From: Guenter Roeck Date: Sat, 14 Dec 2013 17:30:15 +0000 Subject: [lm-sensors] [PATCH 08/14] i8k: Force SMM to run on CPU 0 Message-Id: <1387042221-3671-9-git-send-email-linux@roeck-us.net> List-Id: References: <1387042221-3671-1-git-send-email-linux@roeck-us.net> In-Reply-To: <1387042221-3671-1-git-send-email-linux@roeck-us.net> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Arnd Bergmann , Greg Kroah-Hartman Cc: Andrew Morton , lm-sensors@lm-sensors.org, linux-kernel@vger.kernel.org, Guenter Roeck , Jean Delvare On Studio 1555 with dual-core CPU, reading sensor attributes exported by this driver resulted in random failures combined with system hangups and forced logouts. Information in drivers/firmware/dcdbas.c suggests that SMM accesses must run on CPU 0. With this patch, the problems are gone, suggesting that this is in fact the case. Code derived from drivers/firmware/dcdbas.c. Cc: Jean Delvare Signed-off-by: Guenter Roeck --- drivers/char/i8k.c | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) diff --git a/drivers/char/i8k.c b/drivers/char/i8k.c index ad0acef..264be2b 100644 --- a/drivers/char/i8k.c +++ b/drivers/char/i8k.c @@ -31,6 +31,7 @@ #include #include #include +#include #include @@ -130,6 +131,17 @@ static int i8k_smm(struct smm_regs *regs) { int rc; int eax = regs->eax; + cpumask_var_t old_mask; + + /* SMM requires CPU 0 */ + if (!alloc_cpumask_var(&old_mask, GFP_KERNEL)) + return -ENOMEM; + cpumask_copy(old_mask, ¤t->cpus_allowed); + set_cpus_allowed_ptr(current, cpumask_of(0)); + if (smp_processor_id() != 0) { + rc = -EBUSY; + goto out; + } #if defined(CONFIG_X86_64) asm volatile("pushq %%rax\n\t" @@ -185,9 +197,12 @@ static int i8k_smm(struct smm_regs *regs) : "%ebx", "%ecx", "%edx", "%esi", "%edi", "memory"); #endif if (rc != 0 || (regs->eax & 0xffff) = 0xffff || regs->eax = eax) - return -EINVAL; + rc = -EINVAL; - return 0; +out: + set_cpus_allowed_ptr(current, old_mask); + free_cpumask_var(old_mask); + return rc; } /* -- 1.7.9.7 _______________________________________________ lm-sensors mailing list lm-sensors@lm-sensors.org http://lists.lm-sensors.org/mailman/listinfo/lm-sensors