diff --git a/Makefile b/Makefile index f13f31e1f56b..ff5dd2fcaab1 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ # SPDX-License-Identifier: GPL-2.0 VERSION = 6 PATCHLEVEL = 1 -SUBLEVEL = 109 +SUBLEVEL = 110 EXTRAVERSION = NAME = Curry Ramen diff --git a/arch/arm64/include/asm/acpi.h b/arch/arm64/include/asm/acpi.h index bd68e1b7f29f..702587fda70c 100644 --- a/arch/arm64/include/asm/acpi.h +++ b/arch/arm64/include/asm/acpi.h @@ -97,6 +97,18 @@ static inline u32 get_acpi_id_for_cpu(unsigned int cpu) return acpi_cpu_get_madt_gicc(cpu)->uid; } +static inline int get_cpu_for_acpi_id(u32 uid) +{ + int cpu; + + for (cpu = 0; cpu < nr_cpu_ids; cpu++) + if (acpi_cpu_get_madt_gicc(cpu) && + uid == get_acpi_id_for_cpu(cpu)) + return cpu; + + return -EINVAL; +} + static inline void arch_fix_phys_package_id(int num, u32 slot) { } void __init acpi_init_cpus(void); int apei_claim_sea(struct pt_regs *regs); diff --git a/arch/arm64/kernel/acpi_numa.c b/arch/arm64/kernel/acpi_numa.c index ccbff21ce1fa..2465f291c7e1 100644 --- a/arch/arm64/kernel/acpi_numa.c +++ b/arch/arm64/kernel/acpi_numa.c @@ -34,17 +34,6 @@ int __init acpi_numa_get_nid(unsigned int cpu) return acpi_early_node_map[cpu]; } -static inline int get_cpu_for_acpi_id(u32 uid) -{ - int cpu; - - for (cpu = 0; cpu < nr_cpu_ids; cpu++) - if (uid == get_acpi_id_for_cpu(cpu)) - return cpu; - - return -EINVAL; -} - static int __init acpi_parse_gicc_pxm(union acpi_subtable_headers *header, const unsigned long end) { diff --git a/arch/powerpc/include/asm/nohash/mmu-e500.h b/arch/powerpc/include/asm/nohash/mmu-e500.h index e43a418d3ccd..9b5ba73d33d6 100644 --- a/arch/powerpc/include/asm/nohash/mmu-e500.h +++ b/arch/powerpc/include/asm/nohash/mmu-e500.h @@ -303,8 +303,7 @@ extern unsigned long linear_map_top; extern int book3e_htw_mode; #define PPC_HTW_NONE 0 -#define PPC_HTW_IBM 1 -#define PPC_HTW_E6500 2 +#define PPC_HTW_E6500 1 /* * 64-bit booke platforms don't load the tlb in the tlb miss handler code. diff --git a/arch/powerpc/mm/nohash/Makefile b/arch/powerpc/mm/nohash/Makefile index f3894e79d5f7..24b445a5fcac 100644 --- a/arch/powerpc/mm/nohash/Makefile +++ b/arch/powerpc/mm/nohash/Makefile @@ -3,7 +3,7 @@ ccflags-$(CONFIG_PPC64) := $(NO_MINIMAL_TOC) obj-y += mmu_context.o tlb.o tlb_low.o kup.o -obj-$(CONFIG_PPC_BOOK3E_64) += tlb_low_64e.o book3e_pgtable.o +obj-$(CONFIG_PPC_BOOK3E_64) += tlb_64e.o tlb_low_64e.o book3e_pgtable.o obj-$(CONFIG_40x) += 40x.o obj-$(CONFIG_44x) += 44x.o obj-$(CONFIG_PPC_8xx) += 8xx.o diff --git a/arch/powerpc/mm/nohash/tlb.c b/arch/powerpc/mm/nohash/tlb.c index 2c15c86c7015..c0b643d30fcc 100644 --- a/arch/powerpc/mm/nohash/tlb.c +++ b/arch/powerpc/mm/nohash/tlb.c @@ -110,28 +110,6 @@ struct mmu_psize_def mmu_psize_defs[MMU_PAGE_COUNT] = { }; #endif -/* The variables below are currently only used on 64-bit Book3E - * though this will probably be made common with other nohash - * implementations at some point - */ -#ifdef CONFIG_PPC64 - -int mmu_pte_psize; /* Page size used for PTE pages */ -int mmu_vmemmap_psize; /* Page size used for the virtual mem map */ -int book3e_htw_mode; /* HW tablewalk? Value is PPC_HTW_* */ -unsigned long linear_map_top; /* Top of linear mapping */ - - -/* - * Number of bytes to add to SPRN_SPRG_TLB_EXFRAME on crit/mcheck/debug - * exceptions. This is used for bolted and e6500 TLB miss handlers which - * do not modify this SPRG in the TLB miss code; for other TLB miss handlers, - * this is set to zero. - */ -int extlb_level_exc; - -#endif /* CONFIG_PPC64 */ - #ifdef CONFIG_PPC_E500 /* next_tlbcam_idx is used to round-robin tlbcam entry assignment */ DEFINE_PER_CPU(int, next_tlbcam_idx); @@ -361,381 +339,7 @@ void tlb_flush(struct mmu_gather *tlb) flush_tlb_mm(tlb->mm); } -/* - * Below are functions specific to the 64-bit variant of Book3E though that - * may change in the future - */ - -#ifdef CONFIG_PPC64 - -/* - * Handling of virtual linear page tables or indirect TLB entries - * flushing when PTE pages are freed - */ -void tlb_flush_pgtable(struct mmu_gather *tlb, unsigned long address) -{ - int tsize = mmu_psize_defs[mmu_pte_psize].enc; - - if (book3e_htw_mode != PPC_HTW_NONE) { - unsigned long start = address & PMD_MASK; - unsigned long end = address + PMD_SIZE; - unsigned long size = 1UL << mmu_psize_defs[mmu_pte_psize].shift; - - /* This isn't the most optimal, ideally we would factor out the - * while preempt & CPU mask mucking around, or even the IPI but - * it will do for now - */ - while (start < end) { - __flush_tlb_page(tlb->mm, start, tsize, 1); - start += size; - } - } else { - unsigned long rmask = 0xf000000000000000ul; - unsigned long rid = (address & rmask) | 0x1000000000000000ul; - unsigned long vpte = address & ~rmask; - - vpte = (vpte >> (PAGE_SHIFT - 3)) & ~0xffful; - vpte |= rid; - __flush_tlb_page(tlb->mm, vpte, tsize, 0); - } -} - -static void __init setup_page_sizes(void) -{ - unsigned int tlb0cfg; - unsigned int tlb0ps; - unsigned int eptcfg; - int i, psize; - -#ifdef CONFIG_PPC_E500 - unsigned int mmucfg = mfspr(SPRN_MMUCFG); - int fsl_mmu = mmu_has_feature(MMU_FTR_TYPE_FSL_E); - - if (fsl_mmu && (mmucfg & MMUCFG_MAVN) == MMUCFG_MAVN_V1) { - unsigned int tlb1cfg = mfspr(SPRN_TLB1CFG); - unsigned int min_pg, max_pg; - - min_pg = (tlb1cfg & TLBnCFG_MINSIZE) >> TLBnCFG_MINSIZE_SHIFT; - max_pg = (tlb1cfg & TLBnCFG_MAXSIZE) >> TLBnCFG_MAXSIZE_SHIFT; - - for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) { - struct mmu_psize_def *def; - unsigned int shift; - - def = &mmu_psize_defs[psize]; - shift = def->shift; - - if (shift == 0 || shift & 1) - continue; - - /* adjust to be in terms of 4^shift Kb */ - shift = (shift - 10) >> 1; - - if ((shift >= min_pg) && (shift <= max_pg)) - def->flags |= MMU_PAGE_SIZE_DIRECT; - } - - goto out; - } - - if (fsl_mmu && (mmucfg & MMUCFG_MAVN) == MMUCFG_MAVN_V2) { - u32 tlb1cfg, tlb1ps; - - tlb0cfg = mfspr(SPRN_TLB0CFG); - tlb1cfg = mfspr(SPRN_TLB1CFG); - tlb1ps = mfspr(SPRN_TLB1PS); - eptcfg = mfspr(SPRN_EPTCFG); - - if ((tlb1cfg & TLBnCFG_IND) && (tlb0cfg & TLBnCFG_PT)) - book3e_htw_mode = PPC_HTW_E6500; - - /* - * We expect 4K subpage size and unrestricted indirect size. - * The lack of a restriction on indirect size is a Freescale - * extension, indicated by PSn = 0 but SPSn != 0. - */ - if (eptcfg != 2) - book3e_htw_mode = PPC_HTW_NONE; - - for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) { - struct mmu_psize_def *def = &mmu_psize_defs[psize]; - - if (!def->shift) - continue; - - if (tlb1ps & (1U << (def->shift - 10))) { - def->flags |= MMU_PAGE_SIZE_DIRECT; - - if (book3e_htw_mode && psize == MMU_PAGE_2M) - def->flags |= MMU_PAGE_SIZE_INDIRECT; - } - } - - goto out; - } -#endif - - tlb0cfg = mfspr(SPRN_TLB0CFG); - tlb0ps = mfspr(SPRN_TLB0PS); - eptcfg = mfspr(SPRN_EPTCFG); - - /* Look for supported direct sizes */ - for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) { - struct mmu_psize_def *def = &mmu_psize_defs[psize]; - - if (tlb0ps & (1U << (def->shift - 10))) - def->flags |= MMU_PAGE_SIZE_DIRECT; - } - - /* Indirect page sizes supported ? */ - if ((tlb0cfg & TLBnCFG_IND) == 0 || - (tlb0cfg & TLBnCFG_PT) == 0) - goto out; - - book3e_htw_mode = PPC_HTW_IBM; - - /* Now, we only deal with one IND page size for each - * direct size. Hopefully all implementations today are - * unambiguous, but we might want to be careful in the - * future. - */ - for (i = 0; i < 3; i++) { - unsigned int ps, sps; - - sps = eptcfg & 0x1f; - eptcfg >>= 5; - ps = eptcfg & 0x1f; - eptcfg >>= 5; - if (!ps || !sps) - continue; - for (psize = 0; psize < MMU_PAGE_COUNT; psize++) { - struct mmu_psize_def *def = &mmu_psize_defs[psize]; - - if (ps == (def->shift - 10)) - def->flags |= MMU_PAGE_SIZE_INDIRECT; - if (sps == (def->shift - 10)) - def->ind = ps + 10; - } - } - -out: - /* Cleanup array and print summary */ - pr_info("MMU: Supported page sizes\n"); - for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) { - struct mmu_psize_def *def = &mmu_psize_defs[psize]; - const char *__page_type_names[] = { - "unsupported", - "direct", - "indirect", - "direct & indirect" - }; - if (def->flags == 0) { - def->shift = 0; - continue; - } - pr_info(" %8ld KB as %s\n", 1ul << (def->shift - 10), - __page_type_names[def->flags & 0x3]); - } -} - -static void __init setup_mmu_htw(void) -{ - /* - * If we want to use HW tablewalk, enable it by patching the TLB miss - * handlers to branch to the one dedicated to it. - */ - - switch (book3e_htw_mode) { - case PPC_HTW_IBM: - patch_exception(0x1c0, exc_data_tlb_miss_htw_book3e); - patch_exception(0x1e0, exc_instruction_tlb_miss_htw_book3e); - break; -#ifdef CONFIG_PPC_E500 - case PPC_HTW_E6500: - extlb_level_exc = EX_TLB_SIZE; - patch_exception(0x1c0, exc_data_tlb_miss_e6500_book3e); - patch_exception(0x1e0, exc_instruction_tlb_miss_e6500_book3e); - break; -#endif - } - pr_info("MMU: Book3E HW tablewalk %s\n", - book3e_htw_mode != PPC_HTW_NONE ? "enabled" : "not supported"); -} - -/* - * Early initialization of the MMU TLB code - */ -static void early_init_this_mmu(void) -{ - unsigned int mas4; - - /* Set MAS4 based on page table setting */ - - mas4 = 0x4 << MAS4_WIMGED_SHIFT; - switch (book3e_htw_mode) { - case PPC_HTW_E6500: - mas4 |= MAS4_INDD; - mas4 |= BOOK3E_PAGESZ_2M << MAS4_TSIZED_SHIFT; - mas4 |= MAS4_TLBSELD(1); - mmu_pte_psize = MMU_PAGE_2M; - break; - - case PPC_HTW_IBM: - mas4 |= MAS4_INDD; - mas4 |= BOOK3E_PAGESZ_1M << MAS4_TSIZED_SHIFT; - mmu_pte_psize = MMU_PAGE_1M; - break; - - case PPC_HTW_NONE: - mas4 |= BOOK3E_PAGESZ_4K << MAS4_TSIZED_SHIFT; - mmu_pte_psize = mmu_virtual_psize; - break; - } - mtspr(SPRN_MAS4, mas4); - -#ifdef CONFIG_PPC_E500 - if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { - unsigned int num_cams; - bool map = true; - - /* use a quarter of the TLBCAM for bolted linear map */ - num_cams = (mfspr(SPRN_TLB1CFG) & TLBnCFG_N_ENTRY) / 4; - - /* - * Only do the mapping once per core, or else the - * transient mapping would cause problems. - */ -#ifdef CONFIG_SMP - if (hweight32(get_tensr()) > 1) - map = false; -#endif - - if (map) - linear_map_top = map_mem_in_cams(linear_map_top, - num_cams, false, true); - } -#endif - - /* A sync won't hurt us after mucking around with - * the MMU configuration - */ - mb(); -} - -static void __init early_init_mmu_global(void) -{ - /* XXX This should be decided at runtime based on supported - * page sizes in the TLB, but for now let's assume 16M is - * always there and a good fit (which it probably is) - * - * Freescale booke only supports 4K pages in TLB0, so use that. - */ - if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) - mmu_vmemmap_psize = MMU_PAGE_4K; - else - mmu_vmemmap_psize = MMU_PAGE_16M; - - /* XXX This code only checks for TLB 0 capabilities and doesn't - * check what page size combos are supported by the HW. It - * also doesn't handle the case where a separate array holds - * the IND entries from the array loaded by the PT. - */ - /* Look for supported page sizes */ - setup_page_sizes(); - - /* Look for HW tablewalk support */ - setup_mmu_htw(); - -#ifdef CONFIG_PPC_E500 - if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { - if (book3e_htw_mode == PPC_HTW_NONE) { - extlb_level_exc = EX_TLB_SIZE; - patch_exception(0x1c0, exc_data_tlb_miss_bolted_book3e); - patch_exception(0x1e0, - exc_instruction_tlb_miss_bolted_book3e); - } - } -#endif - - /* Set the global containing the top of the linear mapping - * for use by the TLB miss code - */ - linear_map_top = memblock_end_of_DRAM(); - - ioremap_bot = IOREMAP_BASE; -} - -static void __init early_mmu_set_memory_limit(void) -{ -#ifdef CONFIG_PPC_E500 - if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { - /* - * Limit memory so we dont have linear faults. - * Unlike memblock_set_current_limit, which limits - * memory available during early boot, this permanently - * reduces the memory available to Linux. We need to - * do this because highmem is not supported on 64-bit. - */ - memblock_enforce_memory_limit(linear_map_top); - } -#endif - - memblock_set_current_limit(linear_map_top); -} - -/* boot cpu only */ -void __init early_init_mmu(void) -{ - early_init_mmu_global(); - early_init_this_mmu(); - early_mmu_set_memory_limit(); -} - -void early_init_mmu_secondary(void) -{ - early_init_this_mmu(); -} - -void setup_initial_memory_limit(phys_addr_t first_memblock_base, - phys_addr_t first_memblock_size) -{ - /* On non-FSL Embedded 64-bit, we adjust the RMA size to match - * the bolted TLB entry. We know for now that only 1G - * entries are supported though that may eventually - * change. - * - * on FSL Embedded 64-bit, usually all RAM is bolted, but with - * unusual memory sizes it's possible for some RAM to not be mapped - * (such RAM is not used at all by Linux, since we don't support - * highmem on 64-bit). We limit ppc64_rma_size to what would be - * mappable if this memblock is the only one. Additional memblocks - * can only increase, not decrease, the amount that ends up getting - * mapped. We still limit max to 1G even if we'll eventually map - * more. This is due to what the early init code is set up to do. - * - * We crop it to the size of the first MEMBLOCK to - * avoid going over total available memory just in case... - */ -#ifdef CONFIG_PPC_E500 - if (early_mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { - unsigned long linear_sz; - unsigned int num_cams; - - /* use a quarter of the TLBCAM for bolted linear map */ - num_cams = (mfspr(SPRN_TLB1CFG) & TLBnCFG_N_ENTRY) / 4; - - linear_sz = map_mem_in_cams(first_memblock_size, num_cams, - true, true); - - ppc64_rma_size = min_t(u64, linear_sz, 0x40000000); - } else -#endif - ppc64_rma_size = min_t(u64, first_memblock_size, 0x40000000); - - /* Finally limit subsequent allocations */ - memblock_set_current_limit(first_memblock_base + ppc64_rma_size); -} -#else /* ! CONFIG_PPC64 */ +#ifndef CONFIG_PPC64 void __init early_init_mmu(void) { #ifdef CONFIG_PPC_47x diff --git a/arch/powerpc/mm/nohash/tlb_64e.c b/arch/powerpc/mm/nohash/tlb_64e.c new file mode 100644 index 000000000000..b6af3ec4d001 --- /dev/null +++ b/arch/powerpc/mm/nohash/tlb_64e.c @@ -0,0 +1,361 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Copyright 2008,2009 Ben Herrenschmidt + * IBM Corp. + * + * Derived from arch/ppc/mm/init.c: + * Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org) + * + * Modifications by Paul Mackerras (PowerMac) (paulus@cs.anu.edu.au) + * and Cort Dougan (PReP) (cort@cs.nmt.edu) + * Copyright (C) 1996 Paul Mackerras + * + * Derived from "arch/i386/mm/init.c" + * Copyright (C) 1991, 1992, 1993, 1994 Linus Torvalds + */ + +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#include + +/* The variables below are currently only used on 64-bit Book3E + * though this will probably be made common with other nohash + * implementations at some point + */ +static int mmu_pte_psize; /* Page size used for PTE pages */ +int mmu_vmemmap_psize; /* Page size used for the virtual mem map */ +int book3e_htw_mode; /* HW tablewalk? Value is PPC_HTW_* */ +unsigned long linear_map_top; /* Top of linear mapping */ + + +/* + * Number of bytes to add to SPRN_SPRG_TLB_EXFRAME on crit/mcheck/debug + * exceptions. This is used for bolted and e6500 TLB miss handlers which + * do not modify this SPRG in the TLB miss code; for other TLB miss handlers, + * this is set to zero. + */ +int extlb_level_exc; + +/* + * Handling of virtual linear page tables or indirect TLB entries + * flushing when PTE pages are freed + */ +void tlb_flush_pgtable(struct mmu_gather *tlb, unsigned long address) +{ + int tsize = mmu_psize_defs[mmu_pte_psize].enc; + + if (book3e_htw_mode != PPC_HTW_NONE) { + unsigned long start = address & PMD_MASK; + unsigned long end = address + PMD_SIZE; + unsigned long size = 1UL << mmu_psize_defs[mmu_pte_psize].shift; + + /* This isn't the most optimal, ideally we would factor out the + * while preempt & CPU mask mucking around, or even the IPI but + * it will do for now + */ + while (start < end) { + __flush_tlb_page(tlb->mm, start, tsize, 1); + start += size; + } + } else { + unsigned long rmask = 0xf000000000000000ul; + unsigned long rid = (address & rmask) | 0x1000000000000000ul; + unsigned long vpte = address & ~rmask; + + vpte = (vpte >> (PAGE_SHIFT - 3)) & ~0xffful; + vpte |= rid; + __flush_tlb_page(tlb->mm, vpte, tsize, 0); + } +} + +static void __init setup_page_sizes(void) +{ + unsigned int tlb0cfg; + unsigned int eptcfg; + int psize; + +#ifdef CONFIG_PPC_E500 + unsigned int mmucfg = mfspr(SPRN_MMUCFG); + int fsl_mmu = mmu_has_feature(MMU_FTR_TYPE_FSL_E); + + if (fsl_mmu && (mmucfg & MMUCFG_MAVN) == MMUCFG_MAVN_V1) { + unsigned int tlb1cfg = mfspr(SPRN_TLB1CFG); + unsigned int min_pg, max_pg; + + min_pg = (tlb1cfg & TLBnCFG_MINSIZE) >> TLBnCFG_MINSIZE_SHIFT; + max_pg = (tlb1cfg & TLBnCFG_MAXSIZE) >> TLBnCFG_MAXSIZE_SHIFT; + + for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) { + struct mmu_psize_def *def; + unsigned int shift; + + def = &mmu_psize_defs[psize]; + shift = def->shift; + + if (shift == 0 || shift & 1) + continue; + + /* adjust to be in terms of 4^shift Kb */ + shift = (shift - 10) >> 1; + + if ((shift >= min_pg) && (shift <= max_pg)) + def->flags |= MMU_PAGE_SIZE_DIRECT; + } + + goto out; + } + + if (fsl_mmu && (mmucfg & MMUCFG_MAVN) == MMUCFG_MAVN_V2) { + u32 tlb1cfg, tlb1ps; + + tlb0cfg = mfspr(SPRN_TLB0CFG); + tlb1cfg = mfspr(SPRN_TLB1CFG); + tlb1ps = mfspr(SPRN_TLB1PS); + eptcfg = mfspr(SPRN_EPTCFG); + + if ((tlb1cfg & TLBnCFG_IND) && (tlb0cfg & TLBnCFG_PT)) + book3e_htw_mode = PPC_HTW_E6500; + + /* + * We expect 4K subpage size and unrestricted indirect size. + * The lack of a restriction on indirect size is a Freescale + * extension, indicated by PSn = 0 but SPSn != 0. + */ + if (eptcfg != 2) + book3e_htw_mode = PPC_HTW_NONE; + + for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) { + struct mmu_psize_def *def = &mmu_psize_defs[psize]; + + if (!def->shift) + continue; + + if (tlb1ps & (1U << (def->shift - 10))) { + def->flags |= MMU_PAGE_SIZE_DIRECT; + + if (book3e_htw_mode && psize == MMU_PAGE_2M) + def->flags |= MMU_PAGE_SIZE_INDIRECT; + } + } + + goto out; + } +#endif +out: + /* Cleanup array and print summary */ + pr_info("MMU: Supported page sizes\n"); + for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) { + struct mmu_psize_def *def = &mmu_psize_defs[psize]; + const char *__page_type_names[] = { + "unsupported", + "direct", + "indirect", + "direct & indirect" + }; + if (def->flags == 0) { + def->shift = 0; + continue; + } + pr_info(" %8ld KB as %s\n", 1ul << (def->shift - 10), + __page_type_names[def->flags & 0x3]); + } +} + +static void __init setup_mmu_htw(void) +{ + /* + * If we want to use HW tablewalk, enable it by patching the TLB miss + * handlers to branch to the one dedicated to it. + */ + + switch (book3e_htw_mode) { +#ifdef CONFIG_PPC_E500 + case PPC_HTW_E6500: + extlb_level_exc = EX_TLB_SIZE; + patch_exception(0x1c0, exc_data_tlb_miss_e6500_book3e); + patch_exception(0x1e0, exc_instruction_tlb_miss_e6500_book3e); + break; +#endif + } + pr_info("MMU: Book3E HW tablewalk %s\n", + book3e_htw_mode != PPC_HTW_NONE ? "enabled" : "not supported"); +} + +/* + * Early initialization of the MMU TLB code + */ +static void early_init_this_mmu(void) +{ + unsigned int mas4; + + /* Set MAS4 based on page table setting */ + + mas4 = 0x4 << MAS4_WIMGED_SHIFT; + switch (book3e_htw_mode) { + case PPC_HTW_E6500: + mas4 |= MAS4_INDD; + mas4 |= BOOK3E_PAGESZ_2M << MAS4_TSIZED_SHIFT; + mas4 |= MAS4_TLBSELD(1); + mmu_pte_psize = MMU_PAGE_2M; + break; + + case PPC_HTW_NONE: + mas4 |= BOOK3E_PAGESZ_4K << MAS4_TSIZED_SHIFT; + mmu_pte_psize = mmu_virtual_psize; + break; + } + mtspr(SPRN_MAS4, mas4); + +#ifdef CONFIG_PPC_E500 + if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { + unsigned int num_cams; + bool map = true; + + /* use a quarter of the TLBCAM for bolted linear map */ + num_cams = (mfspr(SPRN_TLB1CFG) & TLBnCFG_N_ENTRY) / 4; + + /* + * Only do the mapping once per core, or else the + * transient mapping would cause problems. + */ +#ifdef CONFIG_SMP + if (hweight32(get_tensr()) > 1) + map = false; +#endif + + if (map) + linear_map_top = map_mem_in_cams(linear_map_top, + num_cams, false, true); + } +#endif + + /* A sync won't hurt us after mucking around with + * the MMU configuration + */ + mb(); +} + +static void __init early_init_mmu_global(void) +{ + /* XXX This should be decided at runtime based on supported + * page sizes in the TLB, but for now let's assume 16M is + * always there and a good fit (which it probably is) + * + * Freescale booke only supports 4K pages in TLB0, so use that. + */ + if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) + mmu_vmemmap_psize = MMU_PAGE_4K; + else + mmu_vmemmap_psize = MMU_PAGE_16M; + + /* XXX This code only checks for TLB 0 capabilities and doesn't + * check what page size combos are supported by the HW. It + * also doesn't handle the case where a separate array holds + * the IND entries from the array loaded by the PT. + */ + /* Look for supported page sizes */ + setup_page_sizes(); + + /* Look for HW tablewalk support */ + setup_mmu_htw(); + +#ifdef CONFIG_PPC_E500 + if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { + if (book3e_htw_mode == PPC_HTW_NONE) { + extlb_level_exc = EX_TLB_SIZE; + patch_exception(0x1c0, exc_data_tlb_miss_bolted_book3e); + patch_exception(0x1e0, + exc_instruction_tlb_miss_bolted_book3e); + } + } +#endif + + /* Set the global containing the top of the linear mapping + * for use by the TLB miss code + */ + linear_map_top = memblock_end_of_DRAM(); + + ioremap_bot = IOREMAP_BASE; +} + +static void __init early_mmu_set_memory_limit(void) +{ +#ifdef CONFIG_PPC_E500 + if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { + /* + * Limit memory so we dont have linear faults. + * Unlike memblock_set_current_limit, which limits + * memory available during early boot, this permanently + * reduces the memory available to Linux. We need to + * do this because highmem is not supported on 64-bit. + */ + memblock_enforce_memory_limit(linear_map_top); + } +#endif + + memblock_set_current_limit(linear_map_top); +} + +/* boot cpu only */ +void __init early_init_mmu(void) +{ + early_init_mmu_global(); + early_init_this_mmu(); + early_mmu_set_memory_limit(); +} + +void early_init_mmu_secondary(void) +{ + early_init_this_mmu(); +} + +void setup_initial_memory_limit(phys_addr_t first_memblock_base, + phys_addr_t first_memblock_size) +{ + /* On non-FSL Embedded 64-bit, we adjust the RMA size to match + * the bolted TLB entry. We know for now that only 1G + * entries are supported though that may eventually + * change. + * + * on FSL Embedded 64-bit, usually all RAM is bolted, but with + * unusual memory sizes it's possible for some RAM to not be mapped + * (such RAM is not used at all by Linux, since we don't support + * highmem on 64-bit). We limit ppc64_rma_size to what would be + * mappable if this memblock is the only one. Additional memblocks + * can only increase, not decrease, the amount that ends up getting + * mapped. We still limit max to 1G even if we'll eventually map + * more. This is due to what the early init code is set up to do. + * + * We crop it to the size of the first MEMBLOCK to + * avoid going over total available memory just in case... + */ +#ifdef CONFIG_PPC_E500 + if (early_mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { + unsigned long linear_sz; + unsigned int num_cams; + + /* use a quarter of the TLBCAM for bolted linear map */ + num_cams = (mfspr(SPRN_TLB1CFG) & TLBnCFG_N_ENTRY) / 4; + + linear_sz = map_mem_in_cams(first_memblock_size, num_cams, + true, true); + + ppc64_rma_size = min_t(u64, linear_sz, 0x40000000); + } else +#endif + ppc64_rma_size = min_t(u64, first_memblock_size, 0x40000000); + + /* Finally limit subsequent allocations */ + memblock_set_current_limit(first_memblock_base + ppc64_rma_size); +} diff --git a/arch/powerpc/mm/nohash/tlb_low_64e.S b/arch/powerpc/mm/nohash/tlb_low_64e.S index 76cf456d7976..d831a111eaba 100644 --- a/arch/powerpc/mm/nohash/tlb_low_64e.S +++ b/arch/powerpc/mm/nohash/tlb_low_64e.S @@ -893,201 +893,6 @@ virt_page_table_tlb_miss_whacko_fault: TLB_MISS_EPILOG_ERROR b exc_data_storage_book3e - -/************************************************************** - * * - * TLB miss handling for Book3E with hw page table support * - * * - **************************************************************/ - - -/* Data TLB miss */ - START_EXCEPTION(data_tlb_miss_htw) - TLB_MISS_PROLOG - - /* Now we handle the fault proper. We only save DEAR in normal - * fault case since that's the only interesting values here. - * We could probably also optimize by not saving SRR0/1 in the - * linear mapping case but I'll leave that for later - */ - mfspr r14,SPRN_ESR - mfspr r16,SPRN_DEAR /* get faulting address */ - srdi r11,r16,44 /* get region */ - xoris r11,r11,0xc - cmpldi cr0,r11,0 /* linear mapping ? */ - beq tlb_load_linear /* yes -> go to linear map load */ - cmpldi cr1,r11,1 /* vmalloc mapping ? */ - - /* We do the user/kernel test for the PID here along with the RW test - */ - srdi. r11,r16,60 /* Check for user region */ - ld r15,PACAPGD(r13) /* Load user pgdir */ - beq htw_tlb_miss - - /* XXX replace the RMW cycles with immediate loads + writes */ -1: mfspr r10,SPRN_MAS1 - rlwinm r10,r10,0,16,1 /* Clear TID */ - mtspr SPRN_MAS1,r10 - ld r15,PACA_KERNELPGD(r13) /* Load kernel pgdir */ - beq+ cr1,htw_tlb_miss - - /* We got a crappy address, just fault with whatever DEAR and ESR - * are here - */ - TLB_MISS_EPILOG_ERROR - b exc_data_storage_book3e - -/* Instruction TLB miss */ - START_EXCEPTION(instruction_tlb_miss_htw) - TLB_MISS_PROLOG - - /* If we take a recursive fault, the second level handler may need - * to know whether we are handling a data or instruction fault in - * order to get to the right store fault handler. We provide that - * info by keeping a crazy value for ESR in r14 - */ - li r14,-1 /* store to exception frame is done later */ - - /* Now we handle the fault proper. We only save DEAR in the non - * linear mapping case since we know the linear mapping case will - * not re-enter. We could indeed optimize and also not save SRR0/1 - * in the linear mapping case but I'll leave that for later - * - * Faulting address is SRR0 which is already in r16 - */ - srdi r11,r16,44 /* get region */ - xoris r11,r11,0xc - cmpldi cr0,r11,0 /* linear mapping ? */ - beq tlb_load_linear /* yes -> go to linear map load */ - cmpldi cr1,r11,1 /* vmalloc mapping ? */ - - /* We do the user/kernel test for the PID here along with the RW test - */ - srdi. r11,r16,60 /* Check for user region */ - ld r15,PACAPGD(r13) /* Load user pgdir */ - beq htw_tlb_miss - - /* XXX replace the RMW cycles with immediate loads + writes */ -1: mfspr r10,SPRN_MAS1 - rlwinm r10,r10,0,16,1 /* Clear TID */ - mtspr SPRN_MAS1,r10 - ld r15,PACA_KERNELPGD(r13) /* Load kernel pgdir */ - beq+ htw_tlb_miss - - /* We got a crappy address, just fault */ - TLB_MISS_EPILOG_ERROR - b exc_instruction_storage_book3e - - -/* - * This is the guts of the second-level TLB miss handler for direct - * misses. We are entered with: - * - * r16 = virtual page table faulting address - * r15 = PGD pointer - * r14 = ESR - * r13 = PACA - * r12 = TLB exception frame in PACA - * r11 = crap (free to use) - * r10 = crap (free to use) - * - * It can be re-entered by the linear mapping miss handler. However, to - * avoid too much complication, it will save/restore things for us - */ -htw_tlb_miss: -#ifdef CONFIG_PPC_KUAP - mfspr r10,SPRN_MAS1 - rlwinm. r10,r10,0,0x3fff0000 - beq- htw_tlb_miss_fault /* KUAP fault */ -#endif - /* Search if we already have a TLB entry for that virtual address, and - * if we do, bail out. - * - * MAS1:IND should be already set based on MAS4 - */ - PPC_TLBSRX_DOT(0,R16) - beq htw_tlb_miss_done - - /* Now, we need to walk the page tables. First check if we are in - * range. - */ - rldicl. r10,r16,64-PGTABLE_EADDR_SIZE,PGTABLE_EADDR_SIZE+4 - bne- htw_tlb_miss_fault - - /* Get the PGD pointer */ - cmpldi cr0,r15,0 - beq- htw_tlb_miss_fault - - /* Get to PGD entry */ - rldicl r11,r16,64-(PGDIR_SHIFT-3),64-PGD_INDEX_SIZE-3 - clrrdi r10,r11,3 - ldx r15,r10,r15 - cmpdi cr0,r15,0 - bge htw_tlb_miss_fault - - /* Get to PUD entry */ - rldicl r11,r16,64-(PUD_SHIFT-3),64-PUD_INDEX_SIZE-3 - clrrdi r10,r11,3 - ldx r15,r10,r15 - cmpdi cr0,r15,0 - bge htw_tlb_miss_fault - - /* Get to PMD entry */ - rldicl r11,r16,64-(PMD_SHIFT-3),64-PMD_INDEX_SIZE-3 - clrrdi r10,r11,3 - ldx r15,r10,r15 - cmpdi cr0,r15,0 - bge htw_tlb_miss_fault - - /* Ok, we're all right, we can now create an indirect entry for - * a 1M or 256M page. - * - * The last trick is now that because we use "half" pages for - * the HTW (1M IND is 2K and 256M IND is 32K) we need to account - * for an added LSB bit to the RPN. For 64K pages, there is no - * problem as we already use 32K arrays (half PTE pages), but for - * 4K page we need to extract a bit from the virtual address and - * insert it into the "PA52" bit of the RPN. - */ - rlwimi r15,r16,32-9,20,20 - /* Now we build the MAS: - * - * MAS 0 : Fully setup with defaults in MAS4 and TLBnCFG - * MAS 1 : Almost fully setup - * - PID already updated by caller if necessary - * - TSIZE for now is base ind page size always - * MAS 2 : Use defaults - * MAS 3+7 : Needs to be done - */ - ori r10,r15,(BOOK3E_PAGESZ_4K << MAS3_SPSIZE_SHIFT) - - srdi r16,r10,32 - mtspr SPRN_MAS3,r10 - mtspr SPRN_MAS7,r16 - - tlbwe - -htw_tlb_miss_done: - /* We don't bother with restoring DEAR or ESR since we know we are - * level 0 and just going back to userland. They are only needed - * if you are going to take an access fault - */ - TLB_MISS_EPILOG_SUCCESS - rfi - -htw_tlb_miss_fault: - /* We need to check if it was an instruction miss. We know this - * though because r14 would contain -1 - */ - cmpdi cr0,r14,-1 - beq 1f - mtspr SPRN_DEAR,r16 - mtspr SPRN_ESR,r14 - TLB_MISS_EPILOG_ERROR - b exc_data_storage_book3e -1: TLB_MISS_EPILOG_ERROR - b exc_instruction_storage_book3e - /* * This is the guts of "any" level TLB miss handler for kernel linear * mapping misses. We are entered with: diff --git a/arch/x86/mm/pti.c b/arch/x86/mm/pti.c index 7b804c34c020..d5f702ea4c78 100644 --- a/arch/x86/mm/pti.c +++ b/arch/x86/mm/pti.c @@ -241,7 +241,7 @@ static pmd_t *pti_user_pagetable_walk_pmd(unsigned long address) * * Returns a pointer to a PTE on success, or NULL on failure. */ -static pte_t *pti_user_pagetable_walk_pte(unsigned long address) +static pte_t *pti_user_pagetable_walk_pte(unsigned long address, bool late_text) { gfp_t gfp = (GFP_KERNEL | __GFP_NOTRACK | __GFP_ZERO); pmd_t *pmd; @@ -251,10 +251,15 @@ static pte_t *pti_user_pagetable_walk_pte(unsigned long address) if (!pmd) return NULL; - /* We can't do anything sensible if we hit a large mapping. */ + /* Large PMD mapping found */ if (pmd_large(*pmd)) { - WARN_ON(1); - return NULL; + /* Clear the PMD if we hit a large mapping from the first round */ + if (late_text) { + set_pmd(pmd, __pmd(0)); + } else { + WARN_ON_ONCE(1); + return NULL; + } } if (pmd_none(*pmd)) { @@ -283,7 +288,7 @@ static void __init pti_setup_vsyscall(void) if (!pte || WARN_ON(level != PG_LEVEL_4K) || pte_none(*pte)) return; - target_pte = pti_user_pagetable_walk_pte(VSYSCALL_ADDR); + target_pte = pti_user_pagetable_walk_pte(VSYSCALL_ADDR, false); if (WARN_ON(!target_pte)) return; @@ -301,7 +306,7 @@ enum pti_clone_level { static void pti_clone_pgtable(unsigned long start, unsigned long end, - enum pti_clone_level level) + enum pti_clone_level level, bool late_text) { unsigned long addr; @@ -390,7 +395,7 @@ pti_clone_pgtable(unsigned long start, unsigned long end, return; /* Allocate PTE in the user page-table */ - target_pte = pti_user_pagetable_walk_pte(addr); + target_pte = pti_user_pagetable_walk_pte(addr, late_text); if (WARN_ON(!target_pte)) return; @@ -452,7 +457,7 @@ static void __init pti_clone_user_shared(void) phys_addr_t pa = per_cpu_ptr_to_phys((void *)va); pte_t *target_pte; - target_pte = pti_user_pagetable_walk_pte(va); + target_pte = pti_user_pagetable_walk_pte(va, false); if (WARN_ON(!target_pte)) return; @@ -475,7 +480,7 @@ static void __init pti_clone_user_shared(void) start = CPU_ENTRY_AREA_BASE; end = start + (PAGE_SIZE * CPU_ENTRY_AREA_PAGES); - pti_clone_pgtable(start, end, PTI_CLONE_PMD); + pti_clone_pgtable(start, end, PTI_CLONE_PMD, false); } #endif /* CONFIG_X86_64 */ @@ -492,11 +497,11 @@ static void __init pti_setup_espfix64(void) /* * Clone the populated PMDs of the entry text and force it RO. */ -static void pti_clone_entry_text(void) +static void pti_clone_entry_text(bool late) { pti_clone_pgtable((unsigned long) __entry_text_start, (unsigned long) __entry_text_end, - PTI_LEVEL_KERNEL_IMAGE); + PTI_LEVEL_KERNEL_IMAGE, late); } /* @@ -571,7 +576,7 @@ static void pti_clone_kernel_text(void) * pti_set_kernel_image_nonglobal() did to clear the * global bit. */ - pti_clone_pgtable(start, end_clone, PTI_LEVEL_KERNEL_IMAGE); + pti_clone_pgtable(start, end_clone, PTI_LEVEL_KERNEL_IMAGE, false); /* * pti_clone_pgtable() will set the global bit in any PMDs @@ -638,8 +643,15 @@ void __init pti_init(void) /* Undo all global bits from the init pagetables in head_64.S: */ pti_set_kernel_image_nonglobal(); + /* Replace some of the global bits just for shared entry text: */ - pti_clone_entry_text(); + /* + * This is very early in boot. Device and Late initcalls can do + * modprobe before free_initmem() and mark_readonly(). This + * pti_clone_entry_text() allows those user-mode-helpers to function, + * but notably the text is still RW. + */ + pti_clone_entry_text(false); pti_setup_espfix64(); pti_setup_vsyscall(); } @@ -656,10 +668,11 @@ void pti_finalize(void) if (!boot_cpu_has(X86_FEATURE_PTI)) return; /* - * We need to clone everything (again) that maps parts of the - * kernel image. + * This is after free_initmem() (all initcalls are done) and we've done + * mark_readonly(). Text is now NX which might've split some PMDs + * relative to the early clone. */ - pti_clone_entry_text(); + pti_clone_entry_text(true); pti_clone_kernel_text(); debug_checkwx_user(); diff --git a/drivers/acpi/acpi_processor.c b/drivers/acpi/acpi_processor.c index 6737b1cbf6d6..8bd5c4fa91f2 100644 --- a/drivers/acpi/acpi_processor.c +++ b/drivers/acpi/acpi_processor.c @@ -373,7 +373,7 @@ static int acpi_processor_add(struct acpi_device *device, result = acpi_processor_get_info(device); if (result) /* Processor is not physically present or unavailable */ - return 0; + goto err_clear_driver_data; BUG_ON(pr->id >= nr_cpu_ids); @@ -388,7 +388,7 @@ static int acpi_processor_add(struct acpi_device *device, "BIOS reported wrong ACPI id %d for the processor\n", pr->id); /* Give up, but do not abort the namespace scan. */ - goto err; + goto err_clear_driver_data; } /* * processor_device_array is not cleared on errors to allow buggy BIOS @@ -400,12 +400,12 @@ static int acpi_processor_add(struct acpi_device *device, dev = get_cpu_device(pr->id); if (!dev) { result = -ENODEV; - goto err; + goto err_clear_per_cpu; } result = acpi_bind_one(dev, device); if (result) - goto err; + goto err_clear_per_cpu; pr->dev = dev; @@ -416,10 +416,11 @@ static int acpi_processor_add(struct acpi_device *device, dev_err(dev, "Processor driver could not be attached\n"); acpi_unbind_one(dev); - err: - free_cpumask_var(pr->throttling.shared_cpu_map); - device->driver_data = NULL; + err_clear_per_cpu: per_cpu(processors, pr->id) = NULL; + err_clear_driver_data: + device->driver_data = NULL; + free_cpumask_var(pr->throttling.shared_cpu_map); err_free_pr: kfree(pr); return result; diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c index 3fa74051f31b..bfd643856f64 100644 --- a/drivers/block/ublk_drv.c +++ b/drivers/block/ublk_drv.c @@ -1915,6 +1915,8 @@ static int ublk_ctrl_start_recovery(struct ublk_device *ub, mutex_lock(&ub->mutex); if (!ublk_can_use_recovery(ub)) goto out_unlock; + if (!ub->nr_queues_ready) + goto out_unlock; /* * START_RECOVERY is only allowd after: * diff --git a/drivers/gpio/gpio-rockchip.c b/drivers/gpio/gpio-rockchip.c index 200e43a6f4b4..3c1e303aaca8 100644 --- a/drivers/gpio/gpio-rockchip.c +++ b/drivers/gpio/gpio-rockchip.c @@ -713,6 +713,7 @@ static int rockchip_gpio_probe(struct platform_device *pdev) return -ENODEV; pctldev = of_pinctrl_get(pctlnp); + of_node_put(pctlnp); if (!pctldev) return -EPROBE_DEFER; diff --git a/drivers/gpio/gpio-zynqmp-modepin.c b/drivers/gpio/gpio-zynqmp-modepin.c index a0d69387c153..2f3c9ebfa78d 100644 --- a/drivers/gpio/gpio-zynqmp-modepin.c +++ b/drivers/gpio/gpio-zynqmp-modepin.c @@ -146,6 +146,7 @@ static const struct of_device_id modepin_platform_id[] = { { .compatible = "xlnx,zynqmp-gpio-modepin", }, { } }; +MODULE_DEVICE_TABLE(of, modepin_platform_id); static struct platform_driver modepin_platform_driver = { .driver = { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index ac773b191071..cd0bccc95205 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -997,6 +997,30 @@ static int amdgpu_display_verify_sizes(struct amdgpu_framebuffer *rfb) block_width = 256 / format_info->cpp[i]; block_height = 1; block_size_log2 = 8; + } else if (AMD_FMT_MOD_GET(TILE_VERSION, modifier) >= AMD_FMT_MOD_TILE_VER_GFX12) { + int swizzle = AMD_FMT_MOD_GET(TILE, modifier); + + switch (swizzle) { + case AMD_FMT_MOD_TILE_GFX12_256B_2D: + block_size_log2 = 8; + break; + case AMD_FMT_MOD_TILE_GFX12_4K_2D: + block_size_log2 = 12; + break; + case AMD_FMT_MOD_TILE_GFX12_64K_2D: + block_size_log2 = 16; + break; + case AMD_FMT_MOD_TILE_GFX12_256K_2D: + block_size_log2 = 18; + break; + default: + drm_dbg_kms(rfb->base.dev, + "Gfx12 swizzle mode with unknown block size: %d\n", swizzle); + return -EINVAL; + } + + get_block_dimensions(block_size_log2, format_info->cpp[i], + &block_width, &block_height); } else { int swizzle = AMD_FMT_MOD_GET(TILE, modifier); @@ -1032,7 +1056,8 @@ static int amdgpu_display_verify_sizes(struct amdgpu_framebuffer *rfb) return ret; } - if (AMD_FMT_MOD_GET(DCC, modifier)) { + if (AMD_FMT_MOD_GET(TILE_VERSION, modifier) <= AMD_FMT_MOD_TILE_VER_GFX11 && + AMD_FMT_MOD_GET(DCC, modifier)) { if (AMD_FMT_MOD_GET(DCC_RETILE, modifier)) { block_size_log2 = get_dcc_block_size(modifier, false, false); get_block_dimensions(block_size_log2 + 8, format_info->cpp[0], diff --git a/drivers/gpu/drm/i915/i915_sw_fence.c b/drivers/gpu/drm/i915/i915_sw_fence.c index 6fc0d1b89690..e664f8e461e6 100644 --- a/drivers/gpu/drm/i915/i915_sw_fence.c +++ b/drivers/gpu/drm/i915/i915_sw_fence.c @@ -51,7 +51,7 @@ static inline void debug_fence_init(struct i915_sw_fence *fence) debug_object_init(fence, &i915_sw_fence_debug_descr); } -static inline void debug_fence_init_onstack(struct i915_sw_fence *fence) +static inline __maybe_unused void debug_fence_init_onstack(struct i915_sw_fence *fence) { debug_object_init_on_stack(fence, &i915_sw_fence_debug_descr); } @@ -77,7 +77,7 @@ static inline void debug_fence_destroy(struct i915_sw_fence *fence) debug_object_destroy(fence, &i915_sw_fence_debug_descr); } -static inline void debug_fence_free(struct i915_sw_fence *fence) +static inline __maybe_unused void debug_fence_free(struct i915_sw_fence *fence) { debug_object_free(fence, &i915_sw_fence_debug_descr); smp_wmb(); /* flush the change in state before reallocation */ @@ -94,7 +94,7 @@ static inline void debug_fence_init(struct i915_sw_fence *fence) { } -static inline void debug_fence_init_onstack(struct i915_sw_fence *fence) +static inline __maybe_unused void debug_fence_init_onstack(struct i915_sw_fence *fence) { } @@ -115,7 +115,7 @@ static inline void debug_fence_destroy(struct i915_sw_fence *fence) { } -static inline void debug_fence_free(struct i915_sw_fence *fence) +static inline __maybe_unused void debug_fence_free(struct i915_sw_fence *fence) { } diff --git a/drivers/net/can/spi/mcp251xfd/mcp251xfd-core.c b/drivers/net/can/spi/mcp251xfd/mcp251xfd-core.c index 1665f78abb5c..a9bafa96e2f9 100644 --- a/drivers/net/can/spi/mcp251xfd/mcp251xfd-core.c +++ b/drivers/net/can/spi/mcp251xfd/mcp251xfd-core.c @@ -2,7 +2,7 @@ // // mcp251xfd - Microchip MCP251xFD Family CAN controller driver // -// Copyright (c) 2019, 2020, 2021 Pengutronix, +// Copyright (c) 2019, 2020, 2021, 2023 Pengutronix, // Marc Kleine-Budde // // Based on: @@ -867,18 +867,18 @@ static int mcp251xfd_get_berr_counter(const struct net_device *ndev, static struct sk_buff * mcp251xfd_alloc_can_err_skb(struct mcp251xfd_priv *priv, - struct can_frame **cf, u32 *timestamp) + struct can_frame **cf, u32 *ts_raw) { struct sk_buff *skb; int err; - err = mcp251xfd_get_timestamp(priv, timestamp); + err = mcp251xfd_get_timestamp_raw(priv, ts_raw); if (err) return NULL; skb = alloc_can_err_skb(priv->ndev, cf); if (skb) - mcp251xfd_skb_set_timestamp(priv, skb, *timestamp); + mcp251xfd_skb_set_timestamp_raw(priv, skb, *ts_raw); return skb; } @@ -889,7 +889,7 @@ static int mcp251xfd_handle_rxovif(struct mcp251xfd_priv *priv) struct mcp251xfd_rx_ring *ring; struct sk_buff *skb; struct can_frame *cf; - u32 timestamp, rxovif; + u32 ts_raw, rxovif; int err, i; stats->rx_over_errors++; @@ -924,14 +924,14 @@ static int mcp251xfd_handle_rxovif(struct mcp251xfd_priv *priv) return err; } - skb = mcp251xfd_alloc_can_err_skb(priv, &cf, ×tamp); + skb = mcp251xfd_alloc_can_err_skb(priv, &cf, &ts_raw); if (!skb) return 0; cf->can_id |= CAN_ERR_CRTL; cf->data[1] = CAN_ERR_CRTL_RX_OVERFLOW; - err = can_rx_offload_queue_timestamp(&priv->offload, skb, timestamp); + err = can_rx_offload_queue_timestamp(&priv->offload, skb, ts_raw); if (err) stats->rx_fifo_errors++; @@ -948,12 +948,12 @@ static int mcp251xfd_handle_txatif(struct mcp251xfd_priv *priv) static int mcp251xfd_handle_ivmif(struct mcp251xfd_priv *priv) { struct net_device_stats *stats = &priv->ndev->stats; - u32 bdiag1, timestamp; + u32 bdiag1, ts_raw; struct sk_buff *skb; struct can_frame *cf = NULL; int err; - err = mcp251xfd_get_timestamp(priv, ×tamp); + err = mcp251xfd_get_timestamp_raw(priv, &ts_raw); if (err) return err; @@ -1035,8 +1035,8 @@ static int mcp251xfd_handle_ivmif(struct mcp251xfd_priv *priv) if (!cf) return 0; - mcp251xfd_skb_set_timestamp(priv, skb, timestamp); - err = can_rx_offload_queue_timestamp(&priv->offload, skb, timestamp); + mcp251xfd_skb_set_timestamp_raw(priv, skb, ts_raw); + err = can_rx_offload_queue_timestamp(&priv->offload, skb, ts_raw); if (err) stats->rx_fifo_errors++; @@ -1049,7 +1049,7 @@ static int mcp251xfd_handle_cerrif(struct mcp251xfd_priv *priv) struct sk_buff *skb; struct can_frame *cf = NULL; enum can_state new_state, rx_state, tx_state; - u32 trec, timestamp; + u32 trec, ts_raw; int err; err = regmap_read(priv->map_reg, MCP251XFD_REG_TREC, &trec); @@ -1079,7 +1079,7 @@ static int mcp251xfd_handle_cerrif(struct mcp251xfd_priv *priv) /* The skb allocation might fail, but can_change_state() * handles cf == NULL. */ - skb = mcp251xfd_alloc_can_err_skb(priv, &cf, ×tamp); + skb = mcp251xfd_alloc_can_err_skb(priv, &cf, &ts_raw); can_change_state(priv->ndev, cf, tx_state, rx_state); if (new_state == CAN_STATE_BUS_OFF) { @@ -1110,7 +1110,7 @@ static int mcp251xfd_handle_cerrif(struct mcp251xfd_priv *priv) cf->data[7] = bec.rxerr; } - err = can_rx_offload_queue_timestamp(&priv->offload, skb, timestamp); + err = can_rx_offload_queue_timestamp(&priv->offload, skb, ts_raw); if (err) stats->rx_fifo_errors++; diff --git a/drivers/net/can/spi/mcp251xfd/mcp251xfd-ring.c b/drivers/net/can/spi/mcp251xfd/mcp251xfd-ring.c index 915d505a304f..0fde8154a649 100644 --- a/drivers/net/can/spi/mcp251xfd/mcp251xfd-ring.c +++ b/drivers/net/can/spi/mcp251xfd/mcp251xfd-ring.c @@ -196,6 +196,7 @@ mcp251xfd_ring_init_rx(struct mcp251xfd_priv *priv, u16 *base, u8 *fifo_nr) int i, j; mcp251xfd_for_each_rx_ring(priv, rx_ring, i) { + rx_ring->last_valid = timecounter_read(&priv->tc); rx_ring->head = 0; rx_ring->tail = 0; rx_ring->base = *base; @@ -513,6 +514,8 @@ int mcp251xfd_ring_alloc(struct mcp251xfd_priv *priv) } rx_ring->obj_num = rx_obj_num; + rx_ring->obj_num_shift_to_u8 = BITS_PER_TYPE(rx_ring->obj_num_shift_to_u8) - + ilog2(rx_obj_num); rx_ring->obj_size = rx_obj_size; priv->rx[i] = rx_ring; } diff --git a/drivers/net/can/spi/mcp251xfd/mcp251xfd-rx.c b/drivers/net/can/spi/mcp251xfd/mcp251xfd-rx.c index ced8d9c81f8c..fe897f3e4c12 100644 --- a/drivers/net/can/spi/mcp251xfd/mcp251xfd-rx.c +++ b/drivers/net/can/spi/mcp251xfd/mcp251xfd-rx.c @@ -2,7 +2,7 @@ // // mcp251xfd - Microchip MCP251xFD Family CAN controller driver // -// Copyright (c) 2019, 2020, 2021 Pengutronix, +// Copyright (c) 2019, 2020, 2021, 2023 Pengutronix, // Marc Kleine-Budde // // Based on: @@ -16,23 +16,14 @@ #include "mcp251xfd.h" -static inline int -mcp251xfd_rx_head_get_from_chip(const struct mcp251xfd_priv *priv, - const struct mcp251xfd_rx_ring *ring, - u8 *rx_head, bool *fifo_empty) +static inline bool mcp251xfd_rx_fifo_sta_empty(const u32 fifo_sta) { - u32 fifo_sta; - int err; + return !(fifo_sta & MCP251XFD_REG_FIFOSTA_TFNRFNIF); +} - err = regmap_read(priv->map_reg, MCP251XFD_REG_FIFOSTA(ring->fifo_nr), - &fifo_sta); - if (err) - return err; - - *rx_head = FIELD_GET(MCP251XFD_REG_FIFOSTA_FIFOCI_MASK, fifo_sta); - *fifo_empty = !(fifo_sta & MCP251XFD_REG_FIFOSTA_TFNRFNIF); - - return 0; +static inline bool mcp251xfd_rx_fifo_sta_full(const u32 fifo_sta) +{ + return fifo_sta & MCP251XFD_REG_FIFOSTA_TFERFFIF; } static inline int @@ -80,29 +71,49 @@ mcp251xfd_check_rx_tail(const struct mcp251xfd_priv *priv, } static int -mcp251xfd_rx_ring_update(const struct mcp251xfd_priv *priv, - struct mcp251xfd_rx_ring *ring) +mcp251xfd_get_rx_len(const struct mcp251xfd_priv *priv, + const struct mcp251xfd_rx_ring *ring, + u8 *len_p) { - u32 new_head; - u8 chip_rx_head; - bool fifo_empty; + const u8 shift = ring->obj_num_shift_to_u8; + u8 chip_head, tail, len; + u32 fifo_sta; int err; - err = mcp251xfd_rx_head_get_from_chip(priv, ring, &chip_rx_head, - &fifo_empty); - if (err || fifo_empty) + err = regmap_read(priv->map_reg, MCP251XFD_REG_FIFOSTA(ring->fifo_nr), + &fifo_sta); + if (err) return err; - /* chip_rx_head, is the next RX-Object filled by the HW. - * The new RX head must be >= the old head. + if (mcp251xfd_rx_fifo_sta_empty(fifo_sta)) { + *len_p = 0; + return 0; + } + + if (mcp251xfd_rx_fifo_sta_full(fifo_sta)) { + *len_p = ring->obj_num; + return 0; + } + + chip_head = FIELD_GET(MCP251XFD_REG_FIFOSTA_FIFOCI_MASK, fifo_sta); + + err = mcp251xfd_check_rx_tail(priv, ring); + if (err) + return err; + tail = mcp251xfd_get_rx_tail(ring); + + /* First shift to full u8. The subtraction works on signed + * values, that keeps the difference steady around the u8 + * overflow. The right shift acts on len, which is an u8. */ - new_head = round_down(ring->head, ring->obj_num) + chip_rx_head; - if (new_head <= ring->head) - new_head += ring->obj_num; + BUILD_BUG_ON(sizeof(ring->obj_num) != sizeof(chip_head)); + BUILD_BUG_ON(sizeof(ring->obj_num) != sizeof(tail)); + BUILD_BUG_ON(sizeof(ring->obj_num) != sizeof(len)); - ring->head = new_head; + len = (chip_head << shift) - (tail << shift); + *len_p = len >> shift; - return mcp251xfd_check_rx_tail(priv, ring); + return 0; } static void @@ -148,8 +159,6 @@ mcp251xfd_hw_rx_obj_to_skb(const struct mcp251xfd_priv *priv, if (!(hw_rx_obj->flags & MCP251XFD_OBJ_FLAGS_RTR)) memcpy(cfd->data, hw_rx_obj->data, cfd->len); - - mcp251xfd_skb_set_timestamp(priv, skb, hw_rx_obj->ts); } static int @@ -160,8 +169,26 @@ mcp251xfd_handle_rxif_one(struct mcp251xfd_priv *priv, struct net_device_stats *stats = &priv->ndev->stats; struct sk_buff *skb; struct canfd_frame *cfd; + u64 timestamp; int err; + /* According to mcp2518fd erratum DS80000789E 6. the FIFOCI + * bits of a FIFOSTA register, here the RX FIFO head index + * might be corrupted and we might process past the RX FIFO's + * head into old CAN frames. + * + * Compare the timestamp of currently processed CAN frame with + * last valid frame received. Abort with -EBADMSG if an old + * CAN frame is detected. + */ + timestamp = timecounter_cyc2time(&priv->tc, hw_rx_obj->ts); + if (timestamp <= ring->last_valid) { + stats->rx_fifo_errors++; + + return -EBADMSG; + } + ring->last_valid = timestamp; + if (hw_rx_obj->flags & MCP251XFD_OBJ_FLAGS_FDF) skb = alloc_canfd_skb(priv->ndev, &cfd); else @@ -172,6 +199,7 @@ mcp251xfd_handle_rxif_one(struct mcp251xfd_priv *priv, return 0; } + mcp251xfd_skb_set_timestamp(skb, timestamp); mcp251xfd_hw_rx_obj_to_skb(priv, hw_rx_obj, skb); err = can_rx_offload_queue_timestamp(&priv->offload, skb, hw_rx_obj->ts); if (err) @@ -197,52 +225,81 @@ mcp251xfd_rx_obj_read(const struct mcp251xfd_priv *priv, return err; } +static int +mcp251xfd_handle_rxif_ring_uinc(const struct mcp251xfd_priv *priv, + struct mcp251xfd_rx_ring *ring, + u8 len) +{ + int offset; + int err; + + if (!len) + return 0; + + ring->head += len; + + /* Increment the RX FIFO tail pointer 'len' times in a + * single SPI message. + * + * Note: + * Calculate offset, so that the SPI transfer ends on + * the last message of the uinc_xfer array, which has + * "cs_change == 0", to properly deactivate the chip + * select. + */ + offset = ARRAY_SIZE(ring->uinc_xfer) - len; + err = spi_sync_transfer(priv->spi, + ring->uinc_xfer + offset, len); + if (err) + return err; + + ring->tail += len; + + return 0; +} + static int mcp251xfd_handle_rxif_ring(struct mcp251xfd_priv *priv, struct mcp251xfd_rx_ring *ring) { struct mcp251xfd_hw_rx_obj_canfd *hw_rx_obj = ring->obj; - u8 rx_tail, len; + u8 rx_tail, len, l; int err, i; - err = mcp251xfd_rx_ring_update(priv, ring); + err = mcp251xfd_get_rx_len(priv, ring, &len); if (err) return err; - while ((len = mcp251xfd_get_rx_linear_len(ring))) { - int offset; - + while ((l = mcp251xfd_get_rx_linear_len(ring, len))) { rx_tail = mcp251xfd_get_rx_tail(ring); err = mcp251xfd_rx_obj_read(priv, ring, hw_rx_obj, - rx_tail, len); + rx_tail, l); if (err) return err; - for (i = 0; i < len; i++) { + for (i = 0; i < l; i++) { err = mcp251xfd_handle_rxif_one(priv, ring, (void *)hw_rx_obj + i * ring->obj_size); - if (err) + + /* -EBADMSG means we're affected by mcp2518fd + * erratum DS80000789E 6., i.e. the timestamp + * in the RX object is older that the last + * valid received CAN frame. Don't process any + * further and mark processed frames as good. + */ + if (err == -EBADMSG) + return mcp251xfd_handle_rxif_ring_uinc(priv, ring, i); + else if (err) return err; } - /* Increment the RX FIFO tail pointer 'len' times in a - * single SPI message. - * - * Note: - * Calculate offset, so that the SPI transfer ends on - * the last message of the uinc_xfer array, which has - * "cs_change == 0", to properly deactivate the chip - * select. - */ - offset = ARRAY_SIZE(ring->uinc_xfer) - len; - err = spi_sync_transfer(priv->spi, - ring->uinc_xfer + offset, len); + err = mcp251xfd_handle_rxif_ring_uinc(priv, ring, l); if (err) return err; - ring->tail += len; + len -= l; } return 0; diff --git a/drivers/net/can/spi/mcp251xfd/mcp251xfd-tef.c b/drivers/net/can/spi/mcp251xfd/mcp251xfd-tef.c index 902eb767426d..8f39730f3122 100644 --- a/drivers/net/can/spi/mcp251xfd/mcp251xfd-tef.c +++ b/drivers/net/can/spi/mcp251xfd/mcp251xfd-tef.c @@ -97,7 +97,7 @@ mcp251xfd_handle_tefif_one(struct mcp251xfd_priv *priv, tef_tail = mcp251xfd_get_tef_tail(priv); skb = priv->can.echo_skb[tef_tail]; if (skb) - mcp251xfd_skb_set_timestamp(priv, skb, hw_tef_obj->ts); + mcp251xfd_skb_set_timestamp_raw(priv, skb, hw_tef_obj->ts); stats->tx_bytes += can_rx_offload_get_echo_skb(&priv->offload, tef_tail, hw_tef_obj->ts, diff --git a/drivers/net/can/spi/mcp251xfd/mcp251xfd-timestamp.c b/drivers/net/can/spi/mcp251xfd/mcp251xfd-timestamp.c index 712e09186987..1db99aabe85c 100644 --- a/drivers/net/can/spi/mcp251xfd/mcp251xfd-timestamp.c +++ b/drivers/net/can/spi/mcp251xfd/mcp251xfd-timestamp.c @@ -2,7 +2,7 @@ // // mcp251xfd - Microchip MCP251xFD Family CAN controller driver // -// Copyright (c) 2021 Pengutronix, +// Copyright (c) 2021, 2023 Pengutronix, // Marc Kleine-Budde // @@ -11,20 +11,20 @@ #include "mcp251xfd.h" -static u64 mcp251xfd_timestamp_read(const struct cyclecounter *cc) +static u64 mcp251xfd_timestamp_raw_read(const struct cyclecounter *cc) { const struct mcp251xfd_priv *priv; - u32 timestamp = 0; + u32 ts_raw = 0; int err; priv = container_of(cc, struct mcp251xfd_priv, cc); - err = mcp251xfd_get_timestamp(priv, ×tamp); + err = mcp251xfd_get_timestamp_raw(priv, &ts_raw); if (err) netdev_err(priv->ndev, "Error %d while reading timestamp. HW timestamps may be inaccurate.", err); - return timestamp; + return ts_raw; } static void mcp251xfd_timestamp_work(struct work_struct *work) @@ -39,21 +39,11 @@ static void mcp251xfd_timestamp_work(struct work_struct *work) MCP251XFD_TIMESTAMP_WORK_DELAY_SEC * HZ); } -void mcp251xfd_skb_set_timestamp(const struct mcp251xfd_priv *priv, - struct sk_buff *skb, u32 timestamp) -{ - struct skb_shared_hwtstamps *hwtstamps = skb_hwtstamps(skb); - u64 ns; - - ns = timecounter_cyc2time(&priv->tc, timestamp); - hwtstamps->hwtstamp = ns_to_ktime(ns); -} - void mcp251xfd_timestamp_init(struct mcp251xfd_priv *priv) { struct cyclecounter *cc = &priv->cc; - cc->read = mcp251xfd_timestamp_read; + cc->read = mcp251xfd_timestamp_raw_read; cc->mask = CYCLECOUNTER_MASK(32); cc->shift = 1; cc->mult = clocksource_hz2mult(priv->can.clock.freq, cc->shift); diff --git a/drivers/net/can/spi/mcp251xfd/mcp251xfd.h b/drivers/net/can/spi/mcp251xfd/mcp251xfd.h index 78d12dda08a0..c07300443c6a 100644 --- a/drivers/net/can/spi/mcp251xfd/mcp251xfd.h +++ b/drivers/net/can/spi/mcp251xfd/mcp251xfd.h @@ -2,7 +2,7 @@ * * mcp251xfd - Microchip MCP251xFD Family CAN controller driver * - * Copyright (c) 2019, 2020, 2021 Pengutronix, + * Copyright (c) 2019, 2020, 2021, 2023 Pengutronix, * Marc Kleine-Budde * Copyright (c) 2019 Martin Sperl */ @@ -549,10 +549,14 @@ struct mcp251xfd_rx_ring { unsigned int head; unsigned int tail; + /* timestamp of the last valid received CAN frame */ + u64 last_valid; + u16 base; u8 nr; u8 fifo_nr; u8 obj_num; + u8 obj_num_shift_to_u8; u8 obj_size; union mcp251xfd_write_reg_buf irq_enable_buf; @@ -793,10 +797,27 @@ mcp251xfd_spi_cmd_write(const struct mcp251xfd_priv *priv, return data; } -static inline int mcp251xfd_get_timestamp(const struct mcp251xfd_priv *priv, - u32 *timestamp) +static inline int mcp251xfd_get_timestamp_raw(const struct mcp251xfd_priv *priv, + u32 *ts_raw) { - return regmap_read(priv->map_reg, MCP251XFD_REG_TBC, timestamp); + return regmap_read(priv->map_reg, MCP251XFD_REG_TBC, ts_raw); +} + +static inline void mcp251xfd_skb_set_timestamp(struct sk_buff *skb, u64 ns) +{ + struct skb_shared_hwtstamps *hwtstamps = skb_hwtstamps(skb); + + hwtstamps->hwtstamp = ns_to_ktime(ns); +} + +static inline +void mcp251xfd_skb_set_timestamp_raw(const struct mcp251xfd_priv *priv, + struct sk_buff *skb, u32 ts_raw) +{ + u64 ns; + + ns = timecounter_cyc2time(&priv->tc, ts_raw); + mcp251xfd_skb_set_timestamp(skb, ns); } static inline u16 mcp251xfd_get_tef_obj_addr(u8 n) @@ -889,18 +910,9 @@ static inline u8 mcp251xfd_get_rx_tail(const struct mcp251xfd_rx_ring *ring) return ring->tail & (ring->obj_num - 1); } -static inline u8 mcp251xfd_get_rx_len(const struct mcp251xfd_rx_ring *ring) -{ - return ring->head - ring->tail; -} - static inline u8 -mcp251xfd_get_rx_linear_len(const struct mcp251xfd_rx_ring *ring) +mcp251xfd_get_rx_linear_len(const struct mcp251xfd_rx_ring *ring, u8 len) { - u8 len; - - len = mcp251xfd_get_rx_len(ring); - return min_t(u8, len, ring->obj_num - mcp251xfd_get_rx_tail(ring)); } @@ -926,8 +938,6 @@ void mcp251xfd_ring_free(struct mcp251xfd_priv *priv); int mcp251xfd_ring_alloc(struct mcp251xfd_priv *priv); int mcp251xfd_handle_rxif(struct mcp251xfd_priv *priv); int mcp251xfd_handle_tefif(struct mcp251xfd_priv *priv); -void mcp251xfd_skb_set_timestamp(const struct mcp251xfd_priv *priv, - struct sk_buff *skb, u32 timestamp); void mcp251xfd_timestamp_init(struct mcp251xfd_priv *priv); void mcp251xfd_timestamp_stop(struct mcp251xfd_priv *priv); diff --git a/drivers/net/ethernet/microsoft/mana/mana.h b/drivers/net/ethernet/microsoft/mana/mana.h index 41c99eabf40a..2b00d6a29117 100644 --- a/drivers/net/ethernet/microsoft/mana/mana.h +++ b/drivers/net/ethernet/microsoft/mana/mana.h @@ -86,6 +86,8 @@ struct mana_txq { atomic_t pending_sends; + bool napi_initialized; + struct mana_stats_tx stats; }; diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index e7d1ce68f05e..b52612eef0a6 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -1391,10 +1391,12 @@ static void mana_destroy_txq(struct mana_port_context *apc) for (i = 0; i < apc->num_queues; i++) { napi = &apc->tx_qp[i].tx_cq.napi; - napi_synchronize(napi); - napi_disable(napi); - netif_napi_del(napi); - + if (apc->tx_qp[i].txq.napi_initialized) { + napi_synchronize(napi); + napi_disable(napi); + netif_napi_del(napi); + apc->tx_qp[i].txq.napi_initialized = false; + } mana_destroy_wq_obj(apc, GDMA_SQ, apc->tx_qp[i].tx_object); mana_deinit_cq(apc, &apc->tx_qp[i].tx_cq); @@ -1450,6 +1452,7 @@ static int mana_create_txq(struct mana_port_context *apc, txq->ndev = net; txq->net_txq = netdev_get_tx_queue(net, i); txq->vp_offset = apc->tx_vp_offset; + txq->napi_initialized = false; skb_queue_head_init(&txq->pending_skbs); memset(&spec, 0, sizeof(spec)); @@ -1514,6 +1517,7 @@ static int mana_create_txq(struct mana_port_context *apc, netif_napi_add_tx(net, &cq->napi, mana_poll); napi_enable(&cq->napi); + txq->napi_initialized = true; mana_gd_ring_cq(cq->gdma_cq, SET_ARM_BIT); } @@ -1525,7 +1529,7 @@ out: } static void mana_destroy_rxq(struct mana_port_context *apc, - struct mana_rxq *rxq, bool validate_state) + struct mana_rxq *rxq, bool napi_initialized) { struct gdma_context *gc = apc->ac->gdma_dev->gdma_context; @@ -1539,15 +1543,15 @@ static void mana_destroy_rxq(struct mana_port_context *apc, napi = &rxq->rx_cq.napi; - if (validate_state) + if (napi_initialized) { napi_synchronize(napi); - napi_disable(napi); + napi_disable(napi); + netif_napi_del(napi); + } xdp_rxq_info_unreg(&rxq->xdp_rxq); - netif_napi_del(napi); - mana_destroy_wq_obj(apc, GDMA_RQ, rxq->rxobj); mana_deinit_cq(apc, &rxq->rx_cq); diff --git a/drivers/nvme/target/tcp.c b/drivers/nvme/target/tcp.c index 76b9eb438268..81574500a57c 100644 --- a/drivers/nvme/target/tcp.c +++ b/drivers/nvme/target/tcp.c @@ -1816,8 +1816,10 @@ static u16 nvmet_tcp_install_queue(struct nvmet_sq *sq) } queue->nr_cmds = sq->size * 2; - if (nvmet_tcp_alloc_cmds(queue)) + if (nvmet_tcp_alloc_cmds(queue)) { + queue->nr_cmds = 0; return NVME_SC_INTERNAL; + } return 0; } diff --git a/drivers/regulator/of_regulator.c b/drivers/regulator/of_regulator.c index c3571781a8cc..1b65e5e4e40f 100644 --- a/drivers/regulator/of_regulator.c +++ b/drivers/regulator/of_regulator.c @@ -767,7 +767,7 @@ restart: memcpy(name, prop->name, i); name[i] = '\0'; tmp = regulator_get(dev, name); - if (!tmp) { + if (IS_ERR(tmp)) { ret = -EINVAL; goto error; } diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 853b1f96b1fd..cca1acf2e037 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1553,7 +1553,6 @@ struct btrfs_drop_extents_args { struct btrfs_file_private { void *filldir_buf; u64 last_index; - bool fsync_skip_inode_lock; }; diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index e23d178f9778..c8231677c79e 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1534,13 +1534,6 @@ again: if (IS_ERR_OR_NULL(dio)) { err = PTR_ERR_OR_ZERO(dio); } else { - struct btrfs_file_private stack_private = { 0 }; - struct btrfs_file_private *private; - const bool have_private = (file->private_data != NULL); - - if (!have_private) - file->private_data = &stack_private; - /* * If we have a synchoronous write, we must make sure the fsync * triggered by the iomap_dio_complete() call below doesn't @@ -1549,13 +1542,10 @@ again: * partial writes due to the input buffer (or parts of it) not * being already faulted in. */ - private = file->private_data; - private->fsync_skip_inode_lock = true; + ASSERT(current->journal_info == NULL); + current->journal_info = BTRFS_TRANS_DIO_WRITE_STUB; err = iomap_dio_complete(dio); - private->fsync_skip_inode_lock = false; - - if (!have_private) - file->private_data = NULL; + current->journal_info = NULL; } /* No increment (+=) because iomap returns a cumulative value. */ @@ -1795,7 +1785,6 @@ static inline bool skip_inode_logging(const struct btrfs_log_ctx *ctx) */ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) { - struct btrfs_file_private *private = file->private_data; struct dentry *dentry = file_dentry(file); struct inode *inode = d_inode(dentry); struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); @@ -1805,7 +1794,13 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) int ret = 0, err; u64 len; bool full_sync; - const bool skip_ilock = (private ? private->fsync_skip_inode_lock : false); + bool skip_ilock = false; + + if (current->journal_info == BTRFS_TRANS_DIO_WRITE_STUB) { + skip_ilock = true; + current->journal_info = NULL; + lockdep_assert_held(&inode->i_rwsem); + } trace_btrfs_sync_file(file, datasync); diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 970ff316069d..8b88446df36d 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -11,6 +11,12 @@ #include "delayed-ref.h" #include "ctree.h" +/* + * Signal that a direct IO write is in progress, to avoid deadlock for sync + * direct IO writes when fsync is called during the direct IO write path. + */ +#define BTRFS_TRANS_DIO_WRITE_STUB ((void *) 1) + enum btrfs_trans_state { TRANS_STATE_RUNNING, TRANS_STATE_COMMIT_START, diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 2ebcd8d30bbd..8e216e004898 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -1506,7 +1506,8 @@ void fuse_send_init(struct fuse_mount *fm) FUSE_ABORT_ERROR | FUSE_MAX_PAGES | FUSE_CACHE_SYMLINKS | FUSE_NO_OPENDIR_SUPPORT | FUSE_EXPLICIT_INVAL_DATA | FUSE_HANDLE_KILLPRIV_V2 | FUSE_SETXATTR_EXT | FUSE_INIT_EXT | - FUSE_SECURITY_CTX; + FUSE_SECURITY_CTX | + FUSE_HAS_EXPIRE_ONLY; #ifdef CONFIG_FUSE_DAX if (fm->fc->dax) flags |= FUSE_MAP_ALIGNMENT; diff --git a/include/uapi/drm/drm_fourcc.h b/include/uapi/drm/drm_fourcc.h index 7fc523147eb8..0d4535aa8258 100644 --- a/include/uapi/drm/drm_fourcc.h +++ b/include/uapi/drm/drm_fourcc.h @@ -1392,6 +1392,7 @@ drm_fourcc_canonicalize_nvidia_format_mod(__u64 modifier) #define AMD_FMT_MOD_TILE_VER_GFX10 2 #define AMD_FMT_MOD_TILE_VER_GFX10_RBPLUS 3 #define AMD_FMT_MOD_TILE_VER_GFX11 4 +#define AMD_FMT_MOD_TILE_VER_GFX12 5 /* * 64K_S is the same for GFX9/GFX10/GFX10_RBPLUS and hence has GFX9 as canonical @@ -1402,6 +1403,8 @@ drm_fourcc_canonicalize_nvidia_format_mod(__u64 modifier) /* * 64K_D for non-32 bpp is the same for GFX9/GFX10/GFX10_RBPLUS and hence has * GFX9 as canonical version. + * + * 64K_D_2D on GFX12 is identical to 64K_D on GFX11. */ #define AMD_FMT_MOD_TILE_GFX9_64K_D 10 #define AMD_FMT_MOD_TILE_GFX9_64K_S_X 25 @@ -1409,6 +1412,21 @@ drm_fourcc_canonicalize_nvidia_format_mod(__u64 modifier) #define AMD_FMT_MOD_TILE_GFX9_64K_R_X 27 #define AMD_FMT_MOD_TILE_GFX11_256K_R_X 31 +/* Gfx12 swizzle modes: + * 0 - LINEAR + * 1 - 256B_2D - 2D block dimensions + * 2 - 4KB_2D + * 3 - 64KB_2D + * 4 - 256KB_2D + * 5 - 4KB_3D - 3D block dimensions + * 6 - 64KB_3D + * 7 - 256KB_3D + */ +#define AMD_FMT_MOD_TILE_GFX12_256B_2D 1 +#define AMD_FMT_MOD_TILE_GFX12_4K_2D 2 +#define AMD_FMT_MOD_TILE_GFX12_64K_2D 3 +#define AMD_FMT_MOD_TILE_GFX12_256K_2D 4 + #define AMD_FMT_MOD_DCC_BLOCK_64B 0 #define AMD_FMT_MOD_DCC_BLOCK_128B 1 #define AMD_FMT_MOD_DCC_BLOCK_256B 2 diff --git a/include/uapi/linux/fuse.h b/include/uapi/linux/fuse.h index f0cb03e3d877..1fcd77f89805 100644 --- a/include/uapi/linux/fuse.h +++ b/include/uapi/linux/fuse.h @@ -365,6 +365,7 @@ struct fuse_file_lock { * FUSE_SECURITY_CTX: add security context to create, mkdir, symlink, and * mknod * FUSE_HAS_INODE_DAX: use per inode DAX + * FUSE_HAS_EXPIRE_ONLY: kernel supports expiry-only entry invalidation */ #define FUSE_ASYNC_READ (1 << 0) #define FUSE_POSIX_LOCKS (1 << 1) @@ -401,6 +402,7 @@ struct fuse_file_lock { /* bits 32..63 get shifted down 32 bits into the flags2 field */ #define FUSE_SECURITY_CTX (1ULL << 32) #define FUSE_HAS_INODE_DAX (1ULL << 33) +#define FUSE_HAS_EXPIRE_ONLY (1ULL << 35) /* * For FUSE < 7.36 FUSE_PASSTHROUGH has value (1 << 31). diff --git a/kernel/bpf/btf.c b/kernel/bpf/btf.c index 0cc1a6ad623a..bffe0ba5537c 100644 --- a/kernel/bpf/btf.c +++ b/kernel/bpf/btf.c @@ -467,10 +467,16 @@ static bool btf_type_is_fwd(const struct btf_type *t) return BTF_INFO_KIND(t->info) == BTF_KIND_FWD; } +static bool btf_type_is_decl_tag(const struct btf_type *t) +{ + return BTF_INFO_KIND(t->info) == BTF_KIND_DECL_TAG; +} + static bool btf_type_nosize(const struct btf_type *t) { return btf_type_is_void(t) || btf_type_is_fwd(t) || - btf_type_is_func(t) || btf_type_is_func_proto(t); + btf_type_is_func(t) || btf_type_is_func_proto(t) || + btf_type_is_decl_tag(t); } static bool btf_type_nosize_or_null(const struct btf_type *t) @@ -493,11 +499,6 @@ static bool btf_type_is_datasec(const struct btf_type *t) return BTF_INFO_KIND(t->info) == BTF_KIND_DATASEC; } -static bool btf_type_is_decl_tag(const struct btf_type *t) -{ - return BTF_INFO_KIND(t->info) == BTF_KIND_DECL_TAG; -} - static bool btf_type_is_decl_tag_target(const struct btf_type *t) { return btf_type_is_func(t) || btf_type_is_struct(t) || diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 98ce3d2b1eb5..a145afb0952f 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -5896,10 +5896,18 @@ static void wq_watchdog_timer_fn(struct timer_list *unused) notrace void wq_watchdog_touch(int cpu) { - if (cpu >= 0) - per_cpu(wq_watchdog_touched_cpu, cpu) = jiffies; + unsigned long thresh = READ_ONCE(wq_watchdog_thresh) * HZ; + unsigned long touch_ts = READ_ONCE(wq_watchdog_touched); + unsigned long now = jiffies; - wq_watchdog_touched = jiffies; + if (cpu >= 0) + per_cpu(wq_watchdog_touched_cpu, cpu) = now; + else + WARN_ONCE(1, "%s should be called with valid CPU", __func__); + + /* Don't unnecessarily store to global cacheline */ + if (time_after(now, touch_ts + thresh / 4)) + WRITE_ONCE(wq_watchdog_touched, jiffies); } static void wq_watchdog_set_thresh(unsigned long thresh) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 58930c982e6a..9adbee3fd5e8 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -5156,12 +5156,29 @@ static struct cftype mem_cgroup_legacy_files[] = { */ static DEFINE_IDR(mem_cgroup_idr); +static DEFINE_SPINLOCK(memcg_idr_lock); + +static int mem_cgroup_alloc_id(void) +{ + int ret; + + idr_preload(GFP_KERNEL); + spin_lock(&memcg_idr_lock); + ret = idr_alloc(&mem_cgroup_idr, NULL, 1, MEM_CGROUP_ID_MAX + 1, + GFP_NOWAIT); + spin_unlock(&memcg_idr_lock); + idr_preload_end(); + return ret; +} static void mem_cgroup_id_remove(struct mem_cgroup *memcg) { if (memcg->id.id > 0) { trace_android_vh_mem_cgroup_id_remove(memcg); + spin_lock(&memcg_idr_lock); idr_remove(&mem_cgroup_idr, memcg->id.id); + spin_unlock(&memcg_idr_lock); + memcg->id.id = 0; } } @@ -5286,8 +5303,7 @@ static struct mem_cgroup *mem_cgroup_alloc(void) if (!memcg) return ERR_PTR(error); - memcg->id.id = idr_alloc(&mem_cgroup_idr, NULL, - 1, MEM_CGROUP_ID_MAX + 1, GFP_KERNEL); + memcg->id.id = mem_cgroup_alloc_id(); if (memcg->id.id < 0) { error = memcg->id.id; goto fail; @@ -5332,7 +5348,9 @@ static struct mem_cgroup *mem_cgroup_alloc(void) INIT_LIST_HEAD(&memcg->deferred_split_queue.split_queue); memcg->deferred_split_queue.split_queue_len = 0; #endif + spin_lock(&memcg_idr_lock); idr_replace(&mem_cgroup_idr, memcg, memcg->id.id); + spin_unlock(&memcg_idr_lock); lru_gen_init_memcg(memcg); trace_android_vh_mem_cgroup_alloc(memcg); return memcg; diff --git a/sound/soc/sof/topology.c b/sound/soc/sof/topology.c index e7305ce57ea1..374c8b1d6958 100644 --- a/sound/soc/sof/topology.c +++ b/sound/soc/sof/topology.c @@ -1817,6 +1817,8 @@ static int sof_link_unload(struct snd_soc_component *scomp, struct snd_soc_dobj if (!slink) return 0; + slink->link->platforms->name = NULL; + kfree(slink->tuples); list_del(&slink->list); kfree(slink->hw_configs); diff --git a/sound/soc/sunxi/sun4i-i2s.c b/sound/soc/sunxi/sun4i-i2s.c index 6028871825ba..47faaf849de0 100644 --- a/sound/soc/sunxi/sun4i-i2s.c +++ b/sound/soc/sunxi/sun4i-i2s.c @@ -100,8 +100,8 @@ #define SUN8I_I2S_CTRL_MODE_PCM (0 << 4) #define SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK BIT(19) -#define SUN8I_I2S_FMT0_LRCLK_POLARITY_INVERTED (1 << 19) -#define SUN8I_I2S_FMT0_LRCLK_POLARITY_NORMAL (0 << 19) +#define SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH (1 << 19) +#define SUN8I_I2S_FMT0_LRCLK_POLARITY_START_LOW (0 << 19) #define SUN8I_I2S_FMT0_LRCK_PERIOD_MASK GENMASK(17, 8) #define SUN8I_I2S_FMT0_LRCK_PERIOD(period) ((period - 1) << 8) #define SUN8I_I2S_FMT0_BCLK_POLARITY_MASK BIT(7) @@ -727,65 +727,37 @@ static int sun4i_i2s_set_soc_fmt(const struct sun4i_i2s *i2s, static int sun8i_i2s_set_soc_fmt(const struct sun4i_i2s *i2s, unsigned int fmt) { - u32 mode, val; + u32 mode, lrclk_pol, bclk_pol, val; u8 offset; - /* - * DAI clock polarity - * - * The setup for LRCK contradicts the datasheet, but under a - * scope it's clear that the LRCK polarity is reversed - * compared to the expected polarity on the bus. - */ - switch (fmt & SND_SOC_DAIFMT_INV_MASK) { - case SND_SOC_DAIFMT_IB_IF: - /* Invert both clocks */ - val = SUN8I_I2S_FMT0_BCLK_POLARITY_INVERTED; - break; - case SND_SOC_DAIFMT_IB_NF: - /* Invert bit clock */ - val = SUN8I_I2S_FMT0_BCLK_POLARITY_INVERTED | - SUN8I_I2S_FMT0_LRCLK_POLARITY_INVERTED; - break; - case SND_SOC_DAIFMT_NB_IF: - /* Invert frame clock */ - val = 0; - break; - case SND_SOC_DAIFMT_NB_NF: - val = SUN8I_I2S_FMT0_LRCLK_POLARITY_INVERTED; - break; - default: - return -EINVAL; - } - - regmap_update_bits(i2s->regmap, SUN4I_I2S_FMT0_REG, - SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK | - SUN8I_I2S_FMT0_BCLK_POLARITY_MASK, - val); - /* DAI Mode */ switch (fmt & SND_SOC_DAIFMT_FORMAT_MASK) { case SND_SOC_DAIFMT_DSP_A: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH; mode = SUN8I_I2S_CTRL_MODE_PCM; offset = 1; break; case SND_SOC_DAIFMT_DSP_B: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH; mode = SUN8I_I2S_CTRL_MODE_PCM; offset = 0; break; case SND_SOC_DAIFMT_I2S: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_LOW; mode = SUN8I_I2S_CTRL_MODE_LEFT; offset = 1; break; case SND_SOC_DAIFMT_LEFT_J: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH; mode = SUN8I_I2S_CTRL_MODE_LEFT; offset = 0; break; case SND_SOC_DAIFMT_RIGHT_J: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH; mode = SUN8I_I2S_CTRL_MODE_RIGHT; offset = 0; break; @@ -803,6 +775,35 @@ static int sun8i_i2s_set_soc_fmt(const struct sun4i_i2s *i2s, SUN8I_I2S_TX_CHAN_OFFSET_MASK, SUN8I_I2S_TX_CHAN_OFFSET(offset)); + /* DAI clock polarity */ + bclk_pol = SUN8I_I2S_FMT0_BCLK_POLARITY_NORMAL; + + switch (fmt & SND_SOC_DAIFMT_INV_MASK) { + case SND_SOC_DAIFMT_IB_IF: + /* Invert both clocks */ + lrclk_pol ^= SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK; + bclk_pol = SUN8I_I2S_FMT0_BCLK_POLARITY_INVERTED; + break; + case SND_SOC_DAIFMT_IB_NF: + /* Invert bit clock */ + bclk_pol = SUN8I_I2S_FMT0_BCLK_POLARITY_INVERTED; + break; + case SND_SOC_DAIFMT_NB_IF: + /* Invert frame clock */ + lrclk_pol ^= SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK; + break; + case SND_SOC_DAIFMT_NB_NF: + /* No inversion */ + break; + default: + return -EINVAL; + } + + regmap_update_bits(i2s->regmap, SUN4I_I2S_FMT0_REG, + SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK | + SUN8I_I2S_FMT0_BCLK_POLARITY_MASK, + lrclk_pol | bclk_pol); + /* DAI clock master masks */ switch (fmt & SND_SOC_DAIFMT_CLOCK_PROVIDER_MASK) { case SND_SOC_DAIFMT_BP_FP: @@ -834,65 +835,37 @@ static int sun8i_i2s_set_soc_fmt(const struct sun4i_i2s *i2s, static int sun50i_h6_i2s_set_soc_fmt(const struct sun4i_i2s *i2s, unsigned int fmt) { - u32 mode, val; + u32 mode, lrclk_pol, bclk_pol, val; u8 offset; - /* - * DAI clock polarity - * - * The setup for LRCK contradicts the datasheet, but under a - * scope it's clear that the LRCK polarity is reversed - * compared to the expected polarity on the bus. - */ - switch (fmt & SND_SOC_DAIFMT_INV_MASK) { - case SND_SOC_DAIFMT_IB_IF: - /* Invert both clocks */ - val = SUN8I_I2S_FMT0_BCLK_POLARITY_INVERTED; - break; - case SND_SOC_DAIFMT_IB_NF: - /* Invert bit clock */ - val = SUN8I_I2S_FMT0_BCLK_POLARITY_INVERTED | - SUN8I_I2S_FMT0_LRCLK_POLARITY_INVERTED; - break; - case SND_SOC_DAIFMT_NB_IF: - /* Invert frame clock */ - val = 0; - break; - case SND_SOC_DAIFMT_NB_NF: - val = SUN8I_I2S_FMT0_LRCLK_POLARITY_INVERTED; - break; - default: - return -EINVAL; - } - - regmap_update_bits(i2s->regmap, SUN4I_I2S_FMT0_REG, - SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK | - SUN8I_I2S_FMT0_BCLK_POLARITY_MASK, - val); - /* DAI Mode */ switch (fmt & SND_SOC_DAIFMT_FORMAT_MASK) { case SND_SOC_DAIFMT_DSP_A: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH; mode = SUN8I_I2S_CTRL_MODE_PCM; offset = 1; break; case SND_SOC_DAIFMT_DSP_B: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH; mode = SUN8I_I2S_CTRL_MODE_PCM; offset = 0; break; case SND_SOC_DAIFMT_I2S: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_LOW; mode = SUN8I_I2S_CTRL_MODE_LEFT; offset = 1; break; case SND_SOC_DAIFMT_LEFT_J: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH; mode = SUN8I_I2S_CTRL_MODE_LEFT; offset = 0; break; case SND_SOC_DAIFMT_RIGHT_J: + lrclk_pol = SUN8I_I2S_FMT0_LRCLK_POLARITY_START_HIGH; mode = SUN8I_I2S_CTRL_MODE_RIGHT; offset = 0; break; @@ -910,6 +883,36 @@ static int sun50i_h6_i2s_set_soc_fmt(const struct sun4i_i2s *i2s, SUN50I_H6_I2S_TX_CHAN_SEL_OFFSET_MASK, SUN50I_H6_I2S_TX_CHAN_SEL_OFFSET(offset)); + /* DAI clock polarity */ + bclk_pol = SUN8I_I2S_FMT0_BCLK_POLARITY_NORMAL; + + switch (fmt & SND_SOC_DAIFMT_INV_MASK) { + case SND_SOC_DAIFMT_IB_IF: + /* Invert both clocks */ + lrclk_pol ^= SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK; + bclk_pol = SUN8I_I2S_FMT0_BCLK_POLARITY_INVERTED; + break; + case SND_SOC_DAIFMT_IB_NF: + /* Invert bit clock */ + bclk_pol = SUN8I_I2S_FMT0_BCLK_POLARITY_INVERTED; + break; + case SND_SOC_DAIFMT_NB_IF: + /* Invert frame clock */ + lrclk_pol ^= SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK; + break; + case SND_SOC_DAIFMT_NB_NF: + /* No inversion */ + break; + default: + return -EINVAL; + } + + regmap_update_bits(i2s->regmap, SUN4I_I2S_FMT0_REG, + SUN8I_I2S_FMT0_LRCLK_POLARITY_MASK | + SUN8I_I2S_FMT0_BCLK_POLARITY_MASK, + lrclk_pol | bclk_pol); + + /* DAI clock master masks */ switch (fmt & SND_SOC_DAIFMT_CLOCK_PROVIDER_MASK) { case SND_SOC_DAIFMT_BP_FP: diff --git a/sound/soc/tegra/tegra210_ahub.c b/sound/soc/tegra/tegra210_ahub.c index b38d205b69cc..dfdcb4580cd7 100644 --- a/sound/soc/tegra/tegra210_ahub.c +++ b/sound/soc/tegra/tegra210_ahub.c @@ -2,7 +2,7 @@ // // tegra210_ahub.c - Tegra210 AHUB driver // -// Copyright (c) 2020-2022, NVIDIA CORPORATION. All rights reserved. +// Copyright (c) 2020-2024, NVIDIA CORPORATION. All rights reserved. #include #include @@ -1401,12 +1401,14 @@ static int tegra_ahub_probe(struct platform_device *pdev) return err; } - err = of_platform_populate(pdev->dev.of_node, NULL, NULL, &pdev->dev); - if (err) - return err; - pm_runtime_enable(&pdev->dev); + err = of_platform_populate(pdev->dev.of_node, NULL, NULL, &pdev->dev); + if (err) { + pm_runtime_disable(&pdev->dev); + return err; + } + return 0; }