e2086aa5e6d2fe87414bb2a2539
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
Since error handling could race with runtime suspend, increase usage_count
for the card device will prevent this race.
Change-Id: Ie95a3c631f519c7993b0032f0b674871b64e4cb6
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
There are three cases, request should not be pulled from the queue:
- it is dcmd request, while dcmd request is in progress
- the CQE is halted, but not because of the runtime suspend
- the cmdq is in error state
When the card is suspended, the CQE is halted. New request coming should
be pulled from the request queue and the runtime resume will be triggered
by mmc_get_card(). There is no race between the pulling request condition
and the runtime suspend flow, because the card marked as suspended after
the CQE is halted.
Change-Id: I126ae689f7fea2e7545dfda7c4c6abda286a0f11
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
After halting the CQE, an unexpected legacy irq may be received by the
controller. The post halt helper function acknowledges this irq and
handles it.
Change-Id: Iaaa4e57fba830d626fad693ff33dd966994bc50f
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
To execute legacy SDHCi suspend/resume flow, CQE should be halted, because
suspend flow uses legacy SDHCi commands (cmd6 flush and cmd5 sleep)
commands.
On resume, there is full card initialization and both the card and the
controller will be in CQ mode after resume. Still need to update CQE state
by cleaning halt bit.
This change implemented in helpers, that are used by both system and
runtime suspend/resume callbacks.
Change-Id: I2ae9b19e2cbde04366e7ecf06377a5efd81e3f26
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Access to the registers of CQE HCI wrapped by increment/decrement of pm
core usage counter for the platform device.
Change-Id: I9da4aa7d28dbf8e1d2bf62f6d5fa0875bd5b6064
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
After the CQE halt on some controllers need additional actions, before
wakeup of waiting thread.
Change-Id: I742baa48eb3f6eccf782d77a03aafe16ab7188f1
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
When enabling cmdq mode for the first time, need to allocate descriptors
table. After controller reset and enabling cmdq next time, still need to
update CQTDLBA and CQTDLBAU registers with already allocated descriptors
table (desc_dma_base).
Change-Id: Ide89eebbce5a193cd44a1ea4ec65403f98e2a7ab
Signed-off-by: Konstantin Dorfman <kdorfman@codeaurora.org>
eMMC cache barrier provides a way to perform delayed
and in-order flushing of cached data. Host can use this
to avoid flushing delays but still maintain the order
between the data requests in cache.
In case the device gets more barrier requests than it
supports, then a barrier request is treated as a normal flush
request in the device.
If the eMMC cache flushing policy is set to 1, then the device
inherently flushes all the cached requests in FIFO order. For such
devices, as per spec, it is redundant to send any barrier requests
to the device. This may add unnecessary overhead to both host and
the device. So make sure to not send barrier requests in such cases.
Change-Id: Ia7af316800a6895942d3cabcd64600d56fab25a6
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
Enable clock gating for cmdq by adding respective
API to clk hold/release. CMDQ also uses the same
legacy clock gating API.
Change-Id: I3d4f28cedb24ff2292ab08bdd7470358cf134dd5
Signed-off-by: Ritesh Harjani <riteshh@codeaurora.org>
MMC_SDHCI_MSM_ICE is enabled only if MMC_SDHCI_MSM config
is enabled. Reorder this so that it shows up appropriately
in the kernel config menu.
Change-Id: Ic6af2979e7c8c56841d509f3f25fadd4dd2fe9d2
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
The platform host has been initialized and should be freed
in case we defer the probe.
Change-Id: Iadce572fa4618b18f2fadf9e00812e75144af1c9
Signed-off-by: Venkat Gopalakrishnan <venkatg@codeaurora.org>
CMDQ is not supported for RPMB partition. Hence, for RPMB requests
the controller is kept in HALT state and then CMDQ is disabled in
the card.
Change-Id: I1242841d5fa063b542e35dcff95694ef5e88737a
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
Discard is supported in CMDQ mode only when device queue is empty.
Hence, discard commands should be sent using DCMD slot with
QBR (Queue Barrier) flag set.
Change-Id: I630091cbd94ffcdcec71626257f912c15fd2e21e
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
[subhashj@codeaurora.org: fixed trivial merge conflicts]
Signed-off-by: Subhash Jadavani <subhashj@codeaurora.org>
This will be needed by core layers to check the status for
CMD13 that is sent in CMDQ mode.
Change-Id: If3d062bad4cf87c2543e6d5345f9ab6a0afa23bf
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
This change adds UHS cards to mmc_sd_get_max_clock() API.
Cards that support UHS can set timing of SDR104 which
supports frequency up to 208Mhz.
Change-Id: I25bcb35aa2cecd98f6f04cd98a616a76c75b6784
Signed-off-by: Talel Shenhar <tatias@codeaurora.org>
Add changes to configure ICE for data encryption/decryption
using CMDQ.
Change-Id: I9a10d18d617102916526e994e9f9d22d2aa5446b
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>
