linux-kernel - Re: [PATCH v5 2/8] acpi: fix enumeration (visited) flags for bus rescans

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

Message-ID: <CAE1zotJg+bCUD+q_cbiO1=abx1PJbUdWVr03uutwR081iG4G1g@mail.gmail.com>
Date:	Wed, 6 Jul 2016 18:37:46 +0300
From:	Octavian Purdila <octavian.purdila@...el.com>
To:	Mika Westerberg <mika.westerberg@...ux.intel.com>
Cc:	"Rafael J . Wysocki" <rjw@...ysocki.net>,
	Len Brown <lenb@...nel.org>,
	Matt Fleming <matt@...eblueprint.co.uk>,
	Mark Brown <broonie@...nel.org>,
	Wolfram Sang <wsa@...-dreams.de>,
	Joel Becker <jlbec@...lplan.org>,
	"linux-acpi@...r.kernel.org" <linux-acpi@...r.kernel.org>,
	linux-efi@...r.kernel.org, linux-i2c <linux-i2c@...r.kernel.org>,
	linux-spi <linux-spi@...r.kernel.org>,
	lkml <linux-kernel@...r.kernel.org>,
	Irina Tirdea <irina.tirdea@...el.com>,
	Leonard Crestez <leonard.crestez@...el.com>
Subject: Re: [PATCH v5 2/8] acpi: fix enumeration (visited) flags for bus rescans

On Wed, Jul 6, 2016 at 6:03 PM, Mika Westerberg
<mika.westerberg@...ux.intel.com> wrote:
> On Fri, Jul 01, 2016 at 11:19:06PM +0300, Octavian Purdila wrote:
>> If the ACPI tables changes as a result of a dinamically loaded table and
>> a bus rescan is required the enumeration/visited flag are not
>> consistent.
>>
>> I2C/SPI are not directly enumerated in acpi_bus_attach(), however the
>> visited flag is set. This makes it impossible to check if an ACPI device
>> has already been enumerated by the I2C and SPI subsystems. To fix this
>> issue we only set the visited flags if the device is not I2C or SPI.
>>
>> With this change we also need to remove setting visited to false from
>> acpi_bus_attach(), otherwise if we rescan already enumerated I2C/SPI
>> devices we try to re-enumerate them.
>>
>> Note that I2C/SPI devices can be enumerated either via a scan handler
>> (when using PRP0001) or via regular device_attach(). In either case
>> the flow goes through acpi_default_enumeration() which makes it the
>> ideal place to mark the ACPI device as enumerated.
>
> Hmm, this somehow fails to enumerate PRP0001 devices.
>
> I'm adding following SSDT to my initrd and the resulting device is never
> enumerated on SPI bus (it exists in /sys/bus/acpi/devices). If I use
> normal ACPI IDs (non-PRP0001) then it works fine.
>

I specifically tested with PRP0001 via initrd with I2C, I am not sure
why it does not work on SPI. I'll give it a try tomorrow when I have
access to my setup.