| 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
| ||
|
Message-ID: <55117373.3020808@amd.com>
Date: Tue, 24 Mar 2015 09:23:47 -0500
From: Suravee Suthikulpanit <Suravee.Suthikulpanit@....com>
To: Mika Westerberg <mika.westerberg@...ux.intel.com>
CC: <rjw@...ysocki.net>, <lv.zheng@...el.com>, <hpa@...ux.intel.com>,
<lenb@...nel.org>, <hdegoede@...hat.com>, <tj@...nel.org>,
<arnd@...db.de>, <mjg59@...f.ucam.org>,
<gregkh@...uxfoundation.org>, <hanjun.guo@...aro.org>,
<al.stone@...aro.org>, <graeme.gregory@...aro.org>,
<leo.duran@....com>, <linux-ide@...r.kernel.org>,
<linux-acpi@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
<linaro-acpi@...ts.linaro.org>
Subject: Re: [V5 PATCH 1/2] ACPI / scan: Add support for ACPI _CLS device
matching
Not sure if this email went out originally. So, I am resending this.
On 3/9/15 10:20, Mika Westerberg wrote:
> On Fri, Mar 06, 2015 at 12:11:41AM -0600, Suravee Suthikulpanit wrote:
>> Device drivers typically use ACPI _HIDs/_CIDs listed in struct device_driver
>> acpi_match_table to match devices. However, for generic drivers, we do not
>> want to list _HID for all supported devices. Also, certain classes of devices
>> do not have _CID (e.g. SATA, USB). Instead, we can leverage ACPI _CLS,
>> which specifies PCI-defined class code (i.e. base-class, subclass and
>> programming interface). This patch adds support for matching ACPI devices using
>> the _CLS method.
>>
>> To support loadable module, current design uses _HID or _CID to match device's
>> modalias. With the new way of matching with _CLS this would requires modification
>> to the current ACPI modalias key to include _CLS. This patch appends PCI-defined
>> class-code to the existing ACPI modalias as following.
>>
>> acpi:<HID>:<CID1>:<CID2>:..:<CIDn>:<bbsspp>:
>> E.g:
>> # cat /sys/devices/platform/AMDI0600:00/modalias
>> acpi:AMDI0600:010601:
>>
>> where bb is th base-class code, ss is te sub-class code, and pp is the
>> programming interface code
>>
>> Since there would not be _HID/_CID in the ACPI matching table of the driver,
>> this patch adds a field to acpi_device_id to specify the matching _CLS.
>>
>> static const struct acpi_device_id ahci_acpi_match[] = {
>> { "", 0, PCI_CLASS_STORAGE_SATA_AHCI },
>
> How about introducing macro like PCI already has and then do it like:
>
> { ACPI_DEVICE_CLASS(PCI_CLASS_STORAGE_SATA_AHCI) },
This is good. I'll update this.
>
> Also should we allow mask here as well? This would allow partial match
> if, for example we are only interested in class and subclass.
Hm, I'm not familiar with how other drivers might be benefit from this.
Could you please give an example of a drivers that only deals with just
the class and/or subclass of devices?
If so, could we use 0xFF to specify the don't care field?
>
>> {},
>> };
>>
>> In this case, the corresponded entry in modules.alias file would be:
>>
>> alias acpi*:010601:* ahci_platform
>>
>> Signed-off-by: Suravee Suthikulpanit <Suravee.Suthikulpanit@....com>
>> ---
>> drivers/acpi/acpica/acutils.h | 3 ++
>> drivers/acpi/acpica/nsxfname.c | 21 ++++++++++--
>> drivers/acpi/acpica/utids.c | 71 +++++++++++++++++++++++++++++++++++++++
>
> I think you need to split the ACPICA part to be a separate patch. Those
> are merged through ACPICA.
I will update this in the next patch set.
>> [....]
>> diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h
>> index e530533..9a42522 100644
>> --- a/include/linux/mod_devicetable.h
>> +++ b/include/linux/mod_devicetable.h
>> @@ -189,6 +189,7 @@ struct css_device_id {
>> struct acpi_device_id {
>> __u8 id[ACPI_ID_LEN];
>> kernel_ulong_t driver_data;
>> + __u32 cls;
>
> It would be nice if we could change ordering here but I understand that
> it breaks quite many drivers. Perhaps we should consider creating
> ACPI_DEVICE() macro and convert existing drivers to that at some point.
Yes, a roughly grep in the drivers directory showing about 112 files at
the moment. If you think this is the right approach going forward, we
can work on cleaning this up on a separate patch series. Please let me
know what you think.
Thanks,
Suravee
>> };
>>
>> #define PNP_ID_LEN 8
>> diff --git a/scripts/mod/devicetable-offsets.c b/scripts/mod/devicetable-offsets.c
>> index f282516..7f68268 100644
>> --- a/scripts/mod/devicetable-offsets.c
>> +++ b/scripts/mod/devicetable-offsets.c
>> @@ -63,6 +63,7 @@ int main(void)
>>
>> DEVID(acpi_device_id);
>> DEVID_FIELD(acpi_device_id, id);
>> + DEVID_FIELD(acpi_device_id, cls);
>>
>> DEVID(pnp_device_id);
>> DEVID_FIELD(pnp_device_id, id);
>> diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c
>> index e614ef6..ba5998c 100644
>> --- a/scripts/mod/file2alias.c
>> +++ b/scripts/mod/file2alias.c
>> @@ -511,12 +511,21 @@ static int do_serio_entry(const char *filename,
>> }
>> ADD_TO_DEVTABLE("serio", serio_device_id, do_serio_entry);
>>
>> -/* looks like: "acpi:ACPI0003 or acpi:PNP0C0B" or "acpi:LNXVIDEO" */
>> +/* looks like: "acpi:ACPI0003" or "acpi:PNP0C0B" or "acpi:LNXVIDEO" or
>> + * "acpi:bbsspp" (bb=base-class, ss=sub-class, pp=prog-if)
>> + *
>> + * NOTE: * Each driver should use one of the following : _HID, _CIDs or _CLS.
>> + */
>> static int do_acpi_entry(const char *filename,
>> void *symval, char *alias)
>> {
>> DEF_FIELD_ADDR(symval, acpi_device_id, id);
>> - sprintf(alias, "acpi*:%s:*", *id);
>> + DEF_FIELD_ADDR(symval, acpi_device_id, cls);
>> +
>> + if (id && strlen((const char *)*id))
>> + sprintf(alias, "acpi*:%s:*", *id);
>> + else if (cls)
>> + sprintf(alias, "acpi*:%06x:*", *cls);
>> return 1;
>> }
>> ADD_TO_DEVTABLE("acpi", acpi_device_id, do_acpi_entry);
>> --
>> 2.1.0
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists