[PATCH v4 1/3] Coresight: Add coresight dummy driver

Hao Zhang quic_hazha at quicinc.com
Tue May 23 23:12:30 PDT 2023



On 5/10/2023 10:56 AM, Hao Zhang wrote:
> Hi Suzuki,
> 
> On 5/5/2023 5:57 PM, Suzuki K Poulose wrote:
>> On 05/05/2023 10:24, Hao Zhang wrote:
>>> Some Coresight devices that kernel don't have permission to access or
>>> configure. So there need driver to register dummy devices as Coresight
>>> devices. It may also be used to define components that may not have
>>> any programming interfaces (e.g, static links), so that paths can be
>>> established in the driver. Provide Coresight API for dummy device
>>> operations, such as enabling and disabling dummy devices. Build the
>>> Coresight path for dummy sink or dummy source for debugging.
>>>
>>> Signed-off-by: Hao Zhang <quic_hazha at quicinc.com>
>>> ---
>>>   drivers/hwtracing/coresight/Kconfig           |  11 ++
>>>   drivers/hwtracing/coresight/Makefile          |   1 +
>>>   drivers/hwtracing/coresight/coresight-dummy.c | 171 ++++++++++++++++++
>>>   include/linux/coresight.h                     |   1 +
>>>   4 files changed, 184 insertions(+)
>>>   create mode 100644 drivers/hwtracing/coresight/coresight-dummy.c
>>>
>>> diff --git a/drivers/hwtracing/coresight/Kconfig 
>>> b/drivers/hwtracing/coresight/Kconfig
>>> index 2b5bbfffbc4f..06f0a7594169 100644
>>> --- a/drivers/hwtracing/coresight/Kconfig
>>> +++ b/drivers/hwtracing/coresight/Kconfig
>>> @@ -236,4 +236,15 @@ config CORESIGHT_TPDA
>>>         To compile this driver as a module, choose M here: the module 
>>> will be
>>>         called coresight-tpda.
>>> +
>>> +config CORESIGHT_DUMMY
>>> +    tristate "Dummy driver support"
>>> +    help
>>> +      Enables support for dummy driver. Dummy driver can be used for
>>> +      CoreSight sources/sinks that are owned and configured by some
>>> +      other subsystem and use Linux drivers to configure rest of trace
>>> +      path > +
>>> +      To compile this driver as a module, choose M here: the module 
>>> will be
>>> +      called coresight-dummy.
>>>   endif
>>> diff --git a/drivers/hwtracing/coresight/Makefile 
>>> b/drivers/hwtracing/coresight/Makefile
>>> index 33bcc3f7b8ae..995d3b2c76df 100644
>>> --- a/drivers/hwtracing/coresight/Makefile
>>> +++ b/drivers/hwtracing/coresight/Makefile
>>> @@ -30,3 +30,4 @@ obj-$(CONFIG_CORESIGHT_TPDA) += coresight-tpda.o
>>>   coresight-cti-y := coresight-cti-core.o    coresight-cti-platform.o \
>>>              coresight-cti-sysfs.o
>>>   obj-$(CONFIG_ULTRASOC_SMB) += ultrasoc-smb.o
>>> +obj-$(CONFIG_CORESIGHT_DUMMY) += coresight-dummy.o
>>> diff --git a/drivers/hwtracing/coresight/coresight-dummy.c 
>>> b/drivers/hwtracing/coresight/coresight-dummy.c
>>> new file mode 100644
>>> index 000000000000..ee9881ff4754
>>> --- /dev/null
>>> +++ b/drivers/hwtracing/coresight/coresight-dummy.c
>>> @@ -0,0 +1,171 @@
>>> +// SPDX-License-Identifier: GPL-2.0
>>> +/*
>>> + * Copyright (c) 2023 Qualcomm Innovation Center, Inc. All rights 
>>> reserved.
>>> + */
>>> +
>>> +#include <linux/kernel.h>
>>> +#include <linux/module.h>
>>> +#include <linux/platform_device.h>
>>> +#include <linux/coresight.h>
>>> +#include <linux/of.h>
>>> +#include <linux/pm_runtime.h>
>>
>> Please follow the alphabetical order for the header files ^
>>
> Sure, I will update this in the next patch series.
> 
>>> +
>>> +#include "coresight-priv.h"
>>> +
>>> +struct dummy_drvdata {
>>> +    struct device            *dev;
>>
>> nit: We don't need this really. And that completely removes the need for
>> drvdata too. See below.
>>
>>> +    struct coresight_device        *csdev;
>>> +};
>>> +
>>> +DEFINE_CORESIGHT_DEVLIST(source_devs, "dummy_source");
>>> +DEFINE_CORESIGHT_DEVLIST(sink_devs, "dummy_sink");
>>> +
>>> +static int dummy_source_enable(struct coresight_device *csdev,
>>> +                   struct perf_event *event, u32 mode)
>>> +{
>>> +    struct dummy_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
>>> +
>>> +    dev_dbg(drvdata->dev, "Dummy source enabled\n");
>>
>>      dev_dbg(csdev->dev.parent, ..");
>>
>> Similarly for all instances below.
>>
>>> +
>>> +    return 0;
>>> +}
>>> +
>>> +static void dummy_source_disable(struct coresight_device *csdev,
>>> +                 struct perf_event *event)
>>> +{
>>> +    struct dummy_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
>>> +
>>> +    dev_dbg(drvdata->dev, "Dummy source disabled\n");
>>> +}
>>> +
>>> +static int dummy_sink_enable(struct coresight_device *csdev, u32 mode,
>>> +                void *data)
>>> +{
>>> +    struct dummy_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
>>> +
>>> +    dev_dbg(drvdata->dev, "Dummy sink enabled\n");
>>> +
>>> +    return 0;
>>> +}
>>> +
>>> +static int dummy_sink_disable(struct coresight_device *csdev)
>>> +{
>>> +    struct dummy_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
>>> +
>>> +    dev_dbg(drvdata->dev, "Dummy sink disabled\n");
>>> +
>>> +    return 0;
>>> +}
>>> +
>>> +static const struct coresight_ops_source dummy_source_ops = {
>>> +    .enable    = dummy_source_enable,
>>> +    .disable = dummy_source_disable,
>>> +};
>>> +
>>> +static const struct coresight_ops dummy_source_cs_ops = {
>>> +    .source_ops = &dummy_source_ops,
>>> +};
>>> +
>>> +static const struct coresight_ops_sink dummy_sink_ops = {
>>> +    .enable    = dummy_sink_enable,
>>> +    .disable = dummy_sink_disable,
>>> +};
>>> +
>>> +static const struct coresight_ops dummy_sink_cs_ops = {
>>> +    .sink_ops = &dummy_sink_ops,
>>> +};
>>> +
>>> +static int dummy_probe(struct platform_device *pdev)
>>> +{
>>> +    struct device *dev = &pdev->dev;
>>> +    struct device_node *node = dev->of_node;
>>> +    struct coresight_platform_data *pdata;
>>> +    struct dummy_drvdata *drvdata;
>>> +    struct coresight_desc desc = { 0 };
>>> +
>>> +    if (of_device_is_compatible(node, "arm,coresight-dummy-source")) {
>>> +
>>> +        desc.name = coresight_alloc_device_name(&source_devs, dev);
>>> +        if (!desc.name)
>>> +            return -ENOMEM;
>>> +
>>> +        desc.type = CORESIGHT_DEV_TYPE_SOURCE;
>>> +        desc.subtype.source_subtype =
>>> +                    CORESIGHT_DEV_SUBTYPE_SOURCE_OTHERS;
>>> +        desc.ops = &dummy_source_cs_ops;
>>> +    } else if (of_device_is_compatible(node, 
>>> "arm,coresight-dummy-sink")) {
>>> +        desc.name = coresight_alloc_device_name(&sink_devs, dev);
>>> +        if (!desc.name)
>>> +            return -ENOMEM;
>>> +
>>> +        desc.type = CORESIGHT_DEV_TYPE_SINK;
>>> +        desc.subtype.sink_subtype = CORESIGHT_DEV_SUBTYPE_SINK_DUMMY;
>>> +        desc.ops = &dummy_sink_cs_ops;
>>> +    } else {
>>> +        dev_err(dev, "Device type not set\n");
>>> +        return -EINVAL;
>>> +    }
>>> +
>>> +    pdata = coresight_get_platform_data(dev);
>>> +    if (IS_ERR(pdata))
>>> +        return PTR_ERR(pdata);
>>> +    pdev->dev.platform_data = pdata;
>>> +
>>> +    drvdata = devm_kzalloc(dev, sizeof(*drvdata), GFP_KERNEL);
>>> +    if (!drvdata)
>>> +        return -ENOMEM; > +
>>> +    drvdata->dev = &pdev->dev;
>>> +    platform_set_drvdata(pdev, drvdata);
>>
>> As above, you may remove the drvdata entirely.
>>
> For some dummy sources, it's needed to allocate traceid for them in 
> kernel driver, so this struct would be useful for that case.
> 
> I will remove it now and upstream it in the further.
> 

Hi Suzuki,

After checking, I find the drvdata struct is required for csdev which 
would store the pointer of coresight_device, and it is also required for 
traceid in the further. So I will keep drvdata in the next patch series.

Thanks,
Hao

> Thanks,
> Hao
> 
>> Otherwise looks good to me
>>
>> Suzuki
>>



More information about the linux-arm-kernel mailing list