[PATCH v3 1/2] cache: Add StarFive StarLink cache management for StarFive JH8100

Joshua Yeong joshua.yeong at starfivetech.com
Wed Apr 24 07:07:54 PDT 2024


Emil Renner Berthing wrote:
> Joshua Yeong wrote:
> > Add StarFive Starlink cache management driver for
> > JH8100 SoC. This driver enables RISC-V non-standard cache operation on
> > JH8100 that does not support Zicbom extension instructions.
> >
> > Signed-off-by: Joshua Yeong <joshua.yeong at starfivetech.com>
> > ---
> >  drivers/cache/Kconfig                   |   9 ++
> >  drivers/cache/Makefile                  |   5 +-
> >  drivers/cache/starfive_starlink_cache.c | 135
> > ++++++++++++++++++++++++
> >  3 files changed, 147 insertions(+), 2 deletions(-)  create mode
> > 100644 drivers/cache/starfive_starlink_cache.c
> >
> > diff --git a/drivers/cache/Kconfig b/drivers/cache/Kconfig index
> > 9345ce4976d7..9181cd391f53 100644
> > --- a/drivers/cache/Kconfig
> > +++ b/drivers/cache/Kconfig
> > @@ -14,4 +14,13 @@ config SIFIVE_CCACHE
> >  	help
> >  	  Support for the composable cache controller on SiFive platforms.
> >
> > +config STARFIVE_STARLINK_CACHE
> > +	bool "StarFive StarLink Cache controller"
> > +	depends on RISCV
> > +	depends on ARCH_STARFIVE
> > +	select RISCV_DMA_NONCOHERENT
> > +	select RISCV_NONSTANDARD_CACHE_OPS
> > +	help
> > +	  Support for the StarLink cache controller on StarFive platforms.
> 
> This is a bit misleading. The JH71x0s don't have this. It's only on the JH8100 so
> far, and hopefully later SoCs will just implement RISC-V standards for this.
> So maybe something like
> 
> "Support for the StarLink cache controller on the StarFive JH8100 SoC."
> 

Hi Emil,

The StarLink-500 cache controller is not designed exclusively for JH8100 SoC. 
While it is true that it currently exists on the StarFive platform, CPU/SoC 
that does not come with Zicbom extensions supported would need to rely 
on this cache drive to do cache management operations. I think we don’t need
to mentioned 'JH8100 SoC' here.

> > +
> >  endmenu
> > diff --git a/drivers/cache/Makefile b/drivers/cache/Makefile index
> > 7657cff3bd6c..55c5e851034d 100644
> > --- a/drivers/cache/Makefile
> > +++ b/drivers/cache/Makefile
> > @@ -1,4 +1,5 @@
> >  # SPDX-License-Identifier: GPL-2.0
> >
> > -obj-$(CONFIG_AX45MP_L2_CACHE)	+= ax45mp_cache.o
> > -obj-$(CONFIG_SIFIVE_CCACHE)	+= sifive_ccache.o
> > +obj-$(CONFIG_AX45MP_L2_CACHE)		+= ax45mp_cache.o
> > +obj-$(CONFIG_SIFIVE_CCACHE)		+= sifive_ccache.o
> > +obj-$(CONFIG_STARFIVE_STARLINK_CACHE)	+= starfive_starlink_cache.o
> > diff --git a/drivers/cache/starfive_starlink_cache.c
> > b/drivers/cache/starfive_starlink_cache.c
> > new file mode 100644
> > index 000000000000..2f1fa6a923ee
> > --- /dev/null
> > +++ b/drivers/cache/starfive_starlink_cache.c
> > @@ -0,0 +1,135 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Cache Management Operations for StarFive's Starlink cache
> > +controller
> > + *
> > + * Copyright (C) 2024 Shanghai StarFive Technology Co., Ltd.
> > + *
> > + * Author: Joshua Yeong <joshua.yeong at starfivetech.com>  */
> > +
> > +#include <linux/bitfield.h>
> > +#include <linux/cacheflush.h>
> > +#include <linux/iopoll.h>
> > +#include <linux/of_address.h>
> > +
> > +#include <asm/dma-noncoherent.h>
> > +
> > +#define STARLINK_CACHE_FLUSH_START_ADDR			0x0
> > +#define STARLINK_CACHE_FLUSH_END_ADDR			0x8
> > +#define STARLINK_CACHE_FLUSH_CTL			0x10
> > +#define STARLINK_CACHE_ALIGN				0x40
> > +
> > +#define STARLINK_CACHE_ADDRESS_RANGE_MASK
> 	GENMASK(39, 0)
> > +#define STARLINK_CACHE_FLUSH_CTL_MODE_MASK
> 	GENMASK(2, 1)
> > +#define STARLINK_CACHE_FLUSH_CTL_ENABLE_MASK		BIT(0)
> > +
> > +#define STARLINK_CACHE_FLUSH_CTL_CLEAN_INVALIDATE	0
> > +#define STARLINK_CACHE_FLUSH_CTL_MAKE_INVALIDATE	1
> > +#define STARLINK_CACHE_FLUSH_CTL_CLEAN_SHARED		2
> > +#define STARLINK_CACHE_FLUSH_POLL_DELAY_US		1
> > +#define STARLINK_CACHE_FLUSH_TIMEOUT_US
> 	5000000
> > +
> > +struct starlink_cache_priv {
> > +	void __iomem *base_addr;
> > +};
> > +
> > +static struct starlink_cache_priv starlink_cache_priv;
> 
> This could just be
> 
> static void __iomem *starlink_cache_base;
> 
> > +
> > +static void starlink_cache_flush_complete(void)
> > +{
> > +	volatile void __iomem *ctl = starlink_cache_priv.base_addr +
> > +				     STARLINK_CACHE_FLUSH_CTL;
> > +	u64 v;
> > +	int ret;
> > +
> > +	ret = readq_poll_timeout_atomic(ctl, v, !(v &
> STARLINK_CACHE_FLUSH_CTL_ENABLE_MASK),
> > +
> 	STARLINK_CACHE_FLUSH_POLL_DELAY_US,
> > +
> 	STARLINK_CACHE_FLUSH_TIMEOUT_US);
> > +	if (ret)
> > +		WARN(1, "StarFive Starlink cache flush operation
> timeout\n"); }
> > +
> > +static void starlink_cache_dma_cache_wback(phys_addr_t paddr,
> > +unsigned long size) {
> > +	writeq(FIELD_PREP(STARLINK_CACHE_ADDRESS_RANGE_MASK,
> paddr),
> > +	       starlink_cache_priv.base_addr +
> STARLINK_CACHE_FLUSH_START_ADDR);
> > +	writeq(FIELD_PREP(STARLINK_CACHE_ADDRESS_RANGE_MASK,
> paddr + size),
> > +	       starlink_cache_priv.base_addr +
> > +STARLINK_CACHE_FLUSH_END_ADDR);
> > +
> > +	mb();
> > +	writeq(FIELD_PREP(STARLINK_CACHE_FLUSH_CTL_MODE_MASK,
> > +			  STARLINK_CACHE_FLUSH_CTL_CLEAN_SHARED),
> > +	       starlink_cache_priv.base_addr + STARLINK_CACHE_FLUSH_CTL);
> > +
> > +	starlink_cache_flush_complete();
> > +}
> > +
> > +static void starlink_cache_dma_cache_invalidate(phys_addr_t paddr,
> > +unsigned long size) {
> > +	writeq(FIELD_PREP(STARLINK_CACHE_ADDRESS_RANGE_MASK,
> paddr),
> > +	       starlink_cache_priv.base_addr +
> STARLINK_CACHE_FLUSH_START_ADDR);
> > +	writeq(FIELD_PREP(STARLINK_CACHE_ADDRESS_RANGE_MASK,
> paddr + size),
> > +	       starlink_cache_priv.base_addr +
> > +STARLINK_CACHE_FLUSH_END_ADDR);
> > +
> > +	mb();
> > +	writeq(FIELD_PREP(STARLINK_CACHE_FLUSH_CTL_MODE_MASK,
> > +			  STARLINK_CACHE_FLUSH_CTL_MAKE_INVALIDATE),
> > +	       starlink_cache_priv.base_addr + STARLINK_CACHE_FLUSH_CTL);
> > +
> > +	starlink_cache_flush_complete();
> > +}
> > +
> > +static void starlink_cache_dma_cache_wback_inv(phys_addr_t paddr,
> > +unsigned long size) {
> > +	writeq(FIELD_PREP(STARLINK_CACHE_ADDRESS_RANGE_MASK,
> paddr),
> > +	       starlink_cache_priv.base_addr +
> STARLINK_CACHE_FLUSH_START_ADDR);
> > +	writeq(FIELD_PREP(STARLINK_CACHE_ADDRESS_RANGE_MASK,
> paddr + size),
> > +	       starlink_cache_priv.base_addr +
> > +STARLINK_CACHE_FLUSH_END_ADDR);
> > +
> > +	mb();
> > +	writeq(FIELD_PREP(STARLINK_CACHE_FLUSH_CTL_MODE_MASK,
> > +			  STARLINK_CACHE_FLUSH_CTL_CLEAN_INVALIDATE),
> > +	       starlink_cache_priv.base_addr + STARLINK_CACHE_FLUSH_CTL);
> > +
> > +	starlink_cache_flush_complete();
> > +}
> > +
> > +static const struct riscv_nonstd_cache_ops starlink_cache_ops = {
> > +	.wback = &starlink_cache_dma_cache_wback,
> > +	.inv = &starlink_cache_dma_cache_invalidate,
> > +	.wback_inv = &starlink_cache_dma_cache_wback_inv,
> > +};
> > +
> > +static const struct of_device_id starlink_cache_ids[] = {
> > +	{ .compatible = "starfive,jh8100-starlink-cache" },
> > +	{ /* sentinel */ }
> > +};
> > +
> > +static int __init starlink_cache_init(void) {
> > +	struct device_node *np;
> > +	u32 block_size = 0;
> 
> You return early if of_property_read_u32() fails, so this doesn't need to be
> initialized.
> 
> > +	int ret;
> > +
> > +	np = of_find_matching_node(NULL, starlink_cache_ids);
> > +	if (!of_device_is_available(np))
> > +		return -ENODEV;
> > +
> > +	ret = of_property_read_u32(np, "cache-block-size", &block_size);
> > +	if (ret)
> > +		return ret;
> > +
> > +	if (block_size % STARLINK_CACHE_ALIGN)
> > +		return -EINVAL;
> > +
> > +	starlink_cache_priv.base_addr = of_iomap(np, 0);
> > +	if (!starlink_cache_priv.base_addr)
> > +		return -ENOMEM;
> > +
> > +	riscv_cbom_block_size = block_size;
> > +	riscv_noncoherent_supported();
> > +	riscv_noncoherent_register_cache_ops(&starlink_cache_ops);
> > +
> > +	return 0;
> > +}
> > +early_initcall(starlink_cache_init);
> 
> The sifive driver gets away with arch_initcall() here. Any particular reason you
> need this earlier?

I will change it to arch_initcall(). Appreciate for the comments.

Thank you.

Regards,
Joshua

> 
> > --
> > 2.25.1
> >
> >
> > _______________________________________________
> > linux-riscv mailing list
> > linux-riscv at lists.infradead.org
> > http://lists.infradead.org/mailman/listinfo/linux-riscv


More information about the linux-riscv mailing list