Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752396AbaLOOq6 (ORCPT ); Mon, 15 Dec 2014 09:46:58 -0500 Received: from youngberry.canonical.com ([91.189.89.112]:38730 "EHLO youngberry.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752875AbaLOO2f (ORCPT ); Mon, 15 Dec 2014 09:28:35 -0500 From: Luis Henriques To: linux-kernel@vger.kernel.org, stable@vger.kernel.org, kernel-team@lists.ubuntu.com Cc: Markos Chandras , linux-mips@linux-mips.org, Ralf Baechle , Luis Henriques Subject: [PATCH 3.16.y-ckt 107/168] MIPS: cpu-probe: Set the FTLB probability bit on supported cores Date: Mon, 15 Dec 2014 14:26:01 +0000 Message-Id: <1418653622-21105-108-git-send-email-luis.henriques@canonical.com> X-Mailer: git-send-email 2.1.3 In-Reply-To: <1418653622-21105-1-git-send-email-luis.henriques@canonical.com> References: <1418653622-21105-1-git-send-email-luis.henriques@canonical.com> X-Extended-Stable: 3.16 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 3.16.7-ckt3 -stable review patch. If anyone has any objections, please let me know. ------------------ From: Markos Chandras commit cf0a8aa0226da5de88011e7f30eff22a894b2f49 upstream. Make use of the Config6/FLTBP bit to set the probability of a TLBWR instruction to hit the FTLB or the VTLB. A value of 0 (which may be the default value on certain cores, such as proAptiv or P5600) means that a TLBWR instruction will never hit the VTLB which leads to performance limitations since it effectively decreases the number of available TLB slots. Signed-off-by: Markos Chandras Reviewed-by: James Hogan Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/8368/ Signed-off-by: Ralf Baechle Signed-off-by: Luis Henriques --- arch/mips/include/asm/mipsregs.h | 2 ++ arch/mips/kernel/cpu-probe.c | 33 ++++++++++++++++++++++++++++++++- 2 files changed, 34 insertions(+), 1 deletion(-) diff --git a/arch/mips/include/asm/mipsregs.h b/arch/mips/include/asm/mipsregs.h index 98e9754a4b6b..6ad0208b50fb 100644 --- a/arch/mips/include/asm/mipsregs.h +++ b/arch/mips/include/asm/mipsregs.h @@ -660,6 +660,8 @@ #define MIPS_CONF6_SYND (_ULCAST_(1) << 13) /* proAptiv FTLB on/off bit */ #define MIPS_CONF6_FTLBEN (_ULCAST_(1) << 15) +/* FTLB probability bits */ +#define MIPS_CONF6_FTLBP_SHIFT (16) #define MIPS_CONF7_WII (_ULCAST_(1) << 31) diff --git a/arch/mips/kernel/cpu-probe.c b/arch/mips/kernel/cpu-probe.c index d74f957c561e..d5006d23ca7b 100644 --- a/arch/mips/kernel/cpu-probe.c +++ b/arch/mips/kernel/cpu-probe.c @@ -179,6 +179,32 @@ static void set_isa(struct cpuinfo_mips *c, unsigned int isa) static char unknown_isa[] = KERN_ERR \ "Unsupported ISA type, c0.config0: %d."; +static unsigned int calculate_ftlb_probability(struct cpuinfo_mips *c) +{ + + unsigned int probability = c->tlbsize / c->tlbsizevtlb; + + /* + * 0 = All TLBWR instructions go to FTLB + * 1 = 15:1: For every 16 TBLWR instructions, 15 go to the + * FTLB and 1 goes to the VTLB. + * 2 = 7:1: As above with 7:1 ratio. + * 3 = 3:1: As above with 3:1 ratio. + * + * Use the linear midpoint as the probability threshold. + */ + if (probability >= 12) + return 1; + else if (probability >= 6) + return 2; + else + /* + * So FTLB is less than 4 times bigger than VTLB. + * A 3:1 ratio can still be useful though. + */ + return 3; +} + static void set_ftlb_enable(struct cpuinfo_mips *c, int enable) { unsigned int config6; @@ -189,9 +215,14 @@ static void set_ftlb_enable(struct cpuinfo_mips *c, int enable) case CPU_P5600: /* proAptiv & related cores use Config6 to enable the FTLB */ config6 = read_c0_config6(); + /* Clear the old probability value */ + config6 &= ~(3 << MIPS_CONF6_FTLBP_SHIFT); if (enable) /* Enable FTLB */ - write_c0_config6(config6 | MIPS_CONF6_FTLBEN); + write_c0_config6(config6 | + (calculate_ftlb_probability(c) + << MIPS_CONF6_FTLBP_SHIFT) + | MIPS_CONF6_FTLBEN); else /* Disable FTLB */ write_c0_config6(config6 & ~MIPS_CONF6_FTLBEN); -- 2.1.3 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/