linux-kernel - Re: [PATCH v3 1/4] ACPI / scan: Fix acpi_is_indirect_io

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: <7d0cb757-a1a1-27bd-170c-87f602d5c22c@redhat.com>
Date:   Tue, 7 Aug 2018 12:41:26 +0200
From:   Hans de Goede <hdegoede@...hat.com>
To:     John Garry <john.garry@...wei.com>,
        "Rafael J . Wysocki" <rjw@...ysocki.net>,
        Len Brown <lenb@...nel.org>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Mika Westerberg <mika.westerberg@...ux.intel.com>,
        Darren Hart <dvhart@...radead.org>,
        Wolfram Sang <wsa@...-dreams.de>
Cc:     Srinivas Pandruvada <srinivas.pandruvada@...ux.intel.com>,
        linux-acpi@...r.kernel.org, platform-driver-x86@...r.kernel.org,
        linux-kernel@...r.kernel.org,
        Heikki Krogerus <heikki.krogerus@...ux.intel.com>,
        linux-i2c@...r.kernel.org, stable@...r.kernel.org,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        Dann Frazier <dann.frazier@...onical.com>,
        Linuxarm <linuxarm@...wei.com>
Subject: Re: [PATCH v3 1/4] ACPI / scan: Fix acpi_is_indirect_io_slave()
 always returning false

Hi,

On 07-08-18 12:05, John Garry wrote:
> On 07/08/2018 09:05, Hans de Goede wrote:
>> Since commit 63347db0affa ("ACPI / scan: Use acpi_bus_get_status() to
>> initialize ACPI_TYPE_DEVICE devs") the status field of normal acpi_devices
>> gets set to 0 by acpi_bus_type_and_status() and filled with its actual
>> value later when acpi_add_single_object() calls acpi_bus_get_status().
>>
>> This means that status is 0 when acpi_device_enumeration_by_parent()
>> gets called which makes the acpi_match_device_ids() call in
>> acpi_is_indirect_io_slave() always return -ENOENT.
> 
> For for the (child) device the status may be zero, but how about the parent device (which is what we care about)? As I see, we would have set the parent device status present field at this stage.

Ah I did not notice you were calling acpi_match_device_ids() on the parents.
Yes I think you are right. This commit is still a necessary part of the
series, but it needs a completely new commit message no longer
referencing acpi_is_indirect_io_slave().

I will submit a v4 of this patch-set with new commit message for this patch.

Regards,

Hans






> 
> Thanks,
> John
> 
>>
>> This commit fixes this by making acpi_bus_type_and_status() set status to
>> ACPI_STA_DEFAULT until acpi_add_single_object() calls acpi_bus_get_status()
>>
>> Fixes: dfda4492322e ("ACPI / scan: Do not enumerate Indirect IO host children")
>> Cc: stable@...r.kernel.org
>> Cc: John Garry <john.garry@...wei.com>
>> Cc: Bjorn Helgaas <bhelgaas@...gle.com>
>> Cc: Dann Frazier <dann.frazier@...onical.com>
>> Signed-off-by: Hans de Goede <hdegoede@...hat.com>
>> ---
>>  drivers/acpi/scan.c | 5 +++--
>>  1 file changed, 3 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
>> index 970dd87d347c..6799d00dd790 100644
>> --- a/drivers/acpi/scan.c
>> +++ b/drivers/acpi/scan.c
>> @@ -1612,7 +1612,8 @@ static int acpi_add_single_object(struct acpi_device **child,
>>       * Note this must be done before the get power-/wakeup_dev-flags calls.
>>       */
>>      if (type == ACPI_BUS_TYPE_DEVICE)
>> -        acpi_bus_get_status(device);
>> +        if (acpi_bus_get_status(device) < 0)
>> +            acpi_set_device_status(device, 0);
>>
>>      acpi_bus_get_power_flags(device);
>>      acpi_bus_get_wakeup_device_flags(device);
>> @@ -1690,7 +1691,7 @@ static int acpi_bus_type_and_status(acpi_handle handle, int *type,
>>           * acpi_add_single_object updates this once we've an acpi_device
>>           * so that acpi_bus_get_status' quirk handling can be used.
>>           */
>> -        *sta = 0;
>> +        *sta = ACPI_STA_DEFAULT;
>>          break;
>>      case ACPI_TYPE_PROCESSOR:
>>          *type = ACPI_BUS_TYPE_PROCESSOR;
>>
> 
>