[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[PATCH v2 1/2] x86/mwait-idle: add 'preferred-cstates' command line option


  • To: "xen-devel@xxxxxxxxxxxxxxxxxxxx" <xen-devel@xxxxxxxxxxxxxxxxxxxx>
  • From: Jan Beulich <jbeulich@xxxxxxxx>
  • Date: Thu, 2 Jun 2022 12:26:44 +0200
  • Arc-authentication-results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=suse.com; dmarc=pass action=none header.from=suse.com; dkim=pass header.d=suse.com; arc=none
  • Arc-message-signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=y8ukhmYyNTQg/Apn6MkKOZE6epFoKJZGr2rITtOrpGk=; b=hwh5GxKSLa41f1i+rs7abwC7An/I8pfkrh6M0OJAPzKkLlur5i7PchAz11W8NkOayQrsHvsffYw+iMvCL7vhJDKXOpTta2MISsdiWIGcpJIzBjT5b+MkSfpniQ3WHTpxaKounweIAkKy+IXGcCTe/6+XZ+LQnOVKJe0aE4xop0SHKKRp4t2SZ0TUiYAZt6c2OzGEH8EyLDdNMg3xmxYitVkoq6TSxP+vrb3oAsY9kCLPOuW8dOvUPHFMNnf6cYJfbrLc938Nm+GfsuBy7mQUOLvHE7s8qOVm4HeNceDXSj7P2mLS9D5iNdbo8Rifxq7KzZfwLzahk80n/QJjzaWY4g==
  • Arc-seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=QiuLGAM/hgaAqSrT8+7X0U8Y53UD/rSAMBXTtb89bGx1VuZkIXXQKufX6mYTOw7cAudRqqvW4Z9qwCt48Ij1+MPLZGeiPXzPio0+lIe1xMVVgEQc1oqWJYU78oE4G14R5TRX/lexV7Bxu3WomOexADb7Iw2XstWEwNSw8I0FDaJspJcGqzqxv7hkqq/3Yi5hNQsPg7PZPWmisem/b9IO4VHHxA2GziaSHJIJyzU8IYgQPzw/TxseqiRFNQ6U4dhgQERhiite1N7pQaO+KTCvo02P4r/tMOH/So57fODo6YiZmPwyt6pDYEq/v7cbYNHvxoa3aw974VVhe9SEPCHAhg==
  • Authentication-results: dkim=none (message not signed) header.d=none;dmarc=none action=none header.from=suse.com;
  • Cc: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>, Wei Liu <wl@xxxxxxx>, Roger Pau Monné <roger.pau@xxxxxxxxxx>
  • Delivery-date: Thu, 02 Jun 2022 10:26:58 +0000
  • List-id: Xen developer discussion <xen-devel.lists.xenproject.org>

From: Artem Bityutskiy <artem.bityutskiy@xxxxxxxxxxxxxxx>

On Sapphire Rapids Xeon (SPR) the C1 and C1E states are basically mutually
exclusive - only one of them can be enabled. By default, 'intel_idle' driver
enables C1 and disables C1E. However, some users prefer to use C1E instead of
C1, because it saves more energy.

This patch adds a new module parameter ('preferred_cstates') for enabling C1E
and disabling C1. Here is the idea behind it.

1. This option has effect only for "mutually exclusive" C-states like C1 and
   C1E on SPR.
2. It does not have any effect on independent C-states, which do not require
   other C-states to be disabled (most states on most platforms as of today).
3. For mutually exclusive C-states, the 'intel_idle' driver always has a
   reasonable default, such as enabling C1 on SPR by default. On other
   platforms, the default may be different.
4. Users can override the default using the 'preferred_cstates' parameter.
5. The parameter accepts the preferred C-states bit-mask, similarly to the
   existing 'states_off' parameter.
6. This parameter is not limited to C1/C1E, and leaves room for supporting
   other mutually exclusive C-states, if they come in the future.

Today 'intel_idle' can only be compiled-in, which means that on SPR, in order
to disable C1 and enable C1E, users should boot with the following kernel
argument: intel_idle.preferred_cstates=4

Signed-off-by: Artem Bityutskiy <artem.bityutskiy@xxxxxxxxxxxxxxx>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
Origin: git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git 
da0e58c038e6

Enable C1E (if requested) not only on the BSP's socket / package. Alter
command line option to fit our model, and extend it to also accept
string form arguments.

Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
---
v2: Also accept string form arguments for command line option. Restore
    C1E-control related enum from Linux, despite our somewhat different
    use (and bigger code churn).

--- a/docs/misc/xen-command-line.pandoc
+++ b/docs/misc/xen-command-line.pandoc
@@ -1885,6 +1885,12 @@ paging controls access to usermode addre
 ### ple_window (Intel)
 > `= <integer>`
 
+### preferred-cstates (x86)
+> `= ( <integer> | List of ( C1 | C1E | C2 | ... )`
+
+This is a mask of C-states which are to be used preferably.  This option is
+applicable only on hardware were certain C-states are exclusive of one another.
+
 ### psr (Intel)
 > `= List of ( cmt:<boolean> | rmid_max:<integer> | cat:<boolean> | 
 > cos_max:<integer> | cdp:<boolean> )`
 
--- a/xen/arch/x86/cpu/mwait-idle.c
+++ b/xen/arch/x86/cpu/mwait-idle.c
@@ -82,10 +82,29 @@ boolean_param("mwait-idle", opt_mwait_id
 
 static unsigned int mwait_substates;
 
+/*
+ * Some platforms come with mutually exclusive C-states, so that if one is
+ * enabled, the other C-states must not be used. Example: C1 and C1E on
+ * Sapphire Rapids platform. This parameter allows for selecting the
+ * preferred C-states among the groups of mutually exclusive C-states - the
+ * selected C-states will be registered, the other C-states from the mutually
+ * exclusive group won't be registered. If the platform has no mutually
+ * exclusive C-states, this parameter has no effect.
+ */
+static unsigned int __ro_after_init preferred_states_mask;
+static char __initdata preferred_states[64];
+string_param("preferred-cstates", preferred_states);
+
 #define LAPIC_TIMER_ALWAYS_RELIABLE 0xFFFFFFFF
 /* Reliable LAPIC Timer States, bit 1 for C1 etc. Default to only C1. */
 static unsigned int lapic_timer_reliable_states = (1 << 1);
 
+enum c1e_promotion {
+       C1E_PROMOTION_PRESERVE,
+       C1E_PROMOTION_ENABLE,
+       C1E_PROMOTION_DISABLE
+};
+
 struct idle_cpu {
        const struct cpuidle_state *state_table;
 
@@ -95,7 +114,7 @@ struct idle_cpu {
         */
        unsigned long auto_demotion_disable_flags;
        bool byt_auto_demotion_disable_flag;
-       bool disable_promotion_to_c1e;
+       enum c1e_promotion c1e_promotion;
 };
 
 static const struct idle_cpu *icpu;
@@ -924,6 +943,15 @@ static void cf_check byt_auto_demotion_d
        wrmsrl(MSR_MC6_DEMOTION_POLICY_CONFIG, 0);
 }
 
+static void cf_check c1e_promotion_enable(void *dummy)
+{
+       uint64_t msr_bits;
+
+       rdmsrl(MSR_IA32_POWER_CTL, msr_bits);
+       msr_bits |= 0x2;
+       wrmsrl(MSR_IA32_POWER_CTL, msr_bits);
+}
+
 static void cf_check c1e_promotion_disable(void *dummy)
 {
        u64 msr_bits;
@@ -936,7 +964,7 @@ static void cf_check c1e_promotion_disab
 static const struct idle_cpu idle_cpu_nehalem = {
        .state_table = nehalem_cstates,
        .auto_demotion_disable_flags = NHM_C1_AUTO_DEMOTE | NHM_C3_AUTO_DEMOTE,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_atom = {
@@ -954,64 +982,64 @@ static const struct idle_cpu idle_cpu_li
 
 static const struct idle_cpu idle_cpu_snb = {
        .state_table = snb_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_byt = {
        .state_table = byt_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
        .byt_auto_demotion_disable_flag = true,
 };
 
 static const struct idle_cpu idle_cpu_cht = {
        .state_table = cht_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
        .byt_auto_demotion_disable_flag = true,
 };
 
 static const struct idle_cpu idle_cpu_ivb = {
        .state_table = ivb_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_ivt = {
        .state_table = ivt_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_hsw = {
        .state_table = hsw_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_bdw = {
        .state_table = bdw_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_skl = {
        .state_table = skl_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_skx = {
        .state_table = skx_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_icx = {
        .state_table = icx_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static struct idle_cpu __read_mostly idle_cpu_spr = {
        .state_table = spr_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_avn = {
        .state_table = avn_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_knl = {
@@ -1020,17 +1048,17 @@ static const struct idle_cpu idle_cpu_kn
 
 static const struct idle_cpu idle_cpu_bxt = {
        .state_table = bxt_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_dnv = {
        .state_table = dnv_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 static const struct idle_cpu idle_cpu_snr = {
        .state_table = snr_cstates,
-       .disable_promotion_to_c1e = true,
+       .c1e_promotion = C1E_PROMOTION_DISABLE,
 };
 
 #define ICPU(model, cpu) \
@@ -1241,6 +1269,25 @@ static void __init skx_idle_state_table_
 }
 
 /*
+ * spr_idle_state_table_update - Adjust Sapphire Rapids idle states table.
+ */
+static void __init spr_idle_state_table_update(void)
+{
+       /* Check if user prefers C1E over C1. */
+       if (preferred_states_mask & BIT(2, U)) {
+               if (preferred_states_mask & BIT(1, U))
+                       /* Both can't be enabled, stick to the defaults. */
+                       return;
+
+               spr_cstates[0].flags |= CPUIDLE_FLAG_DISABLED;
+               spr_cstates[1].flags &= ~CPUIDLE_FLAG_DISABLED;
+
+               /* Request enabling C1E using the "C1E promotion" bit. */
+               idle_cpu_spr.c1e_promotion = C1E_PROMOTION_ENABLE;
+       }
+}
+
+/*
  * mwait_idle_state_table_update()
  *
  * Update the default state_table for this CPU-id
@@ -1261,6 +1308,9 @@ static void __init mwait_idle_state_tabl
        case INTEL_FAM6_SKYLAKE_X:
                skx_idle_state_table_update();
                break;
+       case INTEL_FAM6_SAPPHIRERAPIDS_X:
+               spr_idle_state_table_update();
+               break;
        }
 }
 
@@ -1268,6 +1318,7 @@ static int __init mwait_idle_probe(void)
 {
        unsigned int eax, ebx, ecx;
        const struct x86_cpu_id *id = x86_match_cpu(intel_idle_ids);
+       const char *str;
 
        if (!id) {
                pr_debug(PREFIX "does not run on family %d model %d\n",
@@ -1309,6 +1360,39 @@ static int __init mwait_idle_probe(void)
        pr_debug(PREFIX "lapic_timer_reliable_states %#x\n",
                 lapic_timer_reliable_states);
 
+       str = preferred_states;
+       if (isdigit(str[0]))
+               preferred_states_mask = simple_strtoul(str, &str, 0);
+       else if (str[0])
+       {
+               const char *ss;
+
+               do {
+                       const struct cpuidle_state *state = icpu->state_table;
+                       unsigned int bit = 1;
+
+                       ss = strchr(str, ',');
+                       if (!ss)
+                               ss = strchr(str, '\0');
+
+                       for (; state->name[0]; ++state) {
+                               bit <<= 1;
+                               if (!cmdline_strcmp(str, state->name)) {
+                                       preferred_states_mask |= bit;
+                                       break;
+                               }
+                       }
+                       if (!state->name[0])
+                               break;
+
+                       str = ss + 1;
+           } while (*ss);
+
+           str -= str == ss + 1;
+       }
+       if (str[0])
+               printk("unrecognized \"preferred-cstates=%s\"\n", str);
+
        mwait_idle_state_table_update();
 
        return 0;
@@ -1400,8 +1484,18 @@ static int cf_check mwait_idle_cpu_init(
        if (icpu->byt_auto_demotion_disable_flag)
                on_selected_cpus(cpumask_of(cpu), byt_auto_demotion_disable, 
NULL, 1);
 
-       if (icpu->disable_promotion_to_c1e)
+       switch (icpu->c1e_promotion) {
+       case C1E_PROMOTION_DISABLE:
                on_selected_cpus(cpumask_of(cpu), c1e_promotion_disable, NULL, 
1);
+               break;
+
+       case C1E_PROMOTION_ENABLE:
+               on_selected_cpus(cpumask_of(cpu), c1e_promotion_enable, NULL, 
1);
+               break;
+
+       case C1E_PROMOTION_PRESERVE:
+               break;
+       }
 
        return NOTIFY_DONE;
 }




 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.