From: James Clark <james.clark@linaro.org>
To: Jie Gan <quic_jiegan@quicinc.com>
Cc: Tingwei Zhang <quic_tingweiz@quicinc.com>,
Jinlong Mao <quic_jinlmao@quicinc.com>,
coresight@lists.linaro.org, linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org, devicetree@vger.kernel.org,
linux-arm-msm@vger.kernel.org,
linux-stm32@st-md-mailman.stormreply.com,
Suzuki K Poulose <suzuki.poulose@arm.com>,
Mike Leach <mike.leach@linaro.org>,
Alexander Shishkin <alexander.shishkin@linux.intel.com>,
Maxime Coquelin <mcoquelin.stm32@gmail.com>,
Alexandre Torgue <alexandre.torgue@foss.st.com>,
Rob Herring <robh@kernel.org>,
Krzysztof Kozlowski <krzk+dt@kernel.org>,
Conor Dooley <conor+dt@kernel.org>,
Bjorn Andersson <andersson@kernel.org>,
Konrad Dybcio <konradybcio@kernel.org>
Subject: Re: [PATCH v9 3/6] Coresight: Introduce a new struct coresight_path
Date: Tue, 28 Jan 2025 11:54:50 +0000 [thread overview]
Message-ID: <538ae543-4c43-4df2-8adc-911096fe14dd@linaro.org> (raw)
In-Reply-To: <20250124072537.1801030-4-quic_jiegan@quicinc.com>
On 24/01/2025 7:25 am, Jie Gan wrote:
> Add 'struct coresight_path' to store the data that is needed by
> coresight_enable_path/coresight_disable_path. The structure
> will be transmitted to the helper and sink device to enable
> related funcationalities.
>
> Signed-off-by: Jie Gan <quic_jiegan@quicinc.com>
> ---
> drivers/hwtracing/coresight/coresight-core.c | 87 ++++++++++++++-----
> drivers/hwtracing/coresight/coresight-etb10.c | 3 +-
> .../hwtracing/coresight/coresight-etm-perf.c | 52 ++++++-----
> .../hwtracing/coresight/coresight-etm-perf.h | 2 +-
> drivers/hwtracing/coresight/coresight-priv.h | 21 +++--
> drivers/hwtracing/coresight/coresight-sysfs.c | 32 +++----
> .../hwtracing/coresight/coresight-tmc-etf.c | 3 +-
> .../hwtracing/coresight/coresight-tmc-etr.c | 6 +-
> drivers/hwtracing/coresight/coresight-trbe.c | 4 +-
> drivers/hwtracing/coresight/ultrasoc-smb.c | 3 +-
> 10 files changed, 137 insertions(+), 76 deletions(-)
>
[...]
> INIT_LIST_HEAD(path);
> + cs_path->path = path;
> + /*
> + * Since not all source devices have a defined trace_id function,
> + * make sure to check for it before use.
> + *
> + * Assert the mode is CS_MODE_SYSFS, the trace_id will be assigned
> + * again later if the mode is CS_MODE_PERF.
> + */
> + if (source_ops(source)->trace_id != NULL) {
> + rc = source_ops(source)->trace_id(source, CS_MODE_SYSFS, NULL);
I don't think we should do this. Doesn't this consume two trace IDs for
each session? And I'm not even sure if it's released properly if it's
overwritten.
It should be possible to consolidate the all the trace ID allocation to
a single step when building the path, or another function that gets
called just after the path is built. At the moment the ID can be
allocated from about 5 different places and it's quite hard to
understand, especially with these new changes. I have some of it coded
up, let me finish it off and I can share it.
> + if(IS_VALID_CS_TRACE_ID(rc))
> + cs_path->trace_id = rc;
> + else
> + cs_path->trace_id = 0;
> + }
> + else
> + cs_path->trace_id = 0;
[...]
> +/**
> + * struct coresight_path - data needed by enable/disable path
> + * @handle: perf aux handle for ETM.
> + * @path: path from source to sink.
> + * @trace_id: trace_id of the whole path.
> + */
> +struct coresight_path {
> + struct perf_output_handle *handle;
This is only needed to avoid adding *handle to the enable function call
signature, but having it here implies it needs to be stored. And then we
need to manage the lifecycle of it by nulling it on deletion. All of
this can be avoided by just adding handle to enable().
Unrelated to this patch, but I'm not sure why we were passing around
void* for handle either. It just makes the code hard to read and implies
some flexibility that doesn't exist. It's always "struct
perf_output_handle", so we can change void* to that in the enable
functions. I also have a patch for this that I'll share in a bit.
> + struct list_head *path;
> + u8 trace_id;
> +};
> +
> static inline void coresight_insert_barrier_packet(void *buf)
> {
> if (buf)
> @@ -132,16 +144,15 @@ static inline void CS_UNLOCK(void __iomem *addr)
> } while (0);
> }
>
> -void coresight_disable_path(struct list_head *path);
> -int coresight_enable_path(struct list_head *path, enum cs_mode mode,
> - void *sink_data);
> +void coresight_disable_path(struct coresight_path *cs_path);
> +int coresight_enable_path(struct coresight_path *cs_path, enum cs_mode mode);
> struct coresight_device *coresight_get_sink(struct list_head *path);
This needs to be exported otherwise the build fails because you use it
in a module in another commit. I assume you are building as static?
next prev parent reply other threads:[~2025-01-28 11:54 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-01-24 7:25 [PATCH v9 0/6] Coresight: Add Coresight TMC Control Unit driver Jie Gan
2025-01-24 7:25 ` [PATCH v9 1/6] Coresight: Add support for new APB clock name Jie Gan
2025-01-24 7:25 ` [PATCH v9 2/6] Coresight: Add trace_id function to retrieving the trace ID Jie Gan
2025-01-24 7:25 ` [PATCH v9 3/6] Coresight: Introduce a new struct coresight_path Jie Gan
2025-01-28 11:54 ` James Clark [this message]
2025-01-29 0:57 ` Jie Gan
2025-01-31 16:36 ` [PATCH 0/3] coresight: Alloc trace ID after building the path James Clark
2025-01-31 16:36 ` [PATCH 1/3] coresight: Don't save handle in path James Clark
2025-02-06 3:02 ` Jie Gan
2025-02-06 14:34 ` James Clark
2025-01-31 16:36 ` [PATCH 2/3] coresight: Export coresight_get_sink() James Clark
2025-01-31 16:36 ` [PATCH 3/3] coresight: Alloc trace ID after building the path James Clark
2025-02-05 4:13 ` [PATCH 0/3] " Jie Gan
2025-02-05 7:44 ` Jie Gan
2025-01-31 16:43 ` [PATCH v9 3/6] Coresight: Introduce a new struct coresight_path James Clark
2025-02-05 4:09 ` Jie Gan
2025-01-24 7:25 ` [PATCH v9 4/6] dt-bindings: arm: Add Coresight TMC Control Unit hardware Jie Gan
2025-01-24 7:25 ` [PATCH v9 5/6] Coresight: Add Coresight TMC Control Unit driver Jie Gan
2025-01-28 11:55 ` James Clark
2025-01-29 0:46 ` Jie Gan
2025-01-29 10:35 ` James Clark
2025-01-29 13:02 ` Jie Gan
2025-01-29 14:07 ` James Clark
2025-01-24 7:25 ` [PATCH v9 6/6] arm64: dts: qcom: sa8775p: Add CTCU and ETR nodes Jie Gan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=538ae543-4c43-4df2-8adc-911096fe14dd@linaro.org \
--to=james.clark@linaro.org \
--cc=alexander.shishkin@linux.intel.com \
--cc=alexandre.torgue@foss.st.com \
--cc=andersson@kernel.org \
--cc=conor+dt@kernel.org \
--cc=coresight@lists.linaro.org \
--cc=devicetree@vger.kernel.org \
--cc=konradybcio@kernel.org \
--cc=krzk+dt@kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-stm32@st-md-mailman.stormreply.com \
--cc=mcoquelin.stm32@gmail.com \
--cc=mike.leach@linaro.org \
--cc=quic_jiegan@quicinc.com \
--cc=quic_jinlmao@quicinc.com \
--cc=quic_tingweiz@quicinc.com \
--cc=robh@kernel.org \
--cc=suzuki.poulose@arm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®