From: Jeffrey Hugo <jhugo@codeaurora.org>
To: Stephen Boyd <sboyd@kernel.org>,
Michael Turquette <mturquette@baylibre.com>
Cc: linux-kernel@vger.kernel.org, linux-clk@vger.kernel.org,
Miquel Raynal <miquel.raynal@bootlin.com>,
Jerome Brunet <jbrunet@baylibre.com>,
Russell King <linux@armlinux.org.uk>
Subject: Re: [PATCH 2/9] clk: Introduce get_parent_hw clk op
Date: Mon, 11 Feb 2019 09:09:00 -0700 [thread overview]
Message-ID: <41f8f325-2738-3995-1fc8-354a9412041e@codeaurora.org> (raw)
In-Reply-To: <20190129061021.94775-3-sboyd@kernel.org>
On 1/28/2019 11:10 PM, Stephen Boyd wrote:
> The clk_ops::get_parent function is limited in ability to return errors
> because it returns a u8. A "workaround" to return an error is to return
> a number >= the number of parents of a clk. This will in turn cause the
> framework to "orphan" the clk and make the parent of the clk NULL. This
> isn't really correct, because if an error occurs while reading the
> parents of a clk we should fail the clk registration, instead of
> orphaning the clk and waiting for the clk to appear later.
>
> We really need to have three different return values from the get_parent
> clk op. Something valid for a clk that exists, something invalid for a
> clk that doesn't exist and never will exist or can't be determined
> because the register operation to read the parent failed, and something
> for a clk that doesn't exist because the framework doesn't know about
> what it is. Introduce a new clk_op that can express all of this by
> returning a pointer to the clk_hw of the parent. It's expected that clk
> provider drivers will return a valid pointer when the parent is
> findable, an error pointer like EPROBE_DEFER if their parent provider
> hasn't probed yet but is valid, a NULL pointer if they can't find the
> clk but index is valid, and an error pointer with an appropriate error
> code otherwise.
>
> Cc: Miquel Raynal <miquel.raynal@bootlin.com>
> Cc: Jerome Brunet <jbrunet@baylibre.com>
> Cc: Russell King <linux@armlinux.org.uk>
> Cc: Michael Turquette <mturquette@baylibre.com>
> Signed-off-by: Stephen Boyd <sboyd@kernel.org>
<snip>
> +static int clk_init_parent(struct clk_core *core)
> +{
> + core->parent = __clk_init_parent(core, true);
> + if (IS_ERR(core->parent))
> + return PTR_ERR(core->parent);
> +
> + /*
> + * Populate core->parent if parent has already been clk_core_init'd. If
> + * parent has not yet been clk_core_init'd then place clk in the orphan
> + * list. If clk doesn't have any parents then place it in the root
> + * clk list.
> + *
> + * Every time a new clk is clk_init'd then we walk the list of orphan
> + * clocks and re-parent any that are children of the clock currently
> + * being clk_init'd.
> + */
> + if (core->parent) {
> + hlist_add_head(&core->child_node,
> + &core->parent->children);
> + core->orphan = core->parent->orphan;
> + } else if (!core->num_parents) {
> + hlist_add_head(&core->child_node, &clk_root_list);
> + core->orphan = false;
> + } else {
> + hlist_add_head(&core->child_node, &clk_orphan_list);
Missing "core->orphan = true;"?
The snippet below had that line. Its not clear why it appears to be
dropped here.
> + }
> +
> + return 0;
> +}
<snip>
> @@ -3073,29 +3143,9 @@ static int __clk_core_init(struct clk_core *core)
> "%s: invalid NULL in %s's .parent_names\n",
> __func__, core->name);
>
> - core->parent = __clk_init_parent(core);
> -
> - /*
> - * Populate core->parent if parent has already been clk_core_init'd. If
> - * parent has not yet been clk_core_init'd then place clk in the orphan
> - * list. If clk doesn't have any parents then place it in the root
> - * clk list.
> - *
> - * Every time a new clk is clk_init'd then we walk the list of orphan
> - * clocks and re-parent any that are children of the clock currently
> - * being clk_init'd.
> - */
> - if (core->parent) {
> - hlist_add_head(&core->child_node,
> - &core->parent->children);
> - core->orphan = core->parent->orphan;
> - } else if (!core->num_parents) {
> - hlist_add_head(&core->child_node, &clk_root_list);
> - core->orphan = false;
> - } else {
> - hlist_add_head(&core->child_node, &clk_orphan_list);
> - core->orphan = true;
> - }
> + ret = clk_init_parent(core);
> + if (ret)
> + goto out;
>
> /*
> * optional platform-specific magic
--
Jeffrey Hugo
Qualcomm Datacenter Technologies as an affiliate of Qualcomm
Technologies, Inc.
Qualcomm Technologies, Inc. is a member of the
Code Aurora Forum, a Linux Foundation Collaborative Project.
next prev parent reply other threads:[~2019-02-11 16:09 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-01-29 6:10 [PATCH 0/9] Rewrite clk parent handling Stephen Boyd
2019-01-29 6:10 ` [PATCH 1/9] clk: Combine __clk_get() and __clk_create_clk() Stephen Boyd
2019-01-29 6:10 ` [PATCH 2/9] clk: Introduce get_parent_hw clk op Stephen Boyd
2019-01-29 9:34 ` Jerome Brunet
2019-01-29 21:15 ` Stephen Boyd
2019-01-30 9:53 ` Jerome Brunet
2019-01-30 21:30 ` Stephen Boyd
2019-01-31 18:40 ` Jerome Brunet
2019-02-06 0:01 ` Stephen Boyd
2019-02-13 9:16 ` Jerome Brunet
2019-02-15 17:01 ` Stephen Boyd
2019-02-11 16:09 ` Jeffrey Hugo [this message]
2019-02-15 18:47 ` Stephen Boyd
2019-02-15 19:29 ` Jeffrey Hugo
2019-02-15 21:29 ` Stephen Boyd
2019-02-15 21:34 ` Jeffrey Hugo
2019-01-29 6:10 ` [PATCH 3/9] clk: core: clarify the check for runtime PM Stephen Boyd
2019-01-29 6:10 ` [PATCH 4/9] clk: Introduce of_clk_get_hw_from_clkspec() Stephen Boyd
2019-01-29 6:10 ` [PATCH 5/9] clk: Inform the core about consumer devices Stephen Boyd
2019-01-29 6:10 ` [PATCH 6/9] clk: Move of_clk_*() APIs into clk.c from clkdev.c Stephen Boyd
2019-01-29 6:10 ` [PATCH 7/9] clk: Allow parents to be specified without string names Stephen Boyd
2019-01-29 10:12 ` Jerome Brunet
2019-01-29 18:56 ` Stephen Boyd
2019-01-29 21:08 ` Jerome Brunet
2019-02-13 9:32 ` Jerome Brunet
2019-02-15 21:13 ` Stephen Boyd
2019-01-29 6:10 ` [PATCH 8/9] clk: qcom: gcc-sdm845: Migrate to DT parent mapping Stephen Boyd
2019-01-29 6:10 ` [PATCH 9/9] arm64: dts: qcom: Specify XO clk as input to GCC node Stephen Boyd
2019-01-29 10:12 ` [PATCH 0/9] Rewrite clk parent handling Miquel Raynal
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=41f8f325-2738-3995-1fc8-354a9412041e@codeaurora.org \
--to=jhugo@codeaurora.org \
--cc=jbrunet@baylibre.com \
--cc=linux-clk@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@armlinux.org.uk \
--cc=miquel.raynal@bootlin.com \
--cc=mturquette@baylibre.com \
--cc=sboyd@kernel.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).