From: sudeep.holla@arm.com (Sudeep Holla)
To: linus-amlogic@lists.infradead.org
Subject: [PATCH 06/13] scpi: add priv_scpi_ops and fill legacy structure
Date: Tue, 23 Aug 2016 15:50:01 +0100 [thread overview]
Message-ID: <cfbb96a6-4a90-01f2-591e-8f0853d7c06e@arm.com> (raw)
In-Reply-To: <a0c9c4bc-cd83-c2e7-bfab-b8d24c886349@baylibre.com>
On 23/08/16 09:22, Neil Armstrong wrote:
> On 08/19/2016 06:39 PM, Sudeep Holla wrote:
>>
>>
>> On 18/08/16 11:10, Neil Armstrong wrote:
>>> In order to use the legacy functions variants, add a new priv_scpi_ops
>>> structure that will contain the internal alterne functions and then use these
>>> alternate call in the probe function.
>>>
>>> Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
>>> ---
>>> drivers/firmware/arm_scpi.c | 68 ++++++++++++++++++++++++++++++++++++++++++---
>>> 1 file changed, 64 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/drivers/firmware/arm_scpi.c b/drivers/firmware/arm_scpi.c
>>> index b0d911b..3fe39fe 100644
>>> --- a/drivers/firmware/arm_scpi.c
>>> +++ b/drivers/firmware/arm_scpi.c
>>> @@ -213,6 +213,7 @@ struct scpi_drvinfo {
>>> struct scpi_ops *scpi_ops;
>>> struct scpi_chan *channels;
>>> struct scpi_dvfs_info *dvfs[MAX_DVFS_DOMAINS];
>>> + const struct priv_scpi_ops *ops;
>>> };
>>>
>>> /*
>>> @@ -299,6 +300,17 @@ struct dev_pstate_set {
>>> u8 pstate;
>>> } __packed;
>>>
>>> +struct priv_scpi_ops {
>>> + /* Internal Specific Ops */
>>> + void (*handle_remote_msg)(struct mbox_client *c, void *msg);
>>> + void (*tx_prepare)(struct mbox_client *c, void *msg);
>>> + /* Message Specific Ops */
>>> + int (*init_versions)(struct scpi_drvinfo *info);
>>> + int (*dvfs_get_info)(u8 domain, struct dvfs_info *buf);
>>> + /* System wide Ops */
>>> + struct scpi_ops *scpi_ops;
>>> +};
>>> +
>>
>>
>> I fail to understand the need for this. Can you please explain the issue
>> you would face without this ?
>>
>>> static struct scpi_drvinfo *scpi_info;
>>>
>>> static int scpi_linux_errmap[SCPI_ERR_MAX] = {
>>> @@ -695,9 +707,12 @@ static struct scpi_dvfs_info *scpi_dvfs_get_info(u8 domain)
>>> if (scpi_info->dvfs[domain]) /* data already populated */
>>> return scpi_info->dvfs[domain];
>>>
>>> - ret = scpi_send_message(SCPI_CMD_GET_DVFS_INFO, &domain, sizeof(domain),
>>> + if (scpi_info->ops && scpi_info->ops->dvfs_get_info)
>>> + ret = scpi_info->ops->dvfs_get_info(domain, &buf);
>>> + else
>>> + ret = scpi_send_message(SCPI_CMD_GET_DVFS_INFO,
>>> + &domain, sizeof(domain),
>>> &buf, sizeof(buf));
>>> -
>>> if (ret)
>>> return ERR_PTR(ret);
>>>
>>> @@ -855,6 +870,22 @@ static struct scpi_ops scpi_ops = {
>>> .vendor_send_message = scpi_ext_send_message,
>>> };
>>>
>>> +static struct scpi_ops legacy_scpi_ops = {
>>> + .get_version = scpi_get_version,
>>> + .clk_get_range = NULL,
>>> + .clk_get_val = legacy_scpi_clk_get_val,
>>> + .clk_set_val = legacy_scpi_clk_set_val,
>>> + .dvfs_get_idx = legacy_scpi_dvfs_get_idx,
>>> + .dvfs_set_idx = legacy_scpi_dvfs_set_idx,
>>> + .dvfs_get_info = scpi_dvfs_get_info,
>>> + .sensor_get_capability = legacy_scpi_sensor_get_capability,
>>> + .sensor_get_info = legacy_scpi_sensor_get_info,
>>> + .sensor_get_value = legacy_scpi_sensor_get_value,
>>> + .device_get_power_state = NULL,
>>> + .device_set_power_state = NULL,
>>> + .vendor_send_message = legacy_scpi_send_message,
>>
>> I think we need not have this at all if you follow the suggestion I had
>> in the previous patch. Try and let's see how it would look.
>
> If you confirm you want the if/else as said in patch 4.
>
> But clk_get_range, device_get/set_power_state are not available in legacy,
> I think we should still have this alternate structure.
>
I was thinking of overriding the pointers accordingly at the probe time
as the common list is bigger than the one that differs.
--
Regards,
Sudeep
next prev parent reply other threads:[~2016-08-23 14:50 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-08-18 10:10 [PATCH 00/13] scpi: Add support for legacy SCPI protocol Neil Armstrong
2016-08-18 10:10 ` [PATCH 01/13] scpi: Add vendor_send_message to enable access to vendor commands Neil Armstrong
2016-08-18 15:53 ` Sudeep Holla
2016-08-19 8:00 ` Neil Armstrong
2016-08-19 10:39 ` Sudeep Holla
2016-08-18 10:10 ` [PATCH 02/13] scpi: Add alternative legacy structures and macros Neil Armstrong
2016-08-18 17:16 ` Sudeep Holla
2016-08-23 8:09 ` Neil Armstrong
2016-08-18 10:10 ` [PATCH 03/13] scpi: Add legacy send, prepare and handle remote functions Neil Armstrong
2016-08-19 16:13 ` Sudeep Holla
2016-08-23 8:15 ` Neil Armstrong
2016-08-23 14:42 ` Sudeep Holla
2016-08-18 10:10 ` [PATCH 04/13] scpi: Add legacy SCP functions calling legacy_scpi_send_message Neil Armstrong
2016-08-19 16:22 ` Sudeep Holla
2016-08-23 8:19 ` Neil Armstrong
2016-08-23 14:47 ` Sudeep Holla
2016-08-18 10:10 ` [PATCH 05/13] scpi: move of_match table before probe functions Neil Armstrong
2016-08-19 16:24 ` Sudeep Holla
2016-08-23 8:20 ` Neil Armstrong
2016-08-18 10:10 ` [PATCH 06/13] scpi: add priv_scpi_ops and fill legacy structure Neil Armstrong
2016-08-19 16:39 ` Sudeep Holla
2016-08-23 8:22 ` Neil Armstrong
2016-08-23 14:50 ` Sudeep Holla [this message]
2016-08-18 10:11 ` [PATCH 07/13] scpi: ignore init_versions failure if reported not supported Neil Armstrong
2016-08-19 16:46 ` Sudeep Holla
2016-08-23 8:23 ` Neil Armstrong
2016-08-23 14:54 ` Sudeep Holla
2016-08-23 14:55 ` Neil Armstrong
2016-08-23 15:01 ` Sudeep Holla
2016-08-18 10:11 ` [PATCH 08/13] scpi: add a vendor_msg mechanism in case the mailbox message differs Neil Armstrong
2016-08-19 16:47 ` Sudeep Holla
2016-08-18 10:11 ` [PATCH 09/13] scpi: implement rockchip support via the vendor_msg mechanism Neil Armstrong
2016-08-18 10:11 ` [PATCH 10/13] scpi: grow MAX_DVFS_OPPS to 16 entries Neil Armstrong
2016-08-18 10:11 ` [PATCH 11/13] dt-bindings: Add support for Amlogic GXBB SCPI Interface Neil Armstrong
2016-08-19 13:45 ` Rob Herring
2016-08-18 10:11 ` [PATCH 12/13] ARM64: dts: meson-gxbb: Add SRAM node Neil Armstrong
2016-08-18 10:11 ` [PATCH 13/13] ARM64: dts: meson-gxbb: Add SCPI with cpufreq & sensors Nodes Neil Armstrong
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=cfbb96a6-4a90-01f2-591e-8f0853d7c06e@arm.com \
--to=sudeep.holla@arm.com \
--cc=linus-amlogic@lists.infradead.org \
/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;
as well as URLs for NNTP newsgroup(s).