linux-kernel - Re: [PATCH] power: supply: bq27xxx_battery: do not update cached flags prematurely

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <Z3440OtztmLUv7HA@tp440p.steeds.sam>
Date: Wed, 8 Jan 2025 10:35:28 +0200
From: Sicelo <absicsz@...il.com>
To: linux-pm@...r.kernel.org, Krzysztof Kozlowski <krzk@...nel.org>,
	Pali Rohár <pali@...nel.org>
Cc: linux-pm@...r.kernel.org, sre@...nel.org, linux-kernel@...r.kernel.org,
	maemo-leste@...ts.dyne.org
Subject: Re: [PATCH] power: supply: bq27xxx_battery: do not update cached
 flags prematurely

Hi 

On Mon, Nov 25, 2024 at 06:59:57PM +0100, Pali Rohár wrote:
> cc Krzysztof, author of the mentioned commit.

Yes, apologies for overlooking to cc him.

> On Monday 25 November 2024 17:29:30 Sicelo A. Mhlongo wrote:
> > Commit 243f8ffc883a1 ("power: supply: bq27xxx_battery: Notify also about
> > status changes") intended to notify userspace when the status changes,
> > based on the flags register. However, the cached state is updated too
> > early, before the flags are tested for any changes. Remove the premature
> > update.
> > 
> > Fixes: 243f8ffc883a1 ("power: supply: bq27xxx_battery: Notify also about status changes")
> > Signed-off-by: Sicelo A. Mhlongo <absicsz@...il.com>
> > ---
> >  drivers/power/supply/bq27xxx_battery.c | 1 -
> >  1 file changed, 1 deletion(-)
> > 
> > diff --git a/drivers/power/supply/bq27xxx_battery.c b/drivers/power/supply/bq27xxx_battery.c
> > index 40c5ac7a1118..b2c65fe43d5c 100644
> > --- a/drivers/power/supply/bq27xxx_battery.c
> > +++ b/drivers/power/supply/bq27xxx_battery.c
> > @@ -1913,7 +1913,6 @@ static void bq27xxx_battery_update_unlocked(struct bq27xxx_device_info *di)
> >  		cache.flags = -1; /* read error */
> >  	if (cache.flags >= 0) {
> >  		cache.capacity = bq27xxx_battery_read_soc(di);
> > -		di->cache.flags = cache.flags;
> >  
> >  		/*
> >  		 * On gauges with signed current reporting the current must be
> > -- 
> > 2.45.2

Is there anything that still needs to be done to move this patch along?

Perhaps for clarification, in the middle of bq27xxx_battery_update_unlocked,
there is:

	if ((di->cache.capacity != cache.capacity) ||
	    (di->cache.flags != cache.flags) ||
	    (di->last_status.intval != status.intval)) {
		di->last_status.intval = status.intval;
		power_supply_changed(di->bat);
	}

	if (memcmp(&di->cache, &cache, sizeof(cache)) != 0)
		di->cache = cache;

which compares the new value of cache.flags with the previously cached
values in di->cache, and calls power_supply_changed() if they are
different. For the flags, this test is currently not working, because by
the time it is run, di->cache.flags has already been updated with the
value in cache.flags, so the flags condition is always short-circuited
in the test. Hence my patch ensures di->cache.flags is not updated
before the test.

The flags are correctly updated after the test, in the 

        di->cache = cache;

line, so the removed line is simply not needed.

If there is any adjustment necessary, I will happily do so.

Kind Regards
Sicelo A. Mhlongo