linux-kernel - RE: [PATCH AUTOSEL for 4.9 01/56] ACPICA: Resources: Not a valid resource if buffer length too long

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: <20171122232412.k4ohrqpkc7xroydm@sasha-lappy>
Date:   Wed, 22 Nov 2017 23:24:14 +0000
From:   alexander.levin@...izon.com
To:     "Schmauss, Erik" <erik.schmauss@...el.com>
CC:     "Moore, Robert" <robert.moore@...el.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "stable@...r.kernel.org" <stable@...r.kernel.org>,
        "Wysocki, Rafael J" <rafael.j.wysocki@...el.com>
Subject: RE: [PATCH AUTOSEL for 4.9 01/56] ACPICA: Resources: Not a valid
 resource if buffer length too long

On Mon, Nov 20, 2017 at 05:22:37PM +0000, Schmauss, Erik wrote:
>> -----Original Message-----
>> From: alexander.levin@...izon.com [mailto:alexander.levin@...izon.com]
>> Sent: Thursday, November 16, 2017 4:56 PM
>> To: Schmauss, Erik <erik.schmauss@...el.com>
>> Cc: Moore, Robert <robert.moore@...el.com>; linux-kernel@...r.kernel.org;
>> stable@...r.kernel.org; Wysocki, Rafael J <rafael.j.wysocki@...el.com>
>> Subject: RE: [PATCH AUTOSEL for 4.9 01/56] ACPICA: Resources: Not a valid
>> resource if buffer length too long
>>
>> On Wed, Nov 15, 2017 at 05:05:21PM +0000, Schmauss, Erik wrote:
>> >
>> >
>> >> -----Original Message-----
>> >> From: alexander.levin@...izon.com
>> >> [mailto:alexander.levin@...izon.com]
>> >> Sent: Wednesday, November 15, 2017 8:39 AM
>> >> To: Moore, Robert <robert.moore@...el.com>
>> >> Cc: linux-kernel@...r.kernel.org; stable@...r.kernel.org; Wysocki,
>> >> Rafael J <rafael.j.wysocki@...el.com>; Schmauss, Erik
>> >> <erik.schmauss@...el.com>
>> >> Subject: RE: [PATCH AUTOSEL for 4.9 01/56] ACPICA: Resources: Not a
>> >> valid resource if buffer length too long
>> >>
>> >> On Wed, Nov 15, 2017 at 03:39:22PM +0000, Moore, Robert wrote:
>> >> >> -----Original Message-----
>> >> >> From: alexander.levin@...izon.com
>> >> >> [mailto:alexander.levin@...izon.com]
>> >> >> Sent: Tuesday, November 14, 2017 6:46 PM
>> >> >> To: linux-kernel@...r.kernel.org; stable@...r.kernel.org
>> >> >> Cc: Moore, Robert <robert.moore@...el.com>; Zheng, Lv
>> >> >> <lv.zheng@...el.com>; Wysocki, Rafael J
>> >> >> <rafael.j.wysocki@...el.com>; alexander.levin@...izon.com
>> >> >> Subject: [PATCH AUTOSEL for 4.9 01/56] ACPICA: Resources: Not a
>> >> >> valid resource if buffer length too long
>> >> >>
>> >> >> From: Bob Moore <robert.moore@...el.com>
>> >> >>
>> >> >> [ Upstream commit 57707a9a7780fab426b8ae9b4c7b65b912a748b3 ]
>> >> >>
>> >> >> ACPICA commit 9f76de2d249b18804e35fb55d14b1c2604d627a1
>> >> >> ACPICA commit b2e89d72ef1e9deefd63c3fd1dee90f893575b3a
>> >> >> ACPICA commit 23b5bbe6d78afd3c5abf3adb91a1b098a3000b2e
>> >> >>
>> >> >> The declared buffer length must be the same as the length of the
>> >> >> byte initializer list, otherwise not a valid resource descriptor.
>> >> [snip]
>> >>
>> >> >[Moore, Robert]
>> >> >
>> >> >Please explain what you are doing here.
>> >>
>> >> Proposing this commit for the 4.9 LTS tree.
>> >
>> >What problem are you trying to solve with this change? Are you seeing ACPI
>> errors? If so what are they?
>>
>> Not seeing an actual problem myself. Was this patch supposed to fix a problem
>> or just deal with a theoretical scenario?
>>
>This was supposed to fix issues with our AML disassembler to parse a strange corner in the ASL test suite. I believe this was due to the end tag contains a checksum byte that ACPICA and other ACPI implementations ignore. We thought this was useful in the test suite because we test the disassembler by comparing normally compiled AML with AML that has been compiled, disassembled, and re-compiled. Without this change, the endtag checksum would be over-written to 0 rather than the existing value.
>
>I believe this change broke a few things such as the execution of ConcatResTemplate. We concluded to not use this solution and alter the test suite instead because this endtag byte is usually ignored anyway. To answer your question, this was to deal with a theoretical scenario.

I'll drop it then, thanks Erik!

-- 

Thanks,
Sasha