From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755001Ab2DGAS6 (ORCPT ); Fri, 6 Apr 2012 20:18:58 -0400 Received: from acsinet15.oracle.com ([141.146.126.227]:48980 "EHLO acsinet15.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752811Ab2DGAS5 convert rfc822-to-8bit (ORCPT ); Fri, 6 Apr 2012 20:18:57 -0400 Date: Fri, 6 Apr 2012 20:14:14 -0400 From: Konrad Rzeszutek Wilk To: Lin Ming Cc: linux-kernel@vger.kernel.org Subject: Re: ACPI: Move module parameter gts and bfs to sleep.c Message-ID: <20120407001414.GA12026@phenom.dumpdata.com> References: <20120406222329.GA19837@phenom.dumpdata.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) Content-Transfer-Encoding: 8BIT X-Source-IP: ucsinet21.oracle.com [156.151.31.93] X-CT-RefId: str=0001.0A090204.4F7F87EF.007B,ss=1,re=0.000,fgs=0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, Apr 07, 2012 at 08:10:23AM +0800, Lin Ming wrote: > On Sat, Apr 7, 2012 at 6:23 AM, Konrad Rzeszutek Wilk > wrote: > > Hey Lin, > > > > I was wondering if you are OK with exporting the wake_sleep_flags > > function so that other callers of acpi_enter_sleep (which are > > outside the sleep.c) could use it? > > > > Specifically this patch > > http://git.kernel.org/?p=linux/kernel/git/konrad/xen.git;a=blobdiff;f=include/xen/acpi.h;h=ebaabbbe91877cc9d7188254463b4a591f6bb81d;hp=48a9c0171b658904b33a82df0d4aa31fe8b8ed81;hb=331a7ed09457ca7b6c31170c5d59a09e6f1e02c2;hpb=68504cdd16c26735078f65a0bdc8eeb9f0c493ea > > > > [which I was hoping to post for v3.5] > > .. which in 3.3 would call acpi_enter_sleep_state without > > trouble but with 3.4 it needs a parameter. > > > > This little patch (compile tested, but not actually run): > > > > commit 8f9f29819fa71de17e2e6f38fc50c4156dc3dada > > Author: Konrad Rzeszutek Wilk > > Date:   Fri Apr 6 17:53:51 2012 -0400 > > > >    ACPI: Convert wake_sleep_flags to a value instead of function. > > > >    With commit a2ef5c4fd44ce3922435139393b89f2cce47f576 > >    "ACPI: Move module parameter gts and bfs to sleep.c" the wake_sleep_flags > >    is required when calling acpi_enter_sleep_state, which means > >    that if there are functions outside the sleep.c code they > >    can't get get the wake_sleep_flags values. > > > >    This converts the function in to a exported value and converts > >    the module config operands to a function. > > Why not just export the function? That is possible too. > Do you mean to save a function call by converting it to a value? That was my thinking. > > Looks good to me. > > Thanks, > Lin Ming > > > > >    Signed-off-by: Konrad Rzeszutek Wilk > > > > diff --git a/arch/x86/kernel/acpi/sleep.h b/arch/x86/kernel/acpi/sleep.h > > index 4d3feb5..f84e9cd 100644 > > --- a/arch/x86/kernel/acpi/sleep.h > > +++ b/arch/x86/kernel/acpi/sleep.h > > @@ -9,6 +9,8 @@ extern long saved_magic; > > > >  extern int wakeup_pmode_return; > > > > +extern unsigned wake_sleep_flags; > > + > >  extern unsigned long acpi_copy_wakeup_routine(unsigned long); > >  extern void wakeup_long64(void); > > > > diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c > > index 6a8d4d3..dcee73a 100644 > > --- a/drivers/acpi/sleep.c > > +++ b/drivers/acpi/sleep.c > > @@ -28,23 +28,23 @@ > >  #include "internal.h" > >  #include "sleep.h" > > > > +u8 wake_sleep_flags = ACPI_NO_OPTIONAL_METHODS; > >  static unsigned int gts, bfs; > > -module_param(gts, uint, 0644); > > -module_param(bfs, uint, 0644); > > -MODULE_PARM_DESC(gts, "Enable evaluation of _GTS on suspend."); > > -MODULE_PARM_DESC(bfs, "Enable evaluation of _BFS on resume".); > > - > > -static u8 wake_sleep_flags(void) > > +static int set_param_wake_flag(const char *val, struct kernel_param *kp) > >  { > > -       u8 flags = ACPI_NO_OPTIONAL_METHODS; > > +       int ret = param_set_int(val, kp); > > > >        if (gts) > > -               flags |= ACPI_EXECUTE_GTS; > > +               wake_sleep_flags |= ACPI_EXECUTE_GTS; > >        if (bfs) > > -               flags |= ACPI_EXECUTE_BFS; > > +               wake_sleep_flags |= ACPI_EXECUTE_BFS; > > > > -       return flags; > > +       return ret; > >  } > > +module_param_call(gts, set_param_wake_flag, param_get_int, >s, 0644); > > +module_param_call(bfs, set_param_wake_flag, param_get_int, &bfs, 0644); > > +MODULE_PARM_DESC(gts, "Enable evaluation of _GTS on suspend."); > > +MODULE_PARM_DESC(bfs, "Enable evaluation of _BFS on resume".); > > > >  static u8 sleep_states[ACPI_S_STATE_COUNT]; > > > > @@ -263,7 +263,6 @@ static int acpi_suspend_enter(suspend_state_t pm_state) > >  { > >        acpi_status status = AE_OK; > >        u32 acpi_state = acpi_target_sleep_state; > > -       u8 flags = wake_sleep_flags(); > >        int error; > > > >        ACPI_FLUSH_CPU_CACHE(); > > @@ -271,7 +270,7 @@ static int acpi_suspend_enter(suspend_state_t pm_state) > >        switch (acpi_state) { > >        case ACPI_STATE_S1: > >                barrier(); > > -               status = acpi_enter_sleep_state(acpi_state, flags); > > +               status = acpi_enter_sleep_state(acpi_state, wake_sleep_flags); > >                break; > > > >        case ACPI_STATE_S3: > > @@ -288,7 +287,7 @@ static int acpi_suspend_enter(suspend_state_t pm_state) > >        acpi_write_bit_register(ACPI_BITREG_SCI_ENABLE, 1); > > > >        /* Reprogram control registers and execute _BFS */ > > -       acpi_leave_sleep_state_prep(acpi_state, flags); > > +       acpi_leave_sleep_state_prep(acpi_state, wake_sleep_flags); > > > >        /* ACPI 3.0 specs (P62) says that it's the responsibility > >         * of the OSPM to clear the status bit [ implying that the > > @@ -552,30 +551,27 @@ static int acpi_hibernation_begin(void) > > > >  static int acpi_hibernation_enter(void) > >  { > > -       u8 flags = wake_sleep_flags(); > >        acpi_status status = AE_OK; > > > >        ACPI_FLUSH_CPU_CACHE(); > > > >        /* This shouldn't return.  If it returns, we have a problem */ > > -       status = acpi_enter_sleep_state(ACPI_STATE_S4, flags); > > +       status = acpi_enter_sleep_state(ACPI_STATE_S4, wake_sleep_flags); > >        /* Reprogram control registers and execute _BFS */ > > -       acpi_leave_sleep_state_prep(ACPI_STATE_S4, flags); > > +       acpi_leave_sleep_state_prep(ACPI_STATE_S4, wake_sleep_flags); > > > >        return ACPI_SUCCESS(status) ? 0 : -EFAULT; > >  } > > > >  static void acpi_hibernation_leave(void) > >  { > > -       u8 flags = wake_sleep_flags(); > > - > >        /* > >         * If ACPI is not enabled by the BIOS and the boot kernel, we need to > >         * enable it here. > >         */ > >        acpi_enable(); > >        /* Reprogram control registers and execute _BFS */ > > -       acpi_leave_sleep_state_prep(ACPI_STATE_S4, flags); > > +       acpi_leave_sleep_state_prep(ACPI_STATE_S4, wake_sleep_flags); > >        /* Check the hardware signature */ > >        if (facs && s4_hardware_signature != facs->hardware_signature) { > >                printk(KERN_EMERG "ACPI: Hardware changed while hibernated, " > > @@ -830,12 +826,10 @@ static void acpi_power_off_prepare(void) > > > >  static void acpi_power_off(void) > >  { > > -       u8 flags = wake_sleep_flags(); > > - > >        /* acpi_sleep_prepare(ACPI_STATE_S5) should have already been called */ > >        printk(KERN_DEBUG "%s called\n", __func__); > >        local_irq_disable(); > > -       acpi_enter_sleep_state(ACPI_STATE_S5, flags); > > +       acpi_enter_sleep_state(ACPI_STATE_S5, wake_sleep_flags); > >  } > > > >  /* > > --