From: Sergey Shtylyov <s.shtylyov@omp.ru>
To: "Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>
Cc: "Andrew Lunn" <andrew@lunn.ch>,
"Ulf Hansson" <ulf.hansson@linaro.org>,
"Vignesh Raghavendra" <vigneshr@ti.com>,
"KVM list" <kvm@vger.kernel.org>,
"Rafael J. Wysocki" <rafael@kernel.org>,
linux-iio@vger.kernel.org,
"Linus Walleij" <linus.walleij@linaro.org>,
"Amit Kucheria" <amitk@kernel.org>,
"ALSA Development Mailing List" <alsa-devel@alsa-project.org>,
"Andy Shevchenko" <andriy.shevchenko@linux.intel.com>,
"Jaroslav Kysela" <perex@perex.cz>,
"Guenter Roeck" <groeck@chromium.org>,
"Thierry Reding" <thierry.reding@gmail.com>,
"MTD Maling List" <linux-mtd@lists.infradead.org>,
"Linux I2C" <linux-i2c@vger.kernel.org>,
"Miquel Raynal" <miquel.raynal@bootlin.com>,
linux-phy@lists.infradead.org,
"Jiri Slaby" <jirislaby@kernel.org>,
openipmi-developer@lists.sourceforge.net,
"Khuong Dinh" <khuong@os.amperecomputing.com>,
"Florian Fainelli" <f.fainelli@gmail.com>,
"Matthias Schiffer" <matthias.schiffer@ew.tq-group.com>,
"Joakim Zhang" <qiangqing.zhang@nxp.com>,
"Kamal Dasu" <kdasu.kdev@gmail.com>,
"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
"Lee Jones" <lee.jones@linaro.org>,
"Bartosz Golaszewski" <brgl@bgdev.pl>,
"Daniel Lezcano" <daniel.lezcano@linaro.org>,
"Kishon Vijay Abraham I" <kishon@ti.com>,
"Geert Uytterhoeven" <geert@linux-m68k.org>,
"open list:SERIAL DRIVERS" <linux-serial@vger.kernel.org>,
bcm-kernel-feedback-list <bcm-kernel-feedback-list@broadcom.com>,
"Zhang Rui" <rui.zhang@intel.com>,
"Linux PWM List" <linux-pwm@vger.kernel.org>,
"Robert Richter" <rric@kernel.org>,
"Saravanan Sekar" <sravanhome@gmail.com>,
"Corey Minyard" <minyard@acm.org>,
"Linux PM list" <linux-pm@vger.kernel.org>,
"Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>,
"Mauro Carvalho Chehab" <mchehab@kernel.org>,
"John Garry" <john.garry@huawei.com>,
"Peter Korsgaard" <peter@korsgaard.com>,
"William Breathitt Gray" <vilhelm.gray@gmail.com>,
"Mark Gross" <markgross@kernel.org>,
"open list:GPIO SUBSYSTEM" <linux-gpio@vger.kernel.org>,
"Alex Williamson" <alex.williamson@redhat.com>,
"Mark Brown" <broonie@kernel.org>,
"Borislav Petkov" <bp@alien8.de>,
"Sebastian Reichel" <sre@kernel.org>,
"Eric Auger" <eric.auger@redhat.com>,
"Jakub Kicinski" <kuba@kernel.org>,
"Matthias Brugger" <matthias.bgg@gmail.com>,
"Takashi Iwai" <tiwai@suse.com>,
platform-driver-x86@vger.kernel.org,
"Benson Leung" <bleung@chromium.org>,
"Linux ARM" <linux-arm-kernel@lists.infradead.org>,
linux-edac@vger.kernel.org, "Tony Luck" <tony.luck@intel.com>,
"Mun Yew Tham" <mun.yew.tham@intel.com>,
"Hans de Goede" <hdegoede@redhat.com>,
netdev@vger.kernel.org,
"Yoshihiro Shimoda" <yoshihiro.shimoda.uh@renesas.com>,
"Cornelia Huck" <cohuck@redhat.com>,
"Linux MMC List" <linux-mmc@vger.kernel.org>,
"Liam Girdwood" <lgirdwood@gmail.com>,
linux-spi <linux-spi@vger.kernel.org>,
Linux-Renesas <linux-renesas-soc@vger.kernel.org>,
"Vinod Koul" <vkoul@kernel.org>,
"James Morse" <james.morse@arm.com>,
"Zha Qipeng" <qipeng.zha@intel.com>,
"Pengutronix Kernel Team" <kernel@pengutronix.de>,
"Richard Weinberger" <richard@nod.at>,
"Niklas Söderlund" <niklas.soderlund@ragnatech.se>,
linux-mediatek@lists.infradead.org,
"Brian Norris" <computersforpeace@gmail.com>,
"David S. Miller" <davem@davemloft.net>
Subject: Re: [PATCH 1/2] platform: make platform_get_irq_optional() optional
Date: Wed, 19 Jan 2022 18:34:53 +0300 [thread overview]
Message-ID: <b6b059bd-1491-26e2-9ba4-d34705a7106d@omp.ru> (raw)
In-Reply-To: <20220118222606.3iwuzbenl7g6oeiq@pengutronix.de>
Hello!
On 1/19/22 1:26 AM, Uwe Kleine-König wrote:
[...]
>>>>>>> However for an interupt this cannot work. You will always have to check
>>>>>>> if the irq is actually there or not because if it's not you cannot just
>>>>>>> ignore that. So there is no benefit of an optional irq.
>>>>>>>
>>>>>>> Leaving error message reporting aside, the introduction of
>>>>>>> platform_get_irq_optional() allows to change
>>>>>>>
>>>>>>> irq = platform_get_irq(...);
>>>>>>> if (irq < 0 && irq != -ENXIO) {
>>>>>>> return irq;
>>>>>>> } else if (irq >= 0) {
>>>>>>
>>>>>> Rather (irq > 0) actually, IRQ0 is considered invalid (but still returned).
>>>>>
>>>>> This is a topic I don't feel strong for, so I'm sloppy here. If changing
>>>>> this is all that is needed to convince you of my point ...
>>>>
>>>> Note that we should absolutely (and first of all) stop returning 0 from platform_get_irq()
>>>> on a "real" IRQ0. Handling that "still good" zero absolutely doesn't scale e.g. for the subsystems
>>>> (like libata) which take 0 as an indication that the polling mode should be used... We can't afford
>>>> to be sloppy here. ;-)
>>>
>>> Then maybe do that really first?
>>
>> I'm doing it first already:
>>
>> https://lore.kernel.org/all/5e001ec1-d3f1-bcb8-7f30-a6301fd9930c@omp.ru/
>>
>> This series is atop of the above patch...
>
> Ah, I missed that (probably because I didn't get the cover letter).
>
>>> I didn't recheck, but is this what the
>>> driver changes in your patch is about?
>>
>> Partly, yes. We can afford to play with the meaning of 0 after the above patch.
>
> But the changes that are in patch 1 are all needed?
Yes, they follow from the platform_get_irq_optional() changing the sense of its result...
[...]
>>> For my part I'd say this doesn't justify the change, but at least I
>>> could better life with the reasoning. If you start at:
>>>
>>> irq = platform_get_irq_optional(...)
>>> if (irq < 0 && irq != -ENXIO)
>>> return irq
>>> else if (irq > 0)
>>> setup_irq(irq);
>>> else
>>> setup_polling()
>>>
>>> I'd change that to
>>>
>>> irq = platform_get_irq_optional(...)
>>> if (irq > 0) /* or >= 0 ? */
>>
>> Not >= 0, no...
>>
>>> setup_irq(irq)
>>> else if (irq == -ENXIO)
>>> setup_polling()
>>> else
>>> return irq
>>>
>>> This still has to mention -ENXIO, but this is ok and checking for 0 just
>>> hardcodes a different return value.
>>
>> I think comparing with 0 is simpler (and shorter) than with -ENXIO, if you
>> consider the RISC CPUs, like e.g. MIPS...
>
> Hmm, I don't know MIPS good enough to judge. So I created a small C
MIPS has read-only register 0 (containing 0 :-)) which should simplify things. But
I'd have to check the actual object code... yeah, MIPS has a branching instruction that
compares 2 registers and branches on the result'; with -ENXIO you'd have to load an
immediate value into a register first...
> file:
>
> $ cat test.c
> #include <errno.h>
>
> int platform_get_irq_optional(void);
> void a(void);
>
> int func_0()
> {
> int irq = platform_get_irq_optional();
>
> if (irq == 0)
> a();
> }
>
> int func_enxio()
> {
> int irq = platform_get_irq_optional();
>
> if (irq == -ENXIO)
> a();
> }
>
> With some cross compilers as provided by Debian doing
>
> $CC -c -O3 test.c
Mhm, do we really use -O3 to build the kernel?
> nm --size-sort test.o
>
> I get:
>
> compiler | size of func_0 | size of func_enxio
> ================================+==================|====================
> aarch64-linux-gnu-gcc | 0000000000000024 | 0000000000000028
> arm-linux-gnueabi-gcc | 00000018 | 00000018
> arm-linux-gnueabihf-gcc | 00000010 | 00000012
Hm, 2 bytes only -- perhaps Thumb mode?
> i686-linux-gnu-gcc | 0000002a | 0000002a
Expected.
> mips64el-linux-gnuabi64-gcc | 0000000000000054 | 000000000000005c
That's even more than expected -- 64-bit mode used?
> powerpc-linux-gnu-gcc | 00000058 | 00000058
Well, they say
> s390x-linux-gnu-gcc | 000000000000002e | 0000000000000030
> x86_64-linux-gnu-gcc | 0000000000000022 | 0000000000000022
Again, as expected...
> So you save some bytes indeed.
I see you have a lot of spare time (unlike me!). :-)
>>> Anyhow, I think if you still want to change platform_get_irq_optional
>>> you should add a few patches converting some drivers which demonstrates
>>> the improvement for the callers.
>>
>> Mhm, I did include all the drivers where the IRQ checks have to be modified,
>> not sure what else you want me to touch...
>
> I somehow expected that the changes that are now necessary (or possible)
> to callers makes them prettier somehow. Looking at your patch again:
I think they do...
>
> - drivers/counter/interrupt-cnt.c
> This one is strange in my eyes because it tests the return value of
> gpiod_get_optional against NULL :-(
Mhm, how is this connected with my patch? :-/
> - drivers/edac/xgene_edac.c
> This one just wants a silent irq lookup and then throws away the
> error code returned by platform_get_irq_optional() to return -EINVAL.
> Not so nice, is it?
I have dropped this file from my (to be posted yet) v2... sorry that it
took so long...
> - drivers/gpio/gpio-altera.c
> This one just wants a silent irq lookup. And maybe it should only
> goto skip_irq if the irq was not found, but on an other error code
> abort the probe?!
That's debatable... but anyway it's a matter of a separate (follow up)
patch...
>
> - drivers/gpio/gpio-mvebu.c
> Similar to gpio-altera.c: Wants a silent irq and improved error
> handling.
Same as above...
> - drivers/i2c/busses/i2c-brcmstb.c
> A bit ugly that we now have dev->irq == 0 if the irq isn't available,
> but if requesting the irq failed irq = -1 is used?
This doesn't matter much really but can change to 0, if you want... :-)
>
> - drivers/mmc/host/sh_mmcif.c
> Broken error handling. This one wants to abort on irq[1] < 0 (with
> your changed semantic).
Again, matter of a separate patch (I don't have the guily hardware anymore
but I guess Geert could help with that).
> I stopped here.
Note that most of your complaints are about the existing driver logic --
which my patch just couldn't deal with... I currently don't have the bandwidth
for addressing all your complaints; some (more obvious) I'm goiing to address
as the time permits, the draft patches have been done already...
> It seems quite common that drivers assume a value < 0 returned by
> platform_get_irq means not-found
Of course, before this patch -ENXIO meant IRQ-not-found, many drivers
don't bother to filter it out separately (for simplicity?).
> and don't care for -EPROBE_DEFER (what else can happen?).
Hm, I haven't really seen a lot the probe dererral mishandling in the code
touched by at least my patch #1...
> Changing a relevant function in that mess seems unfortunate here :-\
You seem to have some spare time and I'm getting distracted contrariwise...
want to help? :-)
> Best regards
> Uwe
MBR, Sergey
next prev parent reply other threads:[~2022-01-19 15:35 UTC|newest]
Thread overview: 94+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20220110195449.12448-1-s.shtylyov@omp.ru>
2022-01-10 19:54 ` [PATCH 1/2] platform: make platform_get_irq_optional() optional Sergey Shtylyov
2022-01-10 20:10 ` Uwe Kleine-König
2022-01-10 21:07 ` Andy Shevchenko
2022-01-10 21:44 ` Uwe Kleine-König
2022-01-10 21:18 ` Andrew Lunn
2022-01-12 8:33 ` Geert Uytterhoeven
2022-01-12 8:50 ` Uwe Kleine-König
2022-01-12 10:27 ` Geert Uytterhoeven
2022-01-12 12:27 ` Andy Shevchenko
2022-01-12 13:38 ` Andrew Lunn
2022-01-12 13:51 ` Andy Shevchenko
2022-01-12 13:54 ` Geert Uytterhoeven
2022-01-12 14:41 ` Rafael J. Wysocki
2022-01-12 15:05 ` Sergey Shtylyov
2022-01-12 15:13 ` Hans de Goede
2022-01-12 15:48 ` Rafael J. Wysocki
2022-01-12 16:21 ` Andy Shevchenko
2022-01-12 21:31 ` Uwe Kleine-König
2022-01-12 21:45 ` Mark Brown
2022-01-13 11:08 ` Uwe Kleine-König
2022-01-13 14:45 ` Mark Brown
2022-01-13 19:43 ` [PATCH] driver core: platform: Rename platform_get_irq_optional() to platform_get_irq_silent() Uwe Kleine-König
2022-01-13 20:17 ` Mark Brown
2022-01-13 20:57 ` Sergey Shtylyov
2022-01-13 22:43 ` Uwe Kleine-König
2022-01-14 9:58 ` Geert Uytterhoeven
2022-01-15 15:21 ` Uwe Kleine-König
2022-01-15 21:06 ` Sergey Shtylyov
2022-01-14 11:34 ` Sergey Shtylyov
2022-01-13 20:31 ` Guenter Roeck
2022-01-13 21:42 ` Florian Fainelli
2022-01-13 23:06 ` Uwe Kleine-König
2022-01-14 17:55 ` Sergey Shtylyov
2022-01-15 13:55 ` Uwe Kleine-König
2022-01-19 18:36 ` Andy Shevchenko
2022-01-19 19:44 ` Sergey Shtylyov
2022-01-19 18:40 ` Andy Shevchenko
2022-01-14 6:57 ` Peter Korsgaard
2022-01-14 13:04 ` Andy Shevchenko
2022-01-15 15:45 ` Uwe Kleine-König
2022-01-19 18:51 ` Andy Shevchenko
2022-01-19 19:47 ` Sergey Shtylyov
2022-01-19 20:55 ` Andy Shevchenko
2022-01-20 7:57 ` Uwe Kleine-König
2022-01-24 15:01 ` Andy Shevchenko
2022-01-24 21:02 ` Sergey Shtylyov
2022-01-25 8:25 ` Geert Uytterhoeven
2022-01-25 12:56 ` Matthias Schiffer
2022-01-25 14:01 ` Andy Shevchenko
2022-01-14 19:45 ` Sergey Shtylyov
2022-01-14 20:29 ` Uwe Kleine-König
2022-01-15 13:08 ` Sergey Shtylyov
2022-01-19 18:52 ` Andy Shevchenko
2022-01-13 20:35 ` [PATCH 1/2] platform: make platform_get_irq_optional() optional Sergey Shtylyov
2022-01-14 9:25 ` Uwe Kleine-König
2022-01-14 9:39 ` Geert Uytterhoeven
2022-01-14 19:14 ` Sergey Shtylyov
2022-01-14 20:22 ` Uwe Kleine-König
2022-01-15 20:22 ` Sergey Shtylyov
2022-01-17 8:41 ` Geert Uytterhoeven
2022-01-17 9:24 ` Uwe Kleine-König
2022-01-17 10:35 ` Geert Uytterhoeven
2022-01-17 11:49 ` Uwe Kleine-König
2022-01-17 13:08 ` Geert Uytterhoeven
2022-01-17 17:06 ` Uwe Kleine-König
2022-01-18 8:25 ` Geert Uytterhoeven
2022-01-18 9:09 ` Uwe Kleine-König
2022-01-18 9:37 ` Geert Uytterhoeven
2022-01-18 12:08 ` Uwe Kleine-König
2022-01-18 12:49 ` Geert Uytterhoeven
2022-01-18 14:29 ` Uwe Kleine-König
2022-01-19 16:12 ` Sergey Shtylyov
2022-01-19 18:29 ` Sergey Shtylyov
2022-01-20 11:27 ` Sergey Shtylyov
2022-01-16 18:15 ` Sergey Shtylyov
2022-01-17 8:47 ` Uwe Kleine-König
2022-01-18 20:21 ` Sergey Shtylyov
2022-01-18 22:26 ` Uwe Kleine-König
2022-01-19 15:34 ` Sergey Shtylyov [this message]
2022-01-19 18:58 ` Andy Shevchenko
2022-01-19 19:49 ` Sergey Shtylyov
2022-01-14 18:46 ` Sergey Shtylyov
2022-01-15 18:36 ` [PATCH 1/2] platform: make platform_get_irq_optional() optional (summary) Uwe Kleine-König
2022-01-16 14:19 ` Greg Kroah-Hartman
2022-01-18 9:18 ` Uwe Kleine-König
2022-01-18 9:32 ` Greg Kroah-Hartman
2022-01-19 10:56 ` Sergey Shtylyov
2022-01-19 11:33 ` Uwe Kleine-König
2022-01-19 11:42 ` Sergey Shtylyov
2022-01-19 19:06 ` Andy Shevchenko
2022-01-17 11:57 ` [PATCH 1/2] platform: make platform_get_irq_optional() optional Sergey Shtylyov
2022-01-19 15:02 ` Uwe Kleine-König
2022-01-19 15:50 ` Sergey Shtylyov
2022-01-10 19:54 ` [PATCH 2/2] platform: make platform_get_irq_byname_optional() optional Sergey Shtylyov
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=b6b059bd-1491-26e2-9ba4-d34705a7106d@omp.ru \
--to=s.shtylyov@omp.ru \
--cc=alex.williamson@redhat.com \
--cc=alsa-devel@alsa-project.org \
--cc=amitk@kernel.org \
--cc=andrew@lunn.ch \
--cc=andriy.shevchenko@linux.intel.com \
--cc=bcm-kernel-feedback-list@broadcom.com \
--cc=bleung@chromium.org \
--cc=bp@alien8.de \
--cc=brgl@bgdev.pl \
--cc=broonie@kernel.org \
--cc=cohuck@redhat.com \
--cc=computersforpeace@gmail.com \
--cc=daniel.lezcano@linaro.org \
--cc=davem@davemloft.net \
--cc=eric.auger@redhat.com \
--cc=f.fainelli@gmail.com \
--cc=geert@linux-m68k.org \
--cc=gregkh@linuxfoundation.org \
--cc=groeck@chromium.org \
--cc=hdegoede@redhat.com \
--cc=james.morse@arm.com \
--cc=jirislaby@kernel.org \
--cc=john.garry@huawei.com \
--cc=kdasu.kdev@gmail.com \
--cc=kernel@pengutronix.de \
--cc=khuong@os.amperecomputing.com \
--cc=kishon@ti.com \
--cc=kuba@kernel.org \
--cc=kvm@vger.kernel.org \
--cc=lee.jones@linaro.org \
--cc=lgirdwood@gmail.com \
--cc=linus.walleij@linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-edac@vger.kernel.org \
--cc=linux-gpio@vger.kernel.org \
--cc=linux-i2c@vger.kernel.org \
--cc=linux-iio@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mediatek@lists.infradead.org \
--cc=linux-mmc@vger.kernel.org \
--cc=linux-mtd@lists.infradead.org \
--cc=linux-phy@lists.infradead.org \
--cc=linux-pm@vger.kernel.org \
--cc=linux-pwm@vger.kernel.org \
--cc=linux-renesas-soc@vger.kernel.org \
--cc=linux-serial@vger.kernel.org \
--cc=linux-spi@vger.kernel.org \
--cc=markgross@kernel.org \
--cc=matthias.bgg@gmail.com \
--cc=matthias.schiffer@ew.tq-group.com \
--cc=mchehab@kernel.org \
--cc=minyard@acm.org \
--cc=miquel.raynal@bootlin.com \
--cc=mun.yew.tham@intel.com \
--cc=netdev@vger.kernel.org \
--cc=niklas.soderlund@ragnatech.se \
--cc=openipmi-developer@lists.sourceforge.net \
--cc=perex@perex.cz \
--cc=peter@korsgaard.com \
--cc=platform-driver-x86@vger.kernel.org \
--cc=qiangqing.zhang@nxp.com \
--cc=qipeng.zha@intel.com \
--cc=rafael@kernel.org \
--cc=richard@nod.at \
--cc=rric@kernel.org \
--cc=rui.zhang@intel.com \
--cc=sravanhome@gmail.com \
--cc=sre@kernel.org \
--cc=thierry.reding@gmail.com \
--cc=tiwai@suse.com \
--cc=tony.luck@intel.com \
--cc=u.kleine-koenig@pengutronix.de \
--cc=ulf.hansson@linaro.org \
--cc=vigneshr@ti.com \
--cc=vilhelm.gray@gmail.com \
--cc=vkoul@kernel.org \
--cc=yoshihiro.shimoda.uh@renesas.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;
as well as URLs for NNTP newsgroup(s).