1
linux/arch/riscv/mm/dma-noncoherent.c

158 lines
3.7 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-only
/*
* RISC-V specific functions to support DMA for non-coherent devices
*
* Copyright (c) 2021 Western Digital Corporation or its affiliates.
*/
#include <linux/dma-direct.h>
#include <linux/dma-map-ops.h>
#include <linux/mm.h>
#include <asm/cacheflush.h>
#include <asm/dma-noncoherent.h>
static bool noncoherent_supported __ro_after_init;
riscv: allow kmalloc() caches aligned to the smallest value Currently, riscv defines ARCH_DMA_MINALIGN as L1_CACHE_BYTES, I.E 64Bytes, if CONFIG_RISCV_DMA_NONCOHERENT=y. To support unified kernel Image, usually we have to enable CONFIG_RISCV_DMA_NONCOHERENT, thus it brings some bad effects to coherent platforms: Firstly, it wastes memory, kmalloc-96, kmalloc-32, kmalloc-16 and kmalloc-8 slab caches don't exist any more, they are replaced with either kmalloc-128 or kmalloc-64. Secondly, larger than necessary kmalloc aligned allocations results in unnecessary cache/TLB pressure. This issue also exists on arm64 platforms. From last year, Catalin tried to solve this issue by decoupling ARCH_KMALLOC_MINALIGN from ARCH_DMA_MINALIGN, limiting kmalloc() minimum alignment to dma_get_cache_alignment() and replacing ARCH_KMALLOC_MINALIGN usage in various drivers with ARCH_DMA_MINALIGN etc.[1] One fact we can make use of for riscv: if the CPU doesn't support ZICBOM or T-HEAD CMO, we know the platform is coherent. Based on Catalin's work and above fact, we can easily solve the kmalloc align issue for riscv: we can override dma_get_cache_alignment(), then let it return ARCH_DMA_MINALIGN at the beginning and return 1 once we know the underlying HW neither supports ZICBOM nor supports T-HEAD CMO. So what about if the CPU supports ZICBOM or T-HEAD CMO, but all the devices are dma coherent? Well, we use ARCH_DMA_MINALIGN as the kmalloc minimum alignment, nothing changed in this case. This case can be improved in the future. After this patch, a simple test of booting to a small buildroot rootfs on qemu shows: kmalloc-96 5041 5041 96 ... kmalloc-64 9606 9606 64 ... kmalloc-32 5128 5128 32 ... kmalloc-16 7682 7682 16 ... kmalloc-8 10246 10246 8 ... So we save about 1268KB memory. The saving will be much larger in normal OS env on real HW platforms. Link: https://lore.kernel.org/linux-arm-kernel/20230524171904.3967031-1-catalin.marinas@arm.com/ [1] Signed-off-by: Jisheng Zhang <jszhang@kernel.org> Reviewed-by: Conor Dooley <conor.dooley@microchip.com> Link: https://lore.kernel.org/r/20230718152214.2907-2-jszhang@kernel.org Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
2023-07-18 08:22:13 -07:00
int dma_cache_alignment __ro_after_init = ARCH_DMA_MINALIGN;
EXPORT_SYMBOL_GPL(dma_cache_alignment);
static inline void arch_dma_cache_wback(phys_addr_t paddr, size_t size)
{
void *vaddr = phys_to_virt(paddr);
#ifdef CONFIG_RISCV_NONSTANDARD_CACHE_OPS
if (unlikely(noncoherent_cache_ops.wback)) {
noncoherent_cache_ops.wback(paddr, size);
return;
}
#endif
ALT_CMO_OP(CLEAN, vaddr, size, riscv_cbom_block_size);
}
static inline void arch_dma_cache_inv(phys_addr_t paddr, size_t size)
{
void *vaddr = phys_to_virt(paddr);
#ifdef CONFIG_RISCV_NONSTANDARD_CACHE_OPS
if (unlikely(noncoherent_cache_ops.inv)) {
noncoherent_cache_ops.inv(paddr, size);
return;
}
#endif
ALT_CMO_OP(INVAL, vaddr, size, riscv_cbom_block_size);
}
static inline void arch_dma_cache_wback_inv(phys_addr_t paddr, size_t size)
{
void *vaddr = phys_to_virt(paddr);
#ifdef CONFIG_RISCV_NONSTANDARD_CACHE_OPS
if (unlikely(noncoherent_cache_ops.wback_inv)) {
noncoherent_cache_ops.wback_inv(paddr, size);
return;
}
#endif
ALT_CMO_OP(FLUSH, vaddr, size, riscv_cbom_block_size);
}
static inline bool arch_sync_dma_clean_before_fromdevice(void)
{
return true;
}
static inline bool arch_sync_dma_cpu_needs_post_dma_flush(void)
{
return true;
}
void arch_sync_dma_for_device(phys_addr_t paddr, size_t size,
enum dma_data_direction dir)
{
switch (dir) {
case DMA_TO_DEVICE:
arch_dma_cache_wback(paddr, size);
break;
case DMA_FROM_DEVICE:
if (!arch_sync_dma_clean_before_fromdevice()) {
arch_dma_cache_inv(paddr, size);
break;
}
fallthrough;
case DMA_BIDIRECTIONAL:
/* Skip the invalidate here if it's done later */
if (IS_ENABLED(CONFIG_ARCH_HAS_SYNC_DMA_FOR_CPU) &&
arch_sync_dma_cpu_needs_post_dma_flush())
arch_dma_cache_wback(paddr, size);
else
arch_dma_cache_wback_inv(paddr, size);
break;
default:
break;
}
}
void arch_sync_dma_for_cpu(phys_addr_t paddr, size_t size,
enum dma_data_direction dir)
{
switch (dir) {
case DMA_TO_DEVICE:
break;
case DMA_FROM_DEVICE:
case DMA_BIDIRECTIONAL:
/* FROM_DEVICE invalidate needed if speculative CPU prefetch only */
if (arch_sync_dma_cpu_needs_post_dma_flush())
arch_dma_cache_inv(paddr, size);
break;
default:
break;
}
}
void arch_dma_prep_coherent(struct page *page, size_t size)
{
void *flush_addr = page_address(page);
#ifdef CONFIG_RISCV_NONSTANDARD_CACHE_OPS
if (unlikely(noncoherent_cache_ops.wback_inv)) {
noncoherent_cache_ops.wback_inv(page_to_phys(page), size);
return;
}
#endif
ALT_CMO_OP(FLUSH, flush_addr, size, riscv_cbom_block_size);
}
void arch_setup_dma_ops(struct device *dev, bool coherent)
{
WARN_TAINT(!coherent && riscv_cbom_block_size > ARCH_DMA_MINALIGN,
TAINT_CPU_OUT_OF_SPEC,
"%s %s: ARCH_DMA_MINALIGN smaller than riscv,cbom-block-size (%d < %d)",
dev_driver_string(dev), dev_name(dev),
ARCH_DMA_MINALIGN, riscv_cbom_block_size);
WARN_TAINT(!coherent && !noncoherent_supported, TAINT_CPU_OUT_OF_SPEC,
"%s %s: device non-coherent but no non-coherent operations supported",
dev_driver_string(dev), dev_name(dev));
dev->dma_coherent = coherent;
}
void riscv_noncoherent_supported(void)
{
WARN(!riscv_cbom_block_size,
"Non-coherent DMA support enabled without a block size\n");
noncoherent_supported = true;
}
riscv: allow kmalloc() caches aligned to the smallest value Currently, riscv defines ARCH_DMA_MINALIGN as L1_CACHE_BYTES, I.E 64Bytes, if CONFIG_RISCV_DMA_NONCOHERENT=y. To support unified kernel Image, usually we have to enable CONFIG_RISCV_DMA_NONCOHERENT, thus it brings some bad effects to coherent platforms: Firstly, it wastes memory, kmalloc-96, kmalloc-32, kmalloc-16 and kmalloc-8 slab caches don't exist any more, they are replaced with either kmalloc-128 or kmalloc-64. Secondly, larger than necessary kmalloc aligned allocations results in unnecessary cache/TLB pressure. This issue also exists on arm64 platforms. From last year, Catalin tried to solve this issue by decoupling ARCH_KMALLOC_MINALIGN from ARCH_DMA_MINALIGN, limiting kmalloc() minimum alignment to dma_get_cache_alignment() and replacing ARCH_KMALLOC_MINALIGN usage in various drivers with ARCH_DMA_MINALIGN etc.[1] One fact we can make use of for riscv: if the CPU doesn't support ZICBOM or T-HEAD CMO, we know the platform is coherent. Based on Catalin's work and above fact, we can easily solve the kmalloc align issue for riscv: we can override dma_get_cache_alignment(), then let it return ARCH_DMA_MINALIGN at the beginning and return 1 once we know the underlying HW neither supports ZICBOM nor supports T-HEAD CMO. So what about if the CPU supports ZICBOM or T-HEAD CMO, but all the devices are dma coherent? Well, we use ARCH_DMA_MINALIGN as the kmalloc minimum alignment, nothing changed in this case. This case can be improved in the future. After this patch, a simple test of booting to a small buildroot rootfs on qemu shows: kmalloc-96 5041 5041 96 ... kmalloc-64 9606 9606 64 ... kmalloc-32 5128 5128 32 ... kmalloc-16 7682 7682 16 ... kmalloc-8 10246 10246 8 ... So we save about 1268KB memory. The saving will be much larger in normal OS env on real HW platforms. Link: https://lore.kernel.org/linux-arm-kernel/20230524171904.3967031-1-catalin.marinas@arm.com/ [1] Signed-off-by: Jisheng Zhang <jszhang@kernel.org> Reviewed-by: Conor Dooley <conor.dooley@microchip.com> Link: https://lore.kernel.org/r/20230718152214.2907-2-jszhang@kernel.org Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
2023-07-18 08:22:13 -07:00
void __init riscv_set_dma_cache_alignment(void)
{
if (!noncoherent_supported)
dma_cache_alignment = 1;
}