- 15 May, 2020 5 commits
-
-
Douglas Anderson authored
It has been postulated that the pm_lock is bad for performance because a CPU currently running rpmh_flush() could block other CPUs from coming out of idle. Similarly CPUs coming out of / going into idle all need to contend with each other for the spinlock just to update the variable tracking who's in PM. Let's optimize this a bit. Specifically: - Use a count rather than a bitmask. This is faster to access and also means we can use the atomic_inc_return() function to really detect who the last one to enter PM was. - Accept that it's OK if we race and are doing the flush (because we think we're last) while another CPU is coming out of idle. As long as we block that CPU if/when it tries to do an active-only transfer we're OK. Signed-off-by:
Douglas Anderson <dianders@chromium.org> Reviewed-by:
Stephen Boyd <swboyd@chromium.org> Link: https://lore.kernel.org/r/20200504104917.v6.5.I295cb72bc5334a2af80313cbe97cb5c9dcb1442c@changeidSigned-off-by:
Bjorn Andersson <bjorn.andersson@linaro.org>
-
Douglas Anderson authored
The rpmh-rsc code had both a driver-level lock (sometimes referred to in comments as drv->lock) and a lock per-TCS. The idea was supposed to be that there would be times where you could get by with just locking a TCS lock and therefor other RPMH users wouldn't be blocked. The above didn't work out so well. Looking at tcs_write() the bigger drv->lock was held for most of the function anyway. Only the __tcs_buffer_write() and __tcs_set_trigger() calls were called without holding the drv->lock. It actually turns out that in tcs_write() we don't need to hold the drv->lock for those function calls anyway even if the per-TCS lock isn't there anymore. From the newly added comments in the code, this is because: - We marked "tcs_in_use" under lock. - Once "tcs_in_use" has been marked nobody else could be writing to these registers until the interrupt goes off. - The interrupt can't go off until we trigger w/ the last line of __tcs_set_trigger(). Thus, from a tcs_write() point of view, the per-TCS lock was useless. Looking at rpmh_rsc_write_ctrl_data(), only the per-TCS lock was held. It turns out, though, that this function already needs to be called with the equivalent of the drv->lock held anyway (we either need to hold drv->lock as we will in a future patch or we need to know no other CPUs could be running as happens today). Specifically rpmh_rsc_write_ctrl_data() might be writing to a TCS that has been borrowed for writing an active transation but it never checks this. Let's eliminate this extra overhead and avoid possible AB BA locking headaches. Suggested-by:
Maulik Shah <mkshah@codeaurora.org> Signed-off-by:
-
Douglas Anderson authored
cpu_pm_notify() is basically a wrapper of notifier_call_chain(). notifier_call_chain() doesn't initialize *nr_calls to 0 before it starts incrementing it--presumably it's up to the callers to do this. Unfortunately the callers of cpu_pm_notify() don't init *nr_calls. This potentially means you could get too many or two few calls to CPU_PM_ENTER_FAILED or CPU_CLUSTER_PM_ENTER_FAILED depending on the luck of the stack. Let's fix this. Fixes: ab10023e ("cpu_pm: Add cpu power management notifiers") Cc: stable@vger.kernel.org Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Reviewed-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by:
-
Douglas Anderson authored
When a PM Notifier returns NOTIFY_BAD it doesn't get called with CPU_PM_ENTER_FAILED. It only get called for CPU_PM_ENTER_FAILED if someone else (further down the notifier chain) returns NOTIFY_BAD. Handle this case by taking our CPU out of the list of ones that have entered PM. Without this it's possible we could detect that the last CPU went down (and we would flush) even if some CPU was alive. That's not good since our flushing routines currently assume they're running on the last CPU for mutual exclusion. Fixes: 985427f9 ("soc: qcom: rpmh: Invoke rpmh_flush() for dirty caches") Signed-off-by:
-
Douglas Anderson authored
Our switch statement doesn't have entries for CPU_CLUSTER_PM_ENTER, CPU_CLUSTER_PM_ENTER_FAILED, and CPU_CLUSTER_PM_EXIT and doesn't have a default. This means that we'll try to do a flush in those cases but we won't necessarily be the last CPU down. That's not so ideal since our (lack of) locking assumes we're on the last CPU. Luckily this isn't as big a problem as you'd think since (at least on the SoC I tested) we don't get these notifications except on full system suspend. ...and on full system suspend we get them on the last CPU down. That means that the worst problem we hit is flushing twice. Still, it's good to make it correct. Reviewed-by:
-
- 12 May, 2020 5 commits
-
-
Gustavo A. R. Silva authored
The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member[1][2], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being inadvertently introduced[3] to the codebase from now on. Also, notice that, dynamic memory allocations won't be affected by this change: "Flexible array members have incomplete type, and so the sizeof operator may not be applied. As a quirk of the original implementation of zero-length arrays, sizeof evaluates to zero."[1] sizeof(flexible-array-member) triggers a warning because flexible array members have incomplete type[1]. There are some instances of code in which the sizeof operator is being incorrectly/erroneously applied to zero-length arrays and the result is zero. Such instances may be hiding some bugs. So, this work (flexible-array member conversions) will also help to get completely rid of those sorts of issues. This issue was found with the help of Coccinelle. [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://github.com/KSPP/linux/issues/21 [3] commit 76497732 ("cxgb3/l2t: Fix undefined behaviour") Reviewed-by:Jeffrey Hugo <jhugo@codeaurora.org> Signed-off-by:
Gustavo A. R. Silva <gustavoars@kernel.org> Link: https://lore.kernel.org/r/20200508210805.GA24170@embeddedorSigned-off-by:
-
Douglas Anderson authored
If our data still isn't there after 1 second, shout and give up. Reported-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Douglas Anderson authored
We can make some of the register access functions more readable by factoring out the calculations a little bit. Suggested-by:
-
Vincent Knecht authored
This patch adds missing SoC IDs for MSM8936/39 and their APQ variants. Signed-off-by:
Vincent Knecht <vincent.knecht@mailoo.org> Signed-off-by:
Konrad Dybcio <konradybcio@gmail.com> Link: https://lore.kernel.org/r/20200511212733.214464-1-konradybcio@gmail.comSigned-off-by:
-
Bjorn Andersson authored
Add SM8250 compatible to the qcom_aoss binding and driver. Acked-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
- 21 Apr, 2020 5 commits
-
-
Sibi Sankar authored
The patch fbe639b4: "soc: qcom: Introduce Protection Domain Restart helpers" leads to the following static checker warning: drivers/soc/qcom/pdr_interface.c:158 pdr_register_listener() '(resp.curr_state < (-((~0 >> 1)) - 1)) => (s32min-s32max < s32min)' These are casted to int so they can't be outside of int range. Fixes: fbe639b4 ("soc: qcom: Introduce Protection Domain Restart helpers") Reported-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
Sibi Sankar <sibis@codeaurora.org> Link: https://lore.kernel.org/r/20200415062955.21439-1-sibis@codeaurora.orgSigned-off-by:
-
Douglas Anderson authored
Adding an item into the cache should never be able to make the cache cleaner. Use "|=" rather than "=" to update the dirty flag. Reviewed-by:
Matthias Kaehlcke <mka@chromium.org> Reviewed-by: Maulik Shah <mkshah@codeaurora.org> Thanks, Maulik Reviewed-by:
-
Bjorn Andersson authored
Tested-by:
Vinod Koul <vkoul@kernel.org> Reviewed-by:
-
Christoph Hellwig authored
As far as the device is concerned the dma address is the physical address. There is no need to convert it to a physical address, especially not using dma-direct internals that are not available to drivers and which will interact badly with IOMMUs. Last but not least the commit introducing it claimed to just fix a type issue, but actually changed behavior. Fixes: 6e37ccf7 ("firmware: qcom_scm: Use proper types for dma mappings") Reviewed-by:
Christoph Hellwig <hch@lst.de> Link: https://lore.kernel.org/r/20200414123136.441454-1-hch@lst.deSigned-off-by:
-
Stephan Gerhold authored
Device nodes should be named according to the class of devices they belong to. Change the suggested names of the subnodes to apr-service@<id>, which is already in use in arch/arm64/boot/dts/qcom/sdm845.dtsi. Reviewed-by:
Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Reviewed-by:
Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200415081159.1098-2-stephan@gerhold.netSigned-off-by:
-
- 20 Apr, 2020 4 commits
-
-
Jason Yan authored
The '>' expression itself is bool, no need to convert it to bool again. This fixes the following coccicheck warning: drivers/firmware/qcom_scm.c:946:25-30: WARNING: conversion to bool not needed here Reviewed-by:
Jason Yan <yanaijie@huawei.com> Link: https://lore.kernel.org/r/20200420123516.7888-1-yanaijie@huawei.comSigned-off-by:
-
Stephen Boyd authored
Read the slv_id properly by making sure the 16-bit number is endian swapped from little endian to CPU native before we read it to figure out what to print for the human readable name. Otherwise we may just show that all the elements in the cmd-db are "Unknown" which isn't right. Reviewed-by:
Guenter Roeck <linux@roeck-us.net> Reported-by:
kbuild test robot <lkp@intel.com> Cc: Lina Iyer <ilina@codeaurora.org> Signed-off-by:
-
Stephen Boyd authored
The top few bits aren't relevant to pad out because they're always zero. Let's just print 5 digits instead of 8 so that it's a little shorter and more readable. Reviewed-by:
Lina Iyer <ilina@codeaurora.org> Suggested-by:
-
Stephen Boyd authored
We pass the result of sizeof() here to tell the printk format specifier how many bytes to print. That expects an int though and sizeof() isn't that type. Cast to int to silence this warning: drivers/soc/qcom/cmd-db.c: In function 'cmd_db_debugfs_dump': drivers/soc/qcom/cmd-db.c:281:30: warning: field width specifier '*' expects argument of type 'int', but argument 4 has type 'long unsigned int' [-Wformat=] Reviewed-by:
Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by:
-
- 14 Apr, 2020 21 commits
-
-
John Stultz authored
This patch allow the rpmpd driver to be loaded as a permenent module. Meaning it can be loaded from a module, but then cannot be unloaded. Ideally, it would include a remove hook and related logic, but apparently the genpd code isn't able to track usage and cleaning things up? (See: https://lkml.org/lkml/2019/1/24/38) So making it a permenent module at least improves things slightly over requiring it to be a built in driver. Cc: Todd Kjos <tkjos@google.com> Cc: Saravana Kannan <saravanak@google.com> Cc: Andy Gross <agross@kernel.org> Cc: Bjorn Andersson <bjorn.andersson@linaro.org> Cc: Rajendra Nayak <rnayak@codeaurora.org> Cc: linux-arm-msm@vger.kernel.org Acked-by:
Saravana Kannan <saravanak@google.com> Tested-by:
John Stultz <john.stultz@linaro.org> Link: https://lore.kernel.org/r/20200326224459.105170-2-john.stultz@linaro.orgSigned-off-by:
-
John Stultz authored
This patch allow the rpmhpd driver to be loaded as a permenent module. Meaning it can be loaded from a module, but then cannot be unloaded. Ideally, it would include a remove hook and related logic, but apparently the genpd code isn't able to track usage and cleaning things up? So making it a permenent module at least improves things slightly over requiring it to be a built in driver. Cc: Todd Kjos <tkjos@google.com> Cc: Saravana Kannan <saravanak@google.com> Cc: Andy Gross <agross@kernel.org> Cc: Bjorn Andersson <bjorn.andersson@linaro.org> Cc: Rajendra Nayak <rnayak@codeaurora.org> Cc: linux-arm-msm@vger.kernel.org Acked-by:
-
John Stultz authored
This patch allow the rpmh driver to be loaded as a permenent module. Meaning it can be loaded from a module, but then cannot be unloaded. Ideally, it would include a remove hook and related logic, but the rpmh driver is fairly core to the system, so once its loaded with almost anythign else to get the system to go, the dependencies are not likely to ever also be removed. So making it a permenent module at least improves things slightly over requiring it to be a built in driver. Acked-by:
-
Douglas Anderson authored
The RSC_DRV_IRQ_ENABLE, RSC_DRV_IRQ_STATUS, and RSC_DRV_IRQ_CLEAR registers are not part of TCS 0. Let's not pretend that they are by using read_tcs_reg() and write_tcs_reg() and passing a bogus tcs_id of 0. We could introduce a new wrapper for these registers but it wouldn't buy us much. Let's just read/write directly. Signed-off-by:
-
Douglas Anderson authored
Auditing tcs_invalidate() made me worried. Specifically I saw that it used spin_lock(), not spin_lock_irqsave(). That always worries me unless I can trace for sure that I'm in the interrupt handler or that someone else already disabled interrupts. Looking more at it, there is actually no reason for these locks anyway. Specifically the only reason you'd ever call rpmh_rsc_invalidate() is if you cared that the sleep/wake TCSes were empty. That means that they need to continue to be empty even after rpmh_rsc_invalidate() returns. The only way that can happen is if the caller already has done something to keep all other RPMH users out. It should be noted that even though the caller is only worried about making sleep/wake TCSes empty, they also need to worry about stopping active-only transfers if they need to handle the case where active-only transfers might borrow the wake TCS. At the moment rpmh_rsc_invalidate() is only called in PM code from the last CPU. If that later changes the caller will still need to solve the above problems themselves, so these locks will never be useful. Continuing to audit tcs_invalidate(), I found a bug. The function didn't properly check for a borrowed TCS if we hadn't recently written anything into the TCS. Specifically, if we've never written to the WAKE_TCS (or we've flushed it recently) then tcs->slots is empty. We'll early-out and we'll never call tcs_is_free(). I thought about fixing this bug by either deleting the early check for bitmap_empty() or possibly only doing it if we knew we weren't on a TCS that could be borrowed. However, I think it's better to just delete the checks. As argued above it's up to the caller to make sure that all other users of RPMH are quiet before tcs_invalidate() is called. Since callers need to handle the zero-active-TCS case anyway that means they need to make sure that the active-only transfers are quiet before calling too. The one way tcs_invalidate() gets called today is through rpmh_rsc_cpu_pm_callback() which calls rpmh_rsc_ctrlr_is_busy() to handle this. When we have another path to get to tcs_invalidate() it will also need to come up with something similar and it won't need this extra check either. If we later find some code path that actually needs this check back in (and somehow manages to be race free) we can always add it back in. Signed-off-by:
-
Douglas Anderson authored
The calls rpmh_rsc_write_ctrl_data() and rpmh_rsc_send_data() are only ever called from rpmh.c. We know that rpmh.c already error checked the message. There's no reason to do it again in rpmh-rsc. Suggested-by:
-
Douglas Anderson authored
tcs_is_free() had two checks in it: does the software think that the TCS is free and does the hardware think that the TCS is free. I couldn't figure out in which case the hardware could think that a TCS was in-use but software thought it was free. Apparently there is no case and the extra check can be removed. This apparently has already been done in a downstream patch. Suggested-by:
-
Douglas Anderson authored
I've been pouring through the rpmh-rsc code and trying to understand it. Document everything to the best of my ability. All documentation here is strictly from code analysis--no actual knowledge of the hardware was used. If something is wrong in here I either misunderstood the code, had a typo, or the code has a bug in it leading to my incorrect understanding. In a few places here I have documented things that don't make tons of sense. A future patch will try to address this. While this means I'm adding comments / todos and then later fixing them in the series, it seemed more urgent to get things documented first so that people could understand the later patches. Any comments I adjusted I also tried to make match kernel-doc better. Specifically: - kernel-doc says do not leave a blank line between the function description and the arguments - kernel-doc examples always have things starting w/ a capital and ending with a period. This should be a no-op. It's just comment changes. Signed-off-by:
-
Douglas Anderson authored
The "cmd_cache" in RPMH wasn't terribly sensible. Specifically: - The current code doesn't really detect "conflicts" properly any case where the sequence being checked has more than one entry. One simple way to see this in the current code is that if cmd[0].addr isn't found then cmd[1].addr is never checked. - The code attempted to use the "cmd_cache" to update an existing message in a sleep/wake TCS with new data. The goal appeared to be to update part of a TCS while leaving the rest of the TCS alone. We never actually do this. We always fully invalidate and re-write everything. - If/when we try to optimize things to not fully invalidate / re-write every time we update the TCSes we'll need to think it through very carefully. Specifically requirement of find_match() that the new sequence of addrs must match exactly the old sequence of addrs seems inflexible. It's also not documented in rpmh_write() and rpmh_write_batch(). In any case, if we do decide to require updates to keep the exact same sequence and length then presumably the API and data structures should be updated to understand groups more properly. The current algorithm doesn't really keep track of the length of the old sequence and there are several boundary-condition bugs because of that. Said another way: if we decide to do something like this in the future we should start from scratch and thus find_match() isn't useful to keep around. This patch isn't quite a no-op. Specifically: - It should be a slight performance boost of not searching through so many arrays. - The old code would have done something useful in one case: it would allow someone calling rpmh_write() to override the data that came from rpmh_write_batch(). I don't believe that actually happens in reality. Signed-off-by:
-
Douglas Anderson authored
The get_tcs_of_type() function doesn't provide any value. It's not conceptually difficult to access a value in an array, even if that value is in a structure and we want a pointer to the value. Having the function in there makes me feel like it's doing something fancier like looping or searching. Remove it. Signed-off-by:
-
Douglas Anderson authored
I was trying to write documentation for the functions in rpmh-rsc and I got to tcs_ctrl_write(). The documentation for the function would have been: "This is the core of rpmh_rsc_write_ctrl_data(); all the caller does is error-check and then call this". Having the error checks in a separate function doesn't help for anything since: - There are no other callers that need to bypass the error checks. - It's less documenting. When I read tcs_ctrl_write() I kept wondering if I need to handle cases other than ACTIVE_ONLY or cases with more commands than could fit in a TCS. This is obvious when the error checks and code are together. - The function just isn't that long, so there's no problem understanding the combined function. Things were even more confusing because the two functions names didn't make obvious (at least to me) their relationship. Simplify by folding one function into the other. Signed-off-by:
-
Douglas Anderson authored
Perhaps it's just me, it took a really long time to understand what the register layout of rpmh-rsc was just from the #defines. Let's add a bunch of comments describing which blocks are part of other blocks. Signed-off-by:
-
Douglas Anderson authored
This patch makes two changes, both of which should be no-ops: 1. Make read_tcs_reg() / read_tcs_cmd() symmetric to write_tcs_reg() / write_tcs_cmd(). 2. Change the order of operations in the above functions to make it more obvious to me what the math is doing. Specifically first you want to find the right TCS, then the right register, and then multiply by the command ID if necessary. Signed-off-by:
-
Maulik Shah authored
When there are more than one WAKE TCS available and there is no dedicated ACTIVE TCS available, invalidating all WAKE TCSes and waiting for current transfer to complete in first WAKE TCS blocks using another free WAKE TCS to complete current request. Remove rpmh_rsc_invalidate() to happen from tcs_write() when WAKE TCSes is re-purposed to be used for Active mode. Clear only currently used WAKE TCS's register configuration. Fixes: 2de4b8d3 (drivers: qcom: rpmh-rsc: allow active requests from wake TCS) Signed-off-by:
-
Raju P.L.S.S.S.N authored
For RSCs that have sleep & wake TCS but no dedicated active TCS, wake TCS can be re-purposed to send active requests. Once the active requests are sent and response is received, the active mode configuration needs to be cleared so that controller can use wake TCS for sending wake requests. Introduce enable_tcs_irq() to enable completion IRQ for repurposed TCSes. Fixes: 2de4b8d3 (drivers: qcom: rpmh-rsc: allow active requests from wake TCS) Signed-off-by:
Raju P.L.S.S.S.N <rplsssn@codeaurora.org> [mkshah: call enable_tcs_irq() within drv->lock, update commit message] Signed-off-by:
-
Maulik Shah authored
Add changes to invoke rpmh flush() from CPU PM notification. This is done when the last the cpu is entering deep CPU idle states and controller is not busy. Controllers that have 'HW solver' mode like display RSC do not need to register for CPU PM notification. They may be in autonomous mode executing low power mode and do not require rpmh_flush() to happen from CPU PM notification. Signed-off-by:
-
Maulik Shah authored
TCSes have previously programmed data when rpmh_flush() is called. This can cause old data to trigger along with newly flushed. Fix this by cleaning SLEEP and WAKE TCSes before new data is flushed. With this there is no need to invoke rpmh_rsc_invalidate() call from rpmh_invalidate(). Simplify rpmh_invalidate() by moving invalidate_batch() inside. Fixes: 600513df ("drivers: qcom: rpmh: cache sleep/wake state requests") Signed-off-by:
-
Maulik Shah authored
Currently rpmh ctrlr dirty flag is set for all cases regardless of data is really changed or not. Add changes to update dirty flag when data is changed to newer values. Update dirty flag everytime when data in batch cache is updated since rpmh_flush() may get invoked from any CPU instead of only last CPU going to low power mode. Also move dirty flag updates to happen from within cache_lock and remove unnecessary INIT_LIST_HEAD() call and a default case from switch. Fixes: 600513df ("drivers: qcom: rpmh: cache sleep/wake state requests") Signed-off-by:
Srinivas Rao L <lsrao@codeaurora.org> Reviewed-by:
Evan Green <evgreen@chromium.org> Reviewed-by:
-
Markus Elfring authored
The function platform_get_irq() can log an error already. Thus omit a redundant message for the exception handling in the calling function. This issue was detected by using the Coccinelle software. Signed-off-by:
Markus Elfring <elfring@users.sourceforge.net> Link: https://lore.kernel.org/r/eb92fcfb-6181-1f9d-2601-61e5231bd892@web.deSigned-off-by:
-
Stephen Boyd authored
It's useful for kernel devs to understand what resources and data is stored inside command db. Add a file in debugufs called 'cmd-db' to dump the memory contents and strings for resources along with their addresses. E.g. Command DB DUMP Slave ARC (v16.0) ------------------------- 0x00030000: cx.lvl [00 00 10 00 40 00 80 00 c0 00 00 01 80 01 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00] 0x00030004: cx.tmr 0x00030010: mx.lvl [00 00 10 00 00 01 80 01 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00] 0x00030014: mx.tmr Cc: Lina Iyer <ilina@codeaurora.org> Cc: Maulik Shah <mkshah@codeaurora.org> Signed-off-by:
-
Srinivas Kandagatla authored
Looks like SoC ID is not exported to sysfs for some reason. This patch adds it! This is mostly used by userspace libraries like Snapdragon Neural Processing Engine (SNPE) SDK for checking supported SoC info. Fixes: efb448d0 ("soc: qcom: Add socinfo driver") Reviewed-by:
-
