On 2020-01-22 22:18, Evan Green wrote:
On Wed, Jan 22, 2020 at 12:20 AM Georgi Djakov
<georgi.djakov@xxxxxxxxxx> wrote:
On 1/22/20 08:45, Sibi Sankar wrote:
> Hey Evan,
>
> Thanks for the review!
>
> On 2020-01-22 03:03, Evan Green wrote:
>> On Thu, Jan 9, 2020 at 1:12 PM Sibi Sankar <sibis@xxxxxxxxxxxxxx> wrote:
>>>
>>> On some Qualcomm SoCs, Operating State Manager (OSM) controls the
>>> resources of scaling L3 caches. Add a driver to handle bandwidth
>>> requests to OSM L3 from CPU on SDM845 SoCs.
>>>
>>> Signed-off-by: Sibi Sankar <sibis@xxxxxxxxxxxxxx>
>>> ---
>>> drivers/interconnect/qcom/Kconfig | 7 +
>>> drivers/interconnect/qcom/Makefile | 2 +
>>> drivers/interconnect/qcom/osm-l3.c | 267 +++++++++++++++++++++++++++++
>>> 3 files changed, 276 insertions(+)
>>> create mode 100644 drivers/interconnect/qcom/osm-l3.c
>>>
>>> diff --git a/drivers/interconnect/qcom/Kconfig
>>> b/drivers/interconnect/qcom/Kconfig
>>> index a9bbbdf7400f9..b94d28e7bf700 100644
>>> --- a/drivers/interconnect/qcom/Kconfig
>>> +++ b/drivers/interconnect/qcom/Kconfig
>>> @@ -14,6 +14,13 @@ config INTERCONNECT_QCOM_MSM8974
>>> This is a driver for the Qualcomm Network-on-Chip on msm8974-based
>>> platforms.
>>>
>>> +config INTERCONNECT_QCOM_OSM_L3
>>> + tristate "Qualcomm OSM L3 interconnect driver"
>>> + depends on INTERCONNECT_QCOM || COMPILE_TEST
>>> + help
>>> + Say y here to support the Operating State Manager (OSM) interconnect
>>> + driver which controls the scaling of L3 caches on Qualcomm SoCs.
>>> +
>>> config INTERCONNECT_QCOM_QCS404
>>> tristate "Qualcomm QCS404 interconnect driver"
>>> depends on INTERCONNECT_QCOM
>>> diff --git a/drivers/interconnect/qcom/Makefile
>>> b/drivers/interconnect/qcom/Makefile
>>> index 55ec3c5c89dbd..89fecbd1257c7 100644
>>> --- a/drivers/interconnect/qcom/Makefile
>>> +++ b/drivers/interconnect/qcom/Makefile
>>> @@ -1,5 +1,6 @@
>>> # SPDX-License-Identifier: GPL-2.0
>>>
>>> +icc-osm-l3-objs := osm-l3.o
>>> qnoc-msm8974-objs := msm8974.o
>>> qnoc-qcs404-objs := qcs404.o
>>> qnoc-sc7180-objs := sc7180.o
>>> @@ -12,6 +13,7 @@ icc-smd-rpm-objs := smd-rpm.o
>>> obj-$(CONFIG_INTERCONNECT_QCOM_BCM_VOTER) += icc-bcm-voter.o
>>> obj-$(CONFIG_INTERCONNECT_QCOM_MSM8916) += qnoc-msm8916.o
>>> obj-$(CONFIG_INTERCONNECT_QCOM_MSM8974) += qnoc-msm8974.o
>>> +obj-$(CONFIG_INTERCONNECT_QCOM_OSM_L3) += icc-osm-l3.o
>>> obj-$(CONFIG_INTERCONNECT_QCOM_QCS404) += qnoc-qcs404.o
>>> obj-$(CONFIG_INTERCONNECT_QCOM_RPMH) += icc-rpmh.o
>>> obj-$(CONFIG_INTERCONNECT_QCOM_SC7180) += qnoc-sc7180.o
>>> diff --git a/drivers/interconnect/qcom/osm-l3.c
>>> b/drivers/interconnect/qcom/osm-l3.c
>>> new file mode 100644
>>> index 0000000000000..7fde53c70081e
>>> --- /dev/null
>>> +++ b/drivers/interconnect/qcom/osm-l3.c
>>> @@ -0,0 +1,267 @@
>>> +// SPDX-License-Identifier: GPL-2.0
>>> +/*
>>> + * Copyright (c) 2019, The Linux Foundation. All rights reserved.
>>> + *
>>> + */
>>> +
>>> +#include <dt-bindings/interconnect/qcom,osm-l3.h>
>>> +#include <linux/bitfield.h>
>>> +#include <linux/clk.h>
>>> +#include <linux/interconnect-provider.h>
>>> +#include <linux/io.h>
>>> +#include <linux/kernel.h>
>>> +#include <linux/module.h>
>>> +#include <linux/of_device.h>
>>> +#include <linux/of_platform.h>
>>> +#include <linux/platform_device.h>
>>> +
>>> +#define LUT_MAX_ENTRIES 40U
>>> +#define LUT_SRC GENMASK(31, 30)
>>> +#define LUT_L_VAL GENMASK(7, 0)
>>> +#define LUT_ROW_SIZE 32
>>> +#define CLK_HW_DIV 2
>>> +
>>> +/* Register offsets */
>>> +#define REG_ENABLE 0x0
>>> +#define REG_FREQ_LUT 0x110
>>> +#define REG_PERF_STATE 0x920
>>> +
>>> +#define OSM_L3_MAX_LINKS 1
>>> +#define SDM845_MAX_RSC_NODES 130
>>
>> I'm nervous this define is going to fall out of date with
>> qcom,sdm845.h. I'm worried someone will end up adding a few more nodes
>> that were always there but previously hidden from Linux. Can we put
>> this define in include/dt-bindings/interconnect/qcom,sdm845.h, so at
>> least when that happens they'll come face to face with this define?
>> The same comment goes for the SC7180 define in patch 4.
>
> Yeah both solution require manual
> intervention how about we just go
> with what I proposed below.
>
>>
>> On second thought, this trick only works once. Are we sure there
>> aren't going to be other drivers that might want to tag on
>> interconnect nodes as well? How about instead we just add the enum
>> values below in qcom,sdm845.h as defines?
>
> Georgi/Evan,
> Since qcom,sdm845.h is specific to
> bindings shouldn't I just create a
> .h file with all the enums so that
> it can used across all icc providers
> on SDM845?
This sounds good to me, unless Evan has any objections.
So is this a new .h file with all the node numbers from qcom,sdm845.h
and your new couple of nodes here? That would be fine with me.
Or is it a .h file with only your two new node numbers? My worry there
is when there are two or three other drivers like this one, it will be
difficult to follow the total order of nodes as "base provider', "L3
driver", "new driver 1", "new driver 2".... any thoughts on how we
might address that?
the relative provider numbers from
qcom,sdm845.h have no useful meaning
for other icc providers. However the
enum defined in the sdm845.c which are
the node ids are needed and should be
sufficient to add/link to any icc node
across icc providers. So introducing a
sdm845.h with all the enumbs global node
ids is what I am proposing to do.
-Evan
--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project.
