From: Daniel Kurtz <djkurtz@google.com>
To: Heiner Kallweit <hkallweit1@gmail.com>
Cc: Jean Delvare <jdelvare@suse.de>, Linux I2C <linux-i2c@vger.kernel.org>
Subject: Re: [PATCH 2/2] i2c: i801: Replace waitqueue with completion API
Date: Thu, 27 May 2021 17:13:14 +1000 [thread overview]
Message-ID: <CAGS+omC_Ruk5UJHuPy2PqOhmdnnyGA4RwG+2=7oNRXisPzFxeQ@mail.gmail.com> (raw)
In-Reply-To: <d220a961-6653-ec27-430c-727dbe88b665@gmail.com>
Hi Jean, Heiner,
On Tue, May 25, 2021 at 11:01 PM Heiner Kallweit <hkallweit1@gmail.com> wrote:
>
> On 25.05.2021 14:12, Jean Delvare wrote:
> > Hi Heiner,
> >
> > Adding Daniel Kurtz, who wrote the original wait queue-based code, to
> > Cc. I'm not sure if Daniel is still into kernel code though.
Thanks for thinking of me! I'm still into kernel code, but I do a lot
more reading than writing these days.
> >
> > On Sat, 22 May 2021 00:02:43 +0200, Heiner Kallweit wrote:
> >> Using the completion API is more intuitive and it allows to simplify
> >> the code. Note that we don't have to set priv->status = 0 any longer
> >> with the completion API.
> >
> > OK, but you need to call reinit_completion() instead, which has the
> > same cost. So that's not a good argument ;-)
> >
> Maybe my comment wasn't clear enough. I didn't want to say that we
> save something but that it's safe from a functional point of view
> to remove clearing priv->status.
>
> >> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
> >> ---
> >> drivers/i2c/busses/i2c-i801.c | 48 ++++++++++++++---------------------
> >> 1 file changed, 19 insertions(+), 29 deletions(-)
> >>
> >> diff --git a/drivers/i2c/busses/i2c-i801.c b/drivers/i2c/busses/i2c-i801.c
> >> index bfea94d02..738204d77 100644
> >> --- a/drivers/i2c/busses/i2c-i801.c
> >> +++ b/drivers/i2c/busses/i2c-i801.c
> >> @@ -103,7 +103,7 @@
> >> #include <linux/dmi.h>
> >> #include <linux/slab.h>
> >> #include <linux/string.h>
> >> -#include <linux/wait.h>
> >> +#include <linux/completion.h>
> >> #include <linux/err.h>
> >> #include <linux/platform_device.h>
> >> #include <linux/platform_data/itco_wdt.h>
> >> @@ -270,7 +270,7 @@ struct i801_priv {
> >> unsigned int features;
> >>
> >> /* isr processing */
> >> - wait_queue_head_t waitq;
> >> + struct completion done;
> >> u8 status;
> >>
> >> /* Command state used by isr for byte-by-byte block transactions */
> >> @@ -496,24 +496,19 @@ static int i801_wait_byte_done(struct i801_priv *priv)
> >> static int i801_transaction(struct i801_priv *priv, int xact)
> >> {
> >> int status;
> >> - int result;
> >> + unsigned long result;
> >> const struct i2c_adapter *adap = &priv->adapter;
> >>
> >> - result = i801_check_pre(priv);
> >> - if (result < 0)
> >> - return result;
> >> + status = i801_check_pre(priv);
> >> + if (status < 0)
> >> + return status;
> >>
> >> if (priv->features & FEATURE_IRQ) {
> >> + reinit_completion(&priv->done);
> >> outb_p(xact | SMBHSTCNT_INTREN | SMBHSTCNT_START,
> >> SMBHSTCNT(priv));
> >> - result = wait_event_timeout(priv->waitq,
> >> - (status = priv->status),
> >> - adap->timeout);
> >> - if (!result)
> >> - status = -ETIMEDOUT;
> >> -
> >> - priv->status = 0;
> >> - return i801_check_post(priv, status);
> >> + result = wait_for_completion_timeout(&priv->done, adap->timeout);
> >> + return i801_check_post(priv, result ? priv->status : -ETIMEDOUT);
> >> }
> >>
> >> /* the current contents of SMBHSTCNT can be overwritten, since PEC,
> >> @@ -638,7 +633,7 @@ static irqreturn_t i801_host_notify_isr(struct i801_priv *priv)
> >> * DEV_ERR - Invalid command, NAK or communication timeout
> >> * BUS_ERR - SMI# transaction collision
> >> * FAILED - transaction was canceled due to a KILL request
> >> - * When any of these occur, update ->status and wake up the waitq.
> >> + * When any of these occur, update ->status and signal completion.
> >> * ->status must be cleared before kicking off the next transaction.
> >> *
> >> * 2) For byte-by-byte (I2C read/write) transactions, one BYTE_DONE interrupt
> >> @@ -675,7 +670,7 @@ static irqreturn_t i801_isr(int irq, void *dev_id)
> >> if (status) {
> >> outb_p(status, SMBHSTSTS(priv));
> >> priv->status = status;
> >> - wake_up(&priv->waitq);
> >> + complete(&priv->done);
> >> }
> >>
> >> return IRQ_HANDLED;
> >> @@ -694,15 +689,15 @@ static int i801_block_transaction_byte_by_byte(struct i801_priv *priv,
> >> int i, len;
> >> int smbcmd;
> >> int status;
> >> - int result;
> >> + unsigned long result;
> >> const struct i2c_adapter *adap = &priv->adapter;
> >>
> >> if (command == I2C_SMBUS_BLOCK_PROC_CALL)
> >> return -EOPNOTSUPP;
> >>
> >> - result = i801_check_pre(priv);
> >> - if (result < 0)
> >> - return result;
> >> + status = i801_check_pre(priv);
> >> + if (status < 0)
> >> + return status;
> >>
> >> len = data->block[0];
> >>
> >> @@ -726,15 +721,10 @@ static int i801_block_transaction_byte_by_byte(struct i801_priv *priv,
> >> priv->count = 0;
> >> priv->data = &data->block[1];
> >>
> >> + reinit_completion(&priv->done);
> >> outb_p(priv->cmd | SMBHSTCNT_START, SMBHSTCNT(priv));
> >> - result = wait_event_timeout(priv->waitq,
> >> - (status = priv->status),
> >> - adap->timeout);
> >> - if (!result)
> >> - status = -ETIMEDOUT;
> >> -
> >> - priv->status = 0;
> >> - return i801_check_post(priv, status);
> >> + result = wait_for_completion_timeout(&priv->done, adap->timeout);
> >> + return i801_check_post(priv, result ? priv->status : -ETIMEDOUT);
> >> }
> >>
> >> for (i = 1; i <= len; i++) {
> >> @@ -1889,7 +1879,7 @@ static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id)
> >> }
> >>
> >> if (priv->features & FEATURE_IRQ) {
> >> - init_waitqueue_head(&priv->waitq);
> >> + init_completion(&priv->done);
> >>
> >> err = devm_request_irq(&dev->dev, dev->irq, i801_isr,
> >> IRQF_SHARED,
> >
> > The clean-up isn't massive but I agree the code is a bit more pleasant
> > to read after the change, and I also see the binary size shrink a bit,
> > so I vote for it.
> >
> > Reviewed-by: Jean Delvare <jdelvare@suse.de>
> > Tested-by: Jean Delvare <jdelvare@suse.de>
> >
>
I agree, shorter and sweeter. Thanks for cleaning this up!
Reviewed-by: Daniel Kurtz <djkurtz@chromium.org>
next prev parent reply other threads:[~2021-05-27 7:13 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-21 21:59 [PATCH 0/2] i2c: i801: Replace waitqueue with completion API Heiner Kallweit
2021-05-21 22:00 ` [PATCH 1/2] i2c: i801: Remove unneeded warning after wait_event_timeout timeout Heiner Kallweit
2021-05-25 10:00 ` Jean Delvare
2021-05-27 20:09 ` Wolfram Sang
2021-05-21 22:02 ` [PATCH 2/2] i2c: i801: Replace waitqueue with completion API Heiner Kallweit
2021-05-25 12:12 ` Jean Delvare
2021-05-25 13:01 ` Heiner Kallweit
2021-05-27 7:13 ` Daniel Kurtz [this message]
2021-05-27 20:09 ` Wolfram Sang
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='CAGS+omC_Ruk5UJHuPy2PqOhmdnnyGA4RwG+2=7oNRXisPzFxeQ@mail.gmail.com' \
--to=djkurtz@google.com \
--cc=hkallweit1@gmail.com \
--cc=jdelvare@suse.de \
--cc=linux-i2c@vger.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).