| 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: <5a22f1ff-44e9-4e1a-bdbe-cc168f718693@matthias-fetzer.de>
Date: Fri, 9 Aug 2024 23:06:34 +0200
From: Matthias Fetzer <kontakt@...thias-fetzer.de>
To: Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
Cc: hmh@....eng.br, Hans de Goede <hdegoede@...hat.com>,
ibm-acpi-devel@...ts.sourceforge.net, platform-driver-x86@...r.kernel.org,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] platform/x86: thinkpad_acpi: Add Thinkpad Edge E531 fan
support
Thanks for the review!
Am 08.08.24 um 15:14 schrieb Ilpo Järvinen:
> On Sun, 14 Jul 2024, Matthias Fetzer wrote:
>
>> Fan control on the E531 is done using the ACPI methods FANG and
>> FANW. The correct parameters and register values were found by
>> analyzing EC firmware as well as DSDT. This has been tested on
>> my Thinkpad Edge E531 (6885CTO, BIOS HEET52WW 1.33).
>>
>> Signed-off-by: Matthias Fetzer <kontakt@...thias-fetzer.de>
>> ---
>> drivers/platform/x86/thinkpad_acpi.c | 159 +++++++++++++++++++++++++++
>> 1 file changed, 159 insertions(+)
>>
>> diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c
>> index 397b409064c9..a171a2b39ac9 100644
>> --- a/drivers/platform/x86/thinkpad_acpi.c
>> +++ b/drivers/platform/x86/thinkpad_acpi.c
>> @@ -7751,6 +7751,28 @@ static struct ibm_struct volume_driver_data = {
>> * EC 0x2f (HFSP) might be available *for reading*, but do not use
>> * it for writing.
>> *
>> + * TPACPI_FAN_RD_ACPI_FANG:
>> + * ACPI FANG method: returns fan control register
>> + *
>> + * Takes one parameter which is 0x8100 plus the offset to EC memory
>> + * address 0xf500 and returns the byte at this address.
>> + *
>> + * 0xf500:
>> + * When the value is less than 9 automatic mode is enabled
>> + * 0xf502:
>> + * Contains the current fan speed from 0-100%
>> + * 0xf504:
>> + * Bit 7 has to be set in order to enable manual control by
>> + * writing a value >= 9 to 0xf500
>> + *
>> + * TPACPI_FAN_WR_ACPI_FANW:
>> + * ACPI FANG method: sets fan control registers
>> + *
>> + * Takes 0x8100 plus the offset to EC memory address 0xf500 and the
>> + * value to be written there as parameters.
>> + *
>> + * see TPACPI_FAN_RD_ACPI_FANG
>> + *
>> * TPACPI_FAN_WR_TPEC:
>> * ThinkPad EC register 0x2f (HFSP): fan control loop mode
>> * Supported on almost all ThinkPads
>> @@ -7884,6 +7906,7 @@ enum { /* Fan control constants */
>> enum fan_status_access_mode {
>> TPACPI_FAN_NONE = 0, /* No fan status or control */
>> TPACPI_FAN_RD_ACPI_GFAN, /* Use ACPI GFAN */
>> + TPACPI_FAN_RD_ACPI_FANG, /* Use ACPI FANG */
>> TPACPI_FAN_RD_TPEC, /* Use ACPI EC regs 0x2f, 0x84-0x85 */
>> TPACPI_FAN_RD_TPEC_NS, /* Use non-standard ACPI EC regs (eg: L13 Yoga gen2 etc.) */
>> };
>> @@ -7891,6 +7914,7 @@ enum fan_status_access_mode {
>> enum fan_control_access_mode {
>> TPACPI_FAN_WR_NONE = 0, /* No fan control */
>> TPACPI_FAN_WR_ACPI_SFAN, /* Use ACPI SFAN */
>> + TPACPI_FAN_WR_ACPI_FANW, /* Use ACPI FANW */
>> TPACPI_FAN_WR_TPEC, /* Use ACPI EC reg 0x2f */
>> TPACPI_FAN_WR_ACPI_FANS, /* Use ACPI FANS and EC reg 0x2f */
>> };
>> @@ -7924,9 +7948,13 @@ TPACPI_HANDLE(fans, ec, "FANS"); /* X31, X40, X41 */
>> TPACPI_HANDLE(gfan, ec, "GFAN", /* 570 */
>> "\\FSPD", /* 600e/x, 770e, 770x */
>> ); /* all others */
>> +TPACPI_HANDLE(fang, ec, "FANG", /* E531 */
>> + ); /* all others */
>> TPACPI_HANDLE(sfan, ec, "SFAN", /* 570 */
>> "JFNS", /* 770x-JL */
>> ); /* all others */
>> +TPACPI_HANDLE(fanw, ec, "FANW", /* E531 */
>> + ); /* all others */
>>
>> /*
>> * Unitialized HFSP quirk: ACPI DSDT and EC fail to initialize the
>> @@ -8033,6 +8061,23 @@ static int fan_get_status(u8 *status)
>>
>> break;
>> }
>> + case TPACPI_FAN_RD_ACPI_FANG: {
>> + /* E531 */
>> + int mode, speed;
>> +
>> + if (unlikely(!acpi_evalf(fang_handle, &mode, NULL, "dd", 0x8100)))
>> + return -EIO;
>> + if (unlikely(!acpi_evalf(fang_handle, &speed, NULL, "dd", 0x8102)))
>> + return -EIO;
>> +
>> + if (likely(status)) {
>> + *status = speed * 7 / 100;
>> + if (mode < 9)
>> + *status |= TP_EC_FAN_AUTO;
>> + }
>> +
>> + break;
>> + }
>> case TPACPI_FAN_RD_TPEC:
>> /* all except 570, 600e/x, 770e, 770x */
>> if (unlikely(!acpi_ec_read(fan_status_offset, &s)))
>> @@ -8147,6 +8192,17 @@ static int fan2_get_speed(unsigned int *speed)
>> if (speed)
>> *speed = lo ? FAN_RPM_CAL_CONST / lo : 0;
>> break;
>> + case TPACPI_FAN_RD_ACPI_FANG: {
>> + /* E531 */
>> + int speed_tmp;
>> +
>> + if (unlikely(!acpi_evalf(fang_handle, &speed_tmp, NULL, "dd", 0x8102)))
>> + return -EIO;
>> +
>> + if (likely(speed))
>> + *speed = speed_tmp * 65535 / 100;
>> + break;
>> + }
>>
>> default:
>> return -ENXIO;
>> @@ -8157,6 +8213,7 @@ static int fan2_get_speed(unsigned int *speed)
>>
>> static int fan_set_level(int level)
>> {
>> + int rc;
>> if (!fan_control_allowed)
>> return -EPERM;
>>
>> @@ -8206,6 +8263,36 @@ static int fan_set_level(int level)
>> tp_features.fan_ctrl_status_undef = 0;
>> break;
>>
>> + case TPACPI_FAN_WR_ACPI_FANW:
>> + if ((!(level & TP_EC_FAN_AUTO) &&
>> + ((level < 0) || (level > 7))) ||
>> + (level & TP_EC_FAN_FULLSPEED))
>> + return -EINVAL;
>
> I'd split this into two to make it more readable:
>
> if (!(level & TP_EC_FAN_AUTO) && (level < 0 || level > 7))
> return -EINVAL;
> if (level & TP_EC_FAN_FULLSPEED)
> return -EINVAL;
This is much better, thanks.
>
> I'm not sure if -EINVAL is really the right code to return though in these
> cases.
>
I thought that since those are invalid values/parameters the return code
of -EINVAL
would be a good choice. What do you suggest to use instead?
>> + if (level & TP_EC_FAN_AUTO) {
>> + if (!acpi_evalf(fanw_handle, NULL, NULL, "vdd", 0x8106, 0x05)) {
>
> Curiously enough, the comment above doesn't cover offset 0xf506 but the
> comment mentions 0xf504 that is never touched anywhere? Is that a typo?
>
Good catch! This indeed is a typo.
Powered by blists - more mailing lists