DPDK patches and discussions
 help / color / mirror / Atom feed
From: Jan Viktorin <viktorin@rehivetech.com>
To: david.marchand@6wind.com, David Hunt <david.hunt@intel.com>,
	Thomas Monjalon <thomas.monjalon@6wind.com>,
	Jerin Jacob <jerin.jacob@caviumnetworks.com>
Cc: dev@dpdk.org
Subject: [dpdk-dev] [PATCH v6 04/15] eal/arm: implement rdtsc by PMU or clock_gettime
Date: Tue,  3 Nov 2015 00:47:17 +0100	[thread overview]
Message-ID: <1446508048-16744-5-git-send-email-viktorin@rehivetech.com> (raw)
In-Reply-To: <1446508048-16744-1-git-send-email-viktorin@rehivetech.com>

Enable to choose a preferred way to read timer based on the
configuration entry CONFIG_RTE_ARM_EAL_RDTSC_USE_PMU.
It requires a kernel module that is not included to work.

Based on the patch by David Hunt and Armuta Zende:

  lib: added support for armv7 architecture

Signed-off-by: Jan Viktorin <viktorin@rehivetech.com>
Signed-off-by: Amruta Zende <amruta.zende@intel.com>
Signed-off-by: David Hunt <david.hunt@intel.com>
---
 .../common/include/arch/arm/rte_cycles_32.h        | 38 +++++++++++++++++++++-
 1 file changed, 37 insertions(+), 1 deletion(-)

diff --git a/lib/librte_eal/common/include/arch/arm/rte_cycles_32.h b/lib/librte_eal/common/include/arch/arm/rte_cycles_32.h
index 755cc4a..6c6098e 100644
--- a/lib/librte_eal/common/include/arch/arm/rte_cycles_32.h
+++ b/lib/librte_eal/common/include/arch/arm/rte_cycles_32.h
@@ -54,8 +54,14 @@ extern "C" {
  * @return
  *   The time base for this lcore.
  */
+#ifndef CONFIG_RTE_ARM_EAL_RDTSC_USE_PMU
+
+/**
+ * This call is easily portable to any ARM architecture, however,
+ * it may be damn slow and inprecise for some tasks.
+ */
 static inline uint64_t
-rte_rdtsc(void)
+__rte_rdtsc_syscall(void)
 {
 	struct timespec val;
 	uint64_t v;
@@ -67,6 +73,36 @@ rte_rdtsc(void)
 	v += (uint64_t) val.tv_nsec;
 	return v;
 }
+#define rte_rdtsc __rte_rdtsc_syscall
+
+#else
+
+/**
+ * This function requires to configure the PMCCNTR and enable
+ * userspace access to it:
+ *
+ *      asm volatile("mcr p15, 0, %0, c9, c14, 0" : : "r"(1));
+ *      asm volatile("mcr p15, 0, %0, c9, c12, 0" : : "r"(29));
+ *      asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r"(0x8000000f));
+ *
+ * which is possible only from the priviledged mode (kernel space).
+ */
+static inline uint64_t
+__rte_rdtsc_pmccntr(void)
+{
+	unsigned tsc;
+	uint64_t final_tsc;
+
+	/* Read PMCCNTR */
+	asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r"(tsc));
+	/* 1 tick = 64 clocks */
+	final_tsc = ((uint64_t)tsc) << 6;
+
+	return (uint64_t)final_tsc;
+}
+#define rte_rdtsc __rte_rdtsc_pmccntr
+
+#endif /* RTE_ARM_EAL_RDTSC_USE_PMU */
 
 static inline uint64_t
 rte_rdtsc_precise(void)
-- 
2.6.2

  parent reply	other threads:[~2015-11-02 23:50 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-11-02 23:47 [dpdk-dev] [PATCH v6 00/15] Support ARMv7 architecture Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 01/15] eal/arm: atomic operations for ARM Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 02/15] eal/arm: byte order " Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 03/15] eal/arm: cpu cycle " Jan Viktorin
2015-11-02 23:47 ` Jan Viktorin [this message]
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 05/15] eal/arm: prefetch " Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 06/15] eal/arm: spinlock operations for ARM (without HTM) Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 07/15] eal/arm: vector memcpy for ARM Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 08/15] eal/arm: use vector memcpy only when NEON is enabled Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 09/15] eal/arm: cpu flag checks for ARM Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 10/15] eal/arm: detect arm architecture in cpu flags Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 11/15] eal/arm: rwlock support for ARM Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 12/15] eal/arm: add very incomplete rte_vect Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 13/15] gcc/arm: avoid alignment errors to break build Jan Viktorin
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 14/15] mk: Introduce ARMv7 architecture Jan Viktorin
2015-11-03 10:16   ` Hunt, David
2015-11-03 10:24     ` Bruce Richardson
2015-11-03 11:18       ` Thomas Monjalon
2015-11-03 10:27     ` Jan Viktorin
2015-11-03 11:08       ` Hunt, David
2015-11-02 23:47 ` [dpdk-dev] [PATCH v6 15/15] maintainers: claim responsibility for ARMv7 Jan Viktorin
2015-11-03  4:49 ` [dpdk-dev] [PATCH v6 00/15] Support ARMv7 architecture Jerin Jacob
2015-11-03 11:33   ` Hunt, David
2015-11-03 12:32     ` Jacob,  Jerin
2015-11-16 21:33 ` David Marchand
2015-11-18 21:45   ` Thomas Monjalon

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1446508048-16744-5-git-send-email-viktorin@rehivetech.com \
    --to=viktorin@rehivetech.com \
    --cc=david.hunt@intel.com \
    --cc=david.marchand@6wind.com \
    --cc=dev@dpdk.org \
    --cc=jerin.jacob@caviumnetworks.com \
    --cc=thomas.monjalon@6wind.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).