Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Atish Patra <atishp@rivosinc.com>
To:     linux-kernel@vger.kernel.org
Cc:     Atish Patra <atishp@rivosinc.com>,
        Anup Patel <anup@brainfault.org>,
        Albert Ou <aou@eecs.berkeley.edu>,
        Atish Patra <atishp@atishpatra.org>,
        Daniel Lezcano <daniel.lezcano@linaro.org>,
        Guo Ren <guoren@kernel.org>, Heiko Stuebner <heiko@sntech.de>,
        kvm-riscv@lists.infradead.org, kvm@vger.kernel.org,
        linux-riscv@lists.infradead.org,
        Palmer Dabbelt <palmer@dabbelt.com>,
        Paolo Bonzini <pbonzini@redhat.com>,
        Paul Walmsley <paul.walmsley@sifive.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Tsukasa OI <research_trasio@irq.a4lg.com>,
        Wei Fu <wefu@redhat.com>
Subject: [PATCH v6 3/4] RISC-V: Prefer sstc extension if available
Date:   Thu, 21 Jul 2022 11:12:11 -0700
Message-Id: <20220721181212.3705138-4-atishp@rivosinc.com>
In-Reply-To: <20220721181212.3705138-1-atishp@rivosinc.com>
References: <20220721181212.3705138-1-atishp@rivosinc.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

RISC-V ISA has sstc extension which allows updating the next clock event
via a CSR (stimecmp) instead of an SBI call. This should happen dynamically
if sstc extension is available. Otherwise, it will fallback to SBI call
to maintain backward compatibility.

Reviewed-by: Anup Patel <anup@brainfault.org>
Signed-off-by: Atish Patra <atishp@rivosinc.com>
---
 drivers/clocksource/timer-riscv.c | 24 +++++++++++++++++++++++-
 1 file changed, 23 insertions(+), 1 deletion(-)

diff --git a/drivers/clocksource/timer-riscv.c b/drivers/clocksource/timer-riscv.c
index 593d5a957b69..3f100fb53d82 100644
--- a/drivers/clocksource/timer-riscv.c
+++ b/drivers/clocksource/timer-riscv.c
@@ -7,6 +7,9 @@
  * either be read from the "time" and "timeh" CSRs, and can use the SBI to
  * setup events, or directly accessed using MMIO registers.
  */
+
+#define pr_fmt(fmt) "riscv-timer: " fmt
+
 #include <linux/clocksource.h>
 #include <linux/clockchips.h>
 #include <linux/cpu.h>
@@ -23,11 +26,24 @@
 #include <asm/sbi.h>
 #include <asm/timex.h>
 
+static DEFINE_STATIC_KEY_FALSE(riscv_sstc_available);
+
 static int riscv_clock_next_event(unsigned long delta,
 		struct clock_event_device *ce)
 {
+	u64 next_tval = get_cycles64() + delta;
+
 	csr_set(CSR_IE, IE_TIE);
-	sbi_set_timer(get_cycles64() + delta);
+	if (static_branch_likely(&riscv_sstc_available)) {
+#if defined(CONFIG_32BIT)
+		csr_write(CSR_STIMECMP, next_tval & 0xFFFFFFFF);
+		csr_write(CSR_STIMECMPH, next_tval >> 32);
+#else
+		csr_write(CSR_STIMECMP, next_tval);
+#endif
+	} else
+		sbi_set_timer(next_tval);
+
 	return 0;
 }
 
@@ -165,6 +181,12 @@ static int __init riscv_timer_init_dt(struct device_node *n)
 	if (error)
 		pr_err("cpu hp setup state failed for RISCV timer [%d]\n",
 		       error);
+
+	if (riscv_isa_extension_available(NULL, SSTC)) {
+		pr_info("Timer interrupt in S-mode is available via sstc extension\n");
+		static_branch_enable(&riscv_sstc_available);
+	}
+
 	return error;
 }
 
-- 
2.25.1