[dpdk-dev] [PATCH 13/28] eal/arm64: override I/O device read/write access for arm64
Jianbo Liu
jianbo.liu at linaro.org
Thu Dec 15 11:17:32 CET 2016
On 15 December 2016 at 18:04, Jerin Jacob
<jerin.jacob at caviumnetworks.com> wrote:
> On Thu, Dec 15, 2016 at 05:53:05PM +0800, Jianbo Liu wrote:
>> On 14 December 2016 at 09:55, Jerin Jacob
>> <jerin.jacob at caviumnetworks.com> wrote:
>> > Override the generic I/O device memory read/write access and implement it
>> > using armv8 instructions for arm64.
>> >
>> > Signed-off-by: Jerin Jacob <jerin.jacob at caviumnetworks.com>
>> > ---
>> > lib/librte_eal/common/include/arch/arm/rte_io.h | 4 +
>> > lib/librte_eal/common/include/arch/arm/rte_io_64.h | 183 +++++++++++++++++++++
>> > 2 files changed, 187 insertions(+)
>> > create mode 100644 lib/librte_eal/common/include/arch/arm/rte_io_64.h
>> >
>> > diff --git a/lib/librte_eal/common/include/arch/arm/rte_io.h b/lib/librte_eal/common/include/arch/arm/rte_io.h
>> > index 74c1f2c..9593b42 100644
>> > --- a/lib/librte_eal/common/include/arch/arm/rte_io.h
>> > +++ b/lib/librte_eal/common/include/arch/arm/rte_io.h
>> > @@ -38,7 +38,11 @@
>> > extern "C" {
>> > #endif
>> >
>> > +#ifdef RTE_ARCH_64
>> > +#include "rte_io_64.h"
>> > +#else
>> > #include "generic/rte_io.h"
>> > +#endif
>> >
>> > #ifdef __cplusplus
>> > }
>> > diff --git a/lib/librte_eal/common/include/arch/arm/rte_io_64.h b/lib/librte_eal/common/include/arch/arm/rte_io_64.h
>> > new file mode 100644
>> > index 0000000..09e7a89
>> > --- /dev/null
>> > +++ b/lib/librte_eal/common/include/arch/arm/rte_io_64.h
>> > @@ -0,0 +1,183 @@
>> > +/*
>> > + * BSD LICENSE
>> > + *
>> > + * Copyright (C) Cavium networks Ltd. 2016.
>> > + *
>> > + * Redistribution and use in source and binary forms, with or without
>> > + * modification, are permitted provided that the following conditions
>> > + * are met:
>> > + *
>> > + * * Redistributions of source code must retain the above copyright
>> > + * notice, this list of conditions and the following disclaimer.
>> > + * * Redistributions in binary form must reproduce the above copyright
>> > + * notice, this list of conditions and the following disclaimer in
>> > + * the documentation and/or other materials provided with the
>> > + * distribution.
>> > + * * Neither the name of Cavium networks nor the names of its
>> > + * contributors may be used to endorse or promote products derived
>> > + * from this software without specific prior written permission.
>> > + *
>> > + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
>> > + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
>> > + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
>> > + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
>> > + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
>> > + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
>> > + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
>> > + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
>> > + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
>> > + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
>> > + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
>> > + */
>> > +
>> > +#ifndef _RTE_IO_ARM64_H_
>> > +#define _RTE_IO_ARM64_H_
>> > +
>> > +#ifdef __cplusplus
>> > +extern "C" {
>> > +#endif
>> > +
>> > +#include <stdint.h>
>> > +
>> > +#define RTE_OVERRIDE_IO_H
>> > +
>> > +#include "generic/rte_io.h"
>> > +#include "rte_atomic_64.h"
>> > +
>> > +static inline __attribute__((always_inline)) uint8_t
>> > +__rte_arm64_readb(const volatile void *addr)
>> > +{
>> > + uint8_t val;
>> > +
>> > + asm volatile(
>> > + "ldrb %w[val], [%x[addr]]"
>> > + : [val] "=r" (val)
>> > + : [addr] "r" (addr));
>> > + return val;
>> > +}
>> > +
>> > +static inline __attribute__((always_inline)) uint16_t
>> > +__rte_arm64_readw(const volatile void *addr)
>> > +{
>> > + uint16_t val;
>> > +
>> > + asm volatile(
>> > + "ldrh %w[val], [%x[addr]]"
>> > + : [val] "=r" (val)
>> > + : [addr] "r" (addr));
>> > + return val;
>> > +}
>> > +
>> > +static inline __attribute__((always_inline)) uint32_t
>> > +__rte_arm64_readl(const volatile void *addr)
>> > +{
>> > + uint32_t val;
>> > +
>> > + asm volatile(
>> > + "ldr %w[val], [%x[addr]]"
>> > + : [val] "=r" (val)
>> > + : [addr] "r" (addr));
>> > + return val;
>> > +}
>> > +
>> > +static inline __attribute__((always_inline)) uint64_t
>> > +__rte_arm64_readq(const volatile void *addr)
>> > +{
>> > + uint64_t val;
>> > +
>> > + asm volatile(
>> > + "ldr %x[val], [%x[addr]]"
>> > + : [val] "=r" (val)
>> > + : [addr] "r" (addr));
>> > + return val;
>> > +}
>> > +
>> > +static inline __attribute__((always_inline)) void
>> > +__rte_arm64_writeb(uint8_t val, volatile void *addr)
>> > +{
>> > + asm volatile(
>> > + "strb %w[val], [%x[addr]]"
>> > + :
>> > + : [val] "r" (val), [addr] "r" (addr));
>> > +}
>> > +
>> > +static inline __attribute__((always_inline)) void
>> > +__rte_arm64_writew(uint16_t val, volatile void *addr)
>> > +{
>> > + asm volatile(
>> > + "strh %w[val], [%x[addr]]"
>> > + :
>> > + : [val] "r" (val), [addr] "r" (addr));
>> > +}
>> > +
>> > +static inline __attribute__((always_inline)) void
>> > +__rte_arm64_writel(uint32_t val, volatile void *addr)
>> > +{
>> > + asm volatile(
>> > + "str %w[val], [%x[addr]]"
>> > + :
>> > + : [val] "r" (val), [addr] "r" (addr));
>> > +}
>> > +
>> > +static inline __attribute__((always_inline)) void
>> > +__rte_arm64_writeq(uint64_t val, volatile void *addr)
>> > +{
>> > + asm volatile(
>> > + "str %x[val], [%x[addr]]"
>> > + :
>> > + : [val] "r" (val), [addr] "r" (addr));
>> > +}
>>
>> I'm not quite sure about these overridings. Can you explain the
>> benefit to do so?
>
> Better to be native if there is option. That all. Do you see any issue?
> or what is the real concern?
>
I think it's the same as the generic c version after compiling. Am I right?
If there is no apparent benefit, I don't think we need the overriding.
More information about the dev
mailing list