diff options
Diffstat (limited to '')
50 files changed, 2496 insertions, 1338 deletions
diff --git a/arch/parisc/kernel/Makefile b/arch/parisc/kernel/Makefile index ed0b87908d71..3d138c9cf9ce 100644 --- a/arch/parisc/kernel/Makefile +++ b/arch/parisc/kernel/Makefile @@ -3,14 +3,14 @@ # Makefile for arch/parisc/kernel # -extra-y := head.o vmlinux.lds +extra-y := vmlinux.lds -obj-y := cache.o pacache.o setup.o pdt.o traps.o time.o irq.o \ +obj-y := head.o cache.o pacache.o setup.o pdt.o traps.o time.o irq.o \ pa7300lc.o syscall.o entry.o sys_parisc.o firmware.o \ ptrace.o hardware.o inventory.o drivers.o alternative.o \ signal.o hpmc.o real2.o parisc_ksyms.o unaligned.o \ process.o processor.o pdc_cons.o pdc_chassis.o unwind.o \ - patch.o + patch.o toc.o toc_asm.o ifdef CONFIG_FUNCTION_TRACER # Do not profile debug and lowlevel utilities @@ -31,7 +31,7 @@ obj-$(CONFIG_AUDIT) += audit.o obj64-$(CONFIG_AUDIT) += compat_audit.o # only supported for PCX-W/U in 64-bit mode at the moment obj-$(CONFIG_64BIT) += perf.o perf_asm.o $(obj64-y) -obj-$(CONFIG_PARISC_CPU_TOPOLOGY) += topology.o +obj-$(CONFIG_GENERIC_ARCH_TOPOLOGY) += topology.o obj-$(CONFIG_FUNCTION_TRACER) += ftrace.o obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o obj-$(CONFIG_JUMP_LABEL) += jump_label.o @@ -39,4 +39,8 @@ obj-$(CONFIG_KGDB) += kgdb.o obj-$(CONFIG_KPROBES) += kprobes.o obj-$(CONFIG_KEXEC_CORE) += kexec.o relocate_kernel.o obj-$(CONFIG_KEXEC_FILE) += kexec_file.o -obj-$(CONFIG_TOC) += toc.o toc_asm.o + +# vdso +obj-y += vdso.o +obj-$(CONFIG_64BIT) += vdso64/ +obj-y += vdso32/ diff --git a/arch/parisc/kernel/alternative.c b/arch/parisc/kernel/alternative.c index fa28c4c9f972..66f5672c70bd 100644 --- a/arch/parisc/kernel/alternative.c +++ b/arch/parisc/kernel/alternative.c @@ -8,6 +8,7 @@ #include <asm/processor.h> #include <asm/sections.h> #include <asm/alternative.h> +#include <asm/cacheflush.h> #include <linux/module.h> @@ -25,7 +26,7 @@ void __init_or_module apply_alternatives(struct alt_instr *start, struct alt_instr *entry; int index = 0, applied = 0; int num_cpus = num_online_cpus(); - u32 cond_check; + u16 cond_check; cond_check = ALT_COND_ALWAYS | ((num_cpus == 1) ? ALT_COND_NO_SMP : 0) | @@ -44,8 +45,9 @@ void __init_or_module apply_alternatives(struct alt_instr *start, for (entry = start; entry < end; entry++, index++) { - u32 *from, cond, replacement; - s32 len; + u32 *from, replacement; + u16 cond; + s16 len; from = (u32 *)((ulong)&entry->orig_offset + entry->orig_offset); len = entry->len; @@ -107,5 +109,14 @@ void __init apply_alternatives_all(void) apply_alternatives((struct alt_instr *) &__alt_instructions, (struct alt_instr *) &__alt_instructions_end, NULL); + if (cache_info.dc_size == 0 && cache_info.ic_size == 0) { + pr_info("alternatives: optimizing cache-flushes.\n"); + static_branch_disable(&parisc_has_cache); + } + if (cache_info.dc_size == 0) + static_branch_disable(&parisc_has_dcache); + if (cache_info.ic_size == 0) + static_branch_disable(&parisc_has_icache); + set_kernel_text_rw(0); } diff --git a/arch/parisc/kernel/asm-offsets.c b/arch/parisc/kernel/asm-offsets.c index 55c1c5189c6a..94652e13c260 100644 --- a/arch/parisc/kernel/asm-offsets.c +++ b/arch/parisc/kernel/asm-offsets.c @@ -26,7 +26,11 @@ #include <asm/ptrace.h> #include <asm/processor.h> #include <asm/pdc.h> +#include <uapi/asm/sigcontext.h> +#include <asm/ucontext.h> +#include <asm/rt_sigframe.h> #include <linux/uaccess.h> +#include "signal32.h" /* Add FRAME_SIZE to the size x and align it to y. All definitions * that use align_frame will include space for a frame. @@ -36,7 +40,11 @@ int main(void) { DEFINE(TASK_TI_FLAGS, offsetof(struct task_struct, thread_info.flags)); +#ifdef CONFIG_SMP + DEFINE(TASK_TI_CPU, offsetof(struct task_struct, thread_info.cpu)); +#endif DEFINE(TASK_STACK, offsetof(struct task_struct, stack)); + DEFINE(TASK_PAGEFAULT_DISABLED, offsetof(struct task_struct, pagefault_disabled)); BLANK(); DEFINE(TASK_REGS, offsetof(struct task_struct, thread.regs)); DEFINE(TASK_PT_PSW, offsetof(struct task_struct, thread.regs.gr[ 0])); @@ -214,6 +222,16 @@ int main(void) DEFINE(TI_FLAGS, offsetof(struct thread_info, flags)); DEFINE(TI_PRE_COUNT, offsetof(struct task_struct, thread_info.preempt_count)); BLANK(); + DEFINE(ASM_SIGFRAME_SIZE, PARISC_RT_SIGFRAME_SIZE); + DEFINE(SIGFRAME_CONTEXT_REGS, offsetof(struct rt_sigframe, uc.uc_mcontext) - PARISC_RT_SIGFRAME_SIZE); +#ifdef CONFIG_64BIT + DEFINE(ASM_SIGFRAME_SIZE32, PARISC_RT_SIGFRAME_SIZE32); + DEFINE(SIGFRAME_CONTEXT_REGS32, offsetof(struct compat_rt_sigframe, uc.uc_mcontext) - PARISC_RT_SIGFRAME_SIZE32); +#else + DEFINE(ASM_SIGFRAME_SIZE32, PARISC_RT_SIGFRAME_SIZE); + DEFINE(SIGFRAME_CONTEXT_REGS32, offsetof(struct rt_sigframe, uc.uc_mcontext) - PARISC_RT_SIGFRAME_SIZE); +#endif + BLANK(); DEFINE(ICACHE_BASE, offsetof(struct pdc_cache_info, ic_base)); DEFINE(ICACHE_STRIDE, offsetof(struct pdc_cache_info, ic_stride)); DEFINE(ICACHE_COUNT, offsetof(struct pdc_cache_info, ic_count)); diff --git a/arch/parisc/kernel/cache.c b/arch/parisc/kernel/cache.c index 94150b91c96f..1d3b8bc8a623 100644 --- a/arch/parisc/kernel/cache.c +++ b/arch/parisc/kernel/cache.c @@ -27,6 +27,7 @@ #include <asm/processor.h> #include <asm/sections.h> #include <asm/shmparam.h> +#include <asm/mmu_context.h> int split_tlb __ro_after_init; int dcache_stride __ro_after_init; @@ -38,6 +39,9 @@ EXPORT_SYMBOL(flush_dcache_page_asm); void purge_dcache_page_asm(unsigned long phys_addr, unsigned long vaddr); void flush_icache_page_asm(unsigned long phys_addr, unsigned long vaddr); +/* Internal implementation in arch/parisc/kernel/pacache.S */ +void flush_data_cache_local(void *); /* flushes local data-cache only */ +void flush_instruction_cache_local(void); /* flushes local code-cache only */ /* On some machines (i.e., ones with the Merced bus), there can be * only a single PxTLB broadcast at a time; this must be guaranteed @@ -46,9 +50,6 @@ void flush_icache_page_asm(unsigned long phys_addr, unsigned long vaddr); */ DEFINE_SPINLOCK(pa_tlb_flush_lock); -/* Swapper page setup lock. */ -DEFINE_SPINLOCK(pa_swapper_pg_lock); - #if defined(CONFIG_64BIT) && defined(CONFIG_SMP) int pa_serialize_tlb_flushes __ro_after_init; #endif @@ -58,28 +59,37 @@ struct pdc_cache_info cache_info __ro_after_init; static struct pdc_btlb_info btlb_info __ro_after_init; #endif -#ifdef CONFIG_SMP -void -flush_data_cache(void) +DEFINE_STATIC_KEY_TRUE(parisc_has_cache); +DEFINE_STATIC_KEY_TRUE(parisc_has_dcache); +DEFINE_STATIC_KEY_TRUE(parisc_has_icache); + +static void cache_flush_local_cpu(void *dummy) { - on_each_cpu(flush_data_cache_local, NULL, 1); + if (static_branch_likely(&parisc_has_icache)) + flush_instruction_cache_local(); + if (static_branch_likely(&parisc_has_dcache)) + flush_data_cache_local(NULL); } -void -flush_instruction_cache(void) + +void flush_cache_all_local(void) { - on_each_cpu(flush_instruction_cache_local, NULL, 1); + cache_flush_local_cpu(NULL); } -#endif -void -flush_cache_all_local(void) +void flush_cache_all(void) +{ + if (static_branch_likely(&parisc_has_cache)) + on_each_cpu(cache_flush_local_cpu, NULL, 1); +} + +static inline void flush_data_cache(void) { - flush_instruction_cache_local(NULL); - flush_data_cache_local(NULL); + if (static_branch_likely(&parisc_has_dcache)) + on_each_cpu(flush_data_cache_local, NULL, 1); } -EXPORT_SYMBOL(flush_cache_all_local); -/* Virtual address of pfn. */ + +/* Kernel virtual address of pfn. */ #define pfn_va(pfn) __va(PFN_PHYS(pfn)) void @@ -112,11 +122,13 @@ show_cache_info(struct seq_file *m) cache_info.ic_size/1024 ); if (cache_info.dc_loop != 1) snprintf(buf, 32, "%lu-way associative", cache_info.dc_loop); - seq_printf(m, "D-cache\t\t: %ld KB (%s%s, %s)\n", + seq_printf(m, "D-cache\t\t: %ld KB (%s%s, %s, alias=%d)\n", cache_info.dc_size/1024, (cache_info.dc_conf.cc_wt ? "WT":"WB"), (cache_info.dc_conf.cc_sh ? ", shared I/D":""), - ((cache_info.dc_loop == 1) ? "direct mapped" : buf)); + ((cache_info.dc_loop == 1) ? "direct mapped" : buf), + cache_info.dc_conf.cc_alias + ); seq_printf(m, "ITLB entries\t: %ld\n" "DTLB entries\t: %ld%s\n", cache_info.it_size, cache_info.dt_size, @@ -261,7 +273,7 @@ parisc_cache_init(void) } } -void __init disable_sr_hashing(void) +void disable_sr_hashing(void) { int srhash_type, retval; unsigned long space_bits; @@ -303,6 +315,8 @@ static inline void __flush_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, unsigned long physaddr) { + if (!static_branch_likely(&parisc_has_cache)) + return; preempt_disable(); flush_dcache_page_asm(physaddr, vmaddr); if (vma->vm_flags & VM_EXEC) @@ -310,23 +324,81 @@ __flush_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, preempt_enable(); } -static inline void -__purge_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, - unsigned long physaddr) +static void flush_user_cache_page(struct vm_area_struct *vma, unsigned long vmaddr) { + unsigned long flags, space, pgd, prot; +#ifdef CONFIG_TLB_PTLOCK + unsigned long pgd_lock; +#endif + + vmaddr &= PAGE_MASK; + preempt_disable(); - purge_dcache_page_asm(physaddr, vmaddr); + + /* Set context for flush */ + local_irq_save(flags); + prot = mfctl(8); + space = mfsp(SR_USER); + pgd = mfctl(25); +#ifdef CONFIG_TLB_PTLOCK + pgd_lock = mfctl(28); +#endif + switch_mm_irqs_off(NULL, vma->vm_mm, NULL); + local_irq_restore(flags); + + flush_user_dcache_range_asm(vmaddr, vmaddr + PAGE_SIZE); if (vma->vm_flags & VM_EXEC) - flush_icache_page_asm(physaddr, vmaddr); + flush_user_icache_range_asm(vmaddr, vmaddr + PAGE_SIZE); + flush_tlb_page(vma, vmaddr); + + /* Restore previous context */ + local_irq_save(flags); +#ifdef CONFIG_TLB_PTLOCK + mtctl(pgd_lock, 28); +#endif + mtctl(pgd, 25); + mtsp(space, SR_USER); + mtctl(prot, 8); + local_irq_restore(flags); + preempt_enable(); } +static inline pte_t *get_ptep(struct mm_struct *mm, unsigned long addr) +{ + pte_t *ptep = NULL; + pgd_t *pgd = mm->pgd; + p4d_t *p4d; + pud_t *pud; + pmd_t *pmd; + + if (!pgd_none(*pgd)) { + p4d = p4d_offset(pgd, addr); + if (!p4d_none(*p4d)) { + pud = pud_offset(p4d, addr); + if (!pud_none(*pud)) { + pmd = pmd_offset(pud, addr); + if (!pmd_none(*pmd)) + ptep = pte_offset_map(pmd, addr); + } + } + } + return ptep; +} + +static inline bool pte_needs_flush(pte_t pte) +{ + return (pte_val(pte) & (_PAGE_PRESENT | _PAGE_ACCESSED | _PAGE_NO_CACHE)) + == (_PAGE_PRESENT | _PAGE_ACCESSED); +} + void flush_dcache_page(struct page *page) { struct address_space *mapping = page_mapping_file(page); struct vm_area_struct *mpnt; unsigned long offset; unsigned long addr, old_addr = 0; + unsigned long count = 0; pgoff_t pgoff; if (mapping && !mapping_mapped(mapping)) { @@ -341,33 +413,52 @@ void flush_dcache_page(struct page *page) pgoff = page->index; - /* We have carefully arranged in arch_get_unmapped_area() that + /* + * We have carefully arranged in arch_get_unmapped_area() that * *any* mappings of a file are always congruently mapped (whether * declared as MAP_PRIVATE or MAP_SHARED), so we only need - * to flush one address here for them all to become coherent */ - + * to flush one address here for them all to become coherent + * on machines that support equivalent aliasing + */ flush_dcache_mmap_lock(mapping); vma_interval_tree_foreach(mpnt, &mapping->i_mmap, pgoff, pgoff) { offset = (pgoff - mpnt->vm_pgoff) << PAGE_SHIFT; addr = mpnt->vm_start + offset; + if (parisc_requires_coherency()) { + pte_t *ptep; - /* The TLB is the engine of coherence on parisc: The - * CPU is entitled to speculate any page with a TLB - * mapping, so here we kill the mapping then flush the - * page along a special flush only alias mapping. - * This guarantees that the page is no-longer in the - * cache for any process and nor may it be - * speculatively read in (until the user or kernel - * specifically accesses it, of course) */ - - flush_tlb_page(mpnt, addr); - if (old_addr == 0 || (old_addr & (SHM_COLOUR - 1)) - != (addr & (SHM_COLOUR - 1))) { - __flush_cache_page(mpnt, addr, page_to_phys(page)); - if (parisc_requires_coherency() && old_addr) - printk(KERN_ERR "INEQUIVALENT ALIASES 0x%lx and 0x%lx in file %pD\n", old_addr, addr, mpnt->vm_file); - old_addr = addr; + ptep = get_ptep(mpnt->vm_mm, addr); + if (ptep && pte_needs_flush(*ptep)) + flush_user_cache_page(mpnt, addr); + } else { + /* + * The TLB is the engine of coherence on parisc: + * The CPU is entitled to speculate any page + * with a TLB mapping, so here we kill the + * mapping then flush the page along a special + * flush only alias mapping. This guarantees that + * the page is no-longer in the cache for any + * process and nor may it be speculatively read + * in (until the user or kernel specifically + * accesses it, of course) + */ + flush_tlb_page(mpnt, addr); + if (old_addr == 0 || (old_addr & (SHM_COLOUR - 1)) + != (addr & (SHM_COLOUR - 1))) { + __flush_cache_page(mpnt, addr, page_to_phys(page)); + /* + * Software is allowed to have any number + * of private mappings to a page. + */ + if (!(mpnt->vm_flags & VM_SHARED)) + continue; + if (old_addr) + pr_err("INEQUIVALENT ALIASES 0x%lx and 0x%lx in file %pD\n", + old_addr, addr, mpnt->vm_file); + old_addr = addr; + } } + WARN_ON(++count == 4096); } flush_dcache_mmap_unlock(mapping); } @@ -375,7 +466,6 @@ EXPORT_SYMBOL(flush_dcache_page); /* Defined in arch/parisc/kernel/pacache.S */ EXPORT_SYMBOL(flush_kernel_dcache_range_asm); -EXPORT_SYMBOL(flush_data_cache_local); EXPORT_SYMBOL(flush_kernel_icache_range_asm); #define FLUSH_THRESHOLD 0x80000 /* 0.5MB */ @@ -388,7 +478,7 @@ void __init parisc_setup_cache_timing(void) { unsigned long rangetime, alltime; unsigned long size; - unsigned long threshold; + unsigned long threshold, threshold2; alltime = mfctl(16); flush_data_cache(); @@ -402,11 +492,16 @@ void __init parisc_setup_cache_timing(void) printk(KERN_DEBUG "Whole cache flush %lu cycles, flushing %lu bytes %lu cycles\n", alltime, size, rangetime); - threshold = L1_CACHE_ALIGN(size * alltime / rangetime); - if (threshold > cache_info.dc_size) - threshold = cache_info.dc_size; - if (threshold) - parisc_cache_flush_threshold = threshold; + threshold = L1_CACHE_ALIGN((unsigned long)((uint64_t)size * alltime / rangetime)); + pr_info("Calculated flush threshold is %lu KiB\n", + threshold/1024); + + /* + * The threshold computed above isn't very reliable. The following + * heuristic works reasonably well on c8000/rp3440. + */ + threshold2 = cache_info.dc_size * num_online_cpus(); + parisc_cache_flush_threshold = threshold2; printk(KERN_INFO "Cache flush threshold set to %lu KiB\n", parisc_cache_flush_threshold/1024); @@ -451,30 +546,58 @@ extern void purge_kernel_dcache_page_asm(unsigned long); extern void clear_user_page_asm(void *, unsigned long); extern void copy_user_page_asm(void *, void *, unsigned long); -void flush_kernel_dcache_page_addr(void *addr) +void flush_kernel_dcache_page_addr(const void *addr) { unsigned long flags; flush_kernel_dcache_page_asm(addr); purge_tlb_start(flags); - pdtlb_kernel(addr); + pdtlb(SR_KERNEL, addr); purge_tlb_end(flags); } EXPORT_SYMBOL(flush_kernel_dcache_page_addr); -void copy_user_page(void *vto, void *vfrom, unsigned long vaddr, - struct page *pg) +static void flush_cache_page_if_present(struct vm_area_struct *vma, + unsigned long vmaddr, unsigned long pfn) { - /* Copy using kernel mapping. No coherency is needed (all in - kunmap) for the `to' page. However, the `from' page needs to - be flushed through a mapping equivalent to the user mapping - before it can be accessed through the kernel mapping. */ - preempt_disable(); - flush_dcache_page_asm(__pa(vfrom), vaddr); - copy_page_asm(vto, vfrom); - preempt_enable(); + pte_t *ptep = get_ptep(vma->vm_mm, vmaddr); + + /* + * The pte check is racy and sometimes the flush will trigger + * a non-access TLB miss. Hopefully, the page has already been + * flushed. + */ + if (ptep && pte_needs_flush(*ptep)) + flush_cache_page(vma, vmaddr, pfn); +} + +void copy_user_highpage(struct page *to, struct page *from, + unsigned long vaddr, struct vm_area_struct *vma) +{ + void *kto, *kfrom; + + kfrom = kmap_local_page(from); + kto = kmap_local_page(to); + flush_cache_page_if_present(vma, vaddr, page_to_pfn(from)); + copy_page_asm(kto, kfrom); + kunmap_local(kto); + kunmap_local(kfrom); +} + +void copy_to_user_page(struct vm_area_struct *vma, struct page *page, + unsigned long user_vaddr, void *dst, void *src, int len) +{ + flush_cache_page_if_present(vma, user_vaddr, page_to_pfn(page)); + memcpy(dst, src, len); + flush_kernel_dcache_range_asm((unsigned long)dst, (unsigned long)dst + len); +} + +void copy_from_user_page(struct vm_area_struct *vma, struct page *page, + unsigned long user_vaddr, void *dst, void *src, int len) +{ + flush_cache_page_if_present(vma, user_vaddr, page_to_pfn(page)); + memcpy(dst, src, len); } -EXPORT_SYMBOL(copy_user_page); /* __flush_tlb_range() * @@ -496,139 +619,124 @@ int __flush_tlb_range(unsigned long sid, unsigned long start, but cause a purge request to be broadcast to other TLBs. */ while (start < end) { purge_tlb_start(flags); - mtsp(sid, 1); - pdtlb(start); - pitlb(start); + mtsp(sid, SR_TEMP1); + pdtlb(SR_TEMP1, start); + pitlb(SR_TEMP1, start); purge_tlb_end(flags); start += PAGE_SIZE; } return 0; } -static void cacheflush_h_tmp_function(void *dummy) +static void flush_cache_pages(struct vm_area_struct *vma, unsigned long start, unsigned long end) { - flush_cache_all_local(); -} + unsigned long addr, pfn; + pte_t *ptep; -void flush_cache_all(void) -{ - on_each_cpu(cacheflush_h_tmp_function, NULL, 1); + for (addr = start; addr < end; addr += PAGE_SIZE) { + /* + * The vma can contain pages that aren't present. Although + * the pte search is expensive, we need the pte to find the + * page pfn and to check whether the page should be flushed. + */ + ptep = get_ptep(vma->vm_mm, addr); + if (ptep && pte_needs_flush(*ptep)) { + if (parisc_requires_coherency()) { + flush_user_cache_page(vma, addr); + } else { + pfn = pte_pfn(*ptep); + if (WARN_ON(!pfn_valid(pfn))) + return; + __flush_cache_page(vma, addr, PFN_PHYS(pfn)); + } + } + } } static inline unsigned long mm_total_size(struct mm_struct *mm) { struct vm_area_struct *vma; unsigned long usize = 0; + VMA_ITERATOR(vmi, mm, 0); - for (vma = mm->mmap; vma; vma = vma->vm_next) + for_each_vma(vmi, vma) { + if (usize >= parisc_cache_flush_threshold) + break; usize += vma->vm_end - vma->vm_start; - return usize; -} - -static inline pte_t *get_ptep(pgd_t *pgd, unsigned long addr) -{ - pte_t *ptep = NULL; - - if (!pgd_none(*pgd)) { - p4d_t *p4d = p4d_offset(pgd, addr); - if (!p4d_none(*p4d)) { - pud_t *pud = pud_offset(p4d, addr); - if (!pud_none(*pud)) { - pmd_t *pmd = pmd_offset(pud, addr); - if (!pmd_none(*pmd)) - ptep = pte_offset_map(pmd, addr); - } - } } - return ptep; + return usize; } -static void flush_cache_pages(struct vm_area_struct *vma, struct mm_struct *mm, - unsigned long start, unsigned long end) +void flush_cache_mm(struct mm_struct *mm) { - unsigned long addr, pfn; - pte_t *ptep; - - for (addr = start; addr < end; addr += PAGE_SIZE) { - ptep = get_ptep(mm->pgd, addr); - if (ptep) { - pfn = pte_pfn(*ptep); - flush_cache_page(vma, addr, pfn); - } + struct vm_area_struct *vma; + VMA_ITERATOR(vmi, mm, 0); + + /* + * Flushing the whole cache on each cpu takes forever on + * rp3440, etc. So, avoid it if the mm isn't too big. + * + * Note that we must flush the entire cache on machines + * with aliasing caches to prevent random segmentation + * faults. + */ + if (!parisc_requires_coherency() + || mm_total_size(mm) >= parisc_cache_flush_threshold) { + if (WARN_ON(IS_ENABLED(CONFIG_SMP) && arch_irqs_disabled())) + return; + flush_tlb_all(); + flush_cache_all(); + return; } -} -static void flush_user_cache_tlb(struct vm_area_struct *vma, - unsigned long start, unsigned long end) -{ - flush_user_dcache_range_asm(start, end); - if (vma->vm_flags & VM_EXEC) - flush_user_icache_range_asm(start, end); - flush_tlb_range(vma, start, end); + /* Flush mm */ + for_each_vma(vmi, vma) + flush_cache_pages(vma, vma->vm_start, vma->vm_end); } -void flush_cache_mm(struct mm_struct *mm) +void flush_cache_range(struct vm_area_struct *vma, unsigned long start, unsigned long end) { - struct vm_area_struct *vma; - - /* Flushing the whole cache on each cpu takes forever on - rp3440, etc. So, avoid it if the mm isn't too big. */ - if ((!IS_ENABLED(CONFIG_SMP) || !arch_irqs_disabled()) && - mm_total_size(mm) >= parisc_cache_flush_threshold) { - if (mm->context) - flush_tlb_all(); + if (!parisc_requires_coherency() + || end - start >= parisc_cache_flush_threshold) { + if (WARN_ON(IS_ENABLED(CONFIG_SMP) && arch_irqs_disabled())) + return; + flush_tlb_range(vma, start, end); flush_cache_all(); return; } - preempt_disable(); - if (mm->context == mfsp(3)) { - for (vma = mm->mmap; vma; vma = vma->vm_next) - flush_user_cache_tlb(vma, vma->vm_start, vma->vm_end); - preempt_enable(); - return; - } + flush_cache_pages(vma, start, end); +} - for (vma = mm->mmap; vma; vma = vma->vm_next) - flush_cache_pages(vma, mm, vma->vm_start, vma->vm_end); - preempt_enable(); +void flush_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, unsigned long pfn) +{ + if (WARN_ON(!pfn_valid(pfn))) + return; + if (parisc_requires_coherency()) + flush_user_cache_page(vma, vmaddr); + else + __flush_cache_page(vma, vmaddr, PFN_PHYS(pfn)); } -void flush_cache_range(struct vm_area_struct *vma, - unsigned long start, unsigned long end) +void flush_anon_page(struct vm_area_struct *vma, struct page *page, unsigned long vmaddr) { - if ((!IS_ENABLED(CONFIG_SMP) || !arch_irqs_disabled()) && - end - start >= parisc_cache_flush_threshold) { - if (vma->vm_mm->context) - flush_tlb_range(vma, start, end); - flush_cache_all(); + if (!PageAnon(page)) return; - } - preempt_disable(); - if (vma->vm_mm->context == mfsp(3)) { - flush_user_cache_tlb(vma, start, end); - preempt_enable(); + if (parisc_requires_coherency()) { + if (vma->vm_flags & VM_SHARED) + flush_data_cache(); + else + flush_user_cache_page(vma, vmaddr); return; } - flush_cache_pages(vma, vma->vm_mm, vma->vm_start, vma->vm_end); + flush_tlb_page(vma, vmaddr); + preempt_disable(); + flush_dcache_page_asm(page_to_phys(page), vmaddr); preempt_enable(); } -void -flush_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, unsigned long pfn) -{ - if (pfn_valid(pfn)) { - if (likely(vma->vm_mm->context)) { - flush_tlb_page(vma, vmaddr); - __flush_cache_page(vma, vmaddr, PFN_PHYS(pfn)); - } else { - __purge_cache_page(vma, vmaddr, PFN_PHYS(pfn)); - } - } -} - void flush_kernel_vmap_range(void *vaddr, int size) { unsigned long start = (unsigned long)vaddr; @@ -651,6 +759,9 @@ void invalidate_kernel_vmap_range(void *vaddr, int size) unsigned long start = (unsigned long)vaddr; unsigned long end = start + size; + /* Ensure DMA is complete */ + asm_syncdma(); + if ((!IS_ENABLED(CONFIG_SMP) || !arch_irqs_disabled()) && (unsigned long)size >= parisc_cache_flush_threshold) { flush_tlb_kernel_range(start, end); diff --git a/arch/parisc/kernel/drivers.c b/arch/parisc/kernel/drivers.c index 776d624a7207..e7ee0c0c91d3 100644 --- a/arch/parisc/kernel/drivers.c +++ b/arch/parisc/kernel/drivers.c @@ -520,7 +520,6 @@ alloc_pa_dev(unsigned long hpa, struct hardware_path *mod_path) dev->id.hversion_rev = iodc_data[1] & 0x0f; dev->id.sversion = ((iodc_data[4] & 0x0f) << 16) | (iodc_data[5] << 8) | iodc_data[6]; - dev->hpa.name = parisc_pathname(dev); dev->hpa.start = hpa; /* This is awkward. The STI spec says that gfx devices may occupy * 32MB or 64MB. Unfortunately, we don't know how to tell whether @@ -534,10 +533,10 @@ alloc_pa_dev(unsigned long hpa, struct hardware_path *mod_path) dev->hpa.end = hpa + 0xfff; } dev->hpa.flags = IORESOURCE_MEM; - name = parisc_hardware_description(&dev->id); - if (name) { - strlcpy(dev->name, name, sizeof(dev->name)); - } + dev->hpa.name = dev->name; + name = parisc_hardware_description(&dev->id) ? : "unknown"; + snprintf(dev->name, sizeof(dev->name), "%s [%s]", + name, parisc_pathname(dev)); /* Silently fail things like mouse ports which are subsumed within * the keyboard controller @@ -883,15 +882,13 @@ void __init walk_central_bus(void) &root); } -static void print_parisc_device(struct parisc_device *dev) +static __init void print_parisc_device(struct parisc_device *dev) { - char hw_path[64]; - static int count; + static int count __initdata; - print_pa_hwpath(dev, hw_path); - pr_info("%d. %s at %pap [%s] { %d, 0x%x, 0x%.3x, 0x%.5x }", - ++count, dev->name, &(dev->hpa.start), hw_path, dev->id.hw_type, - dev->id.hversion_rev, dev->id.hversion, dev->id.sversion); + pr_info("%d. %s at %pap { type:%d, hv:%#x, sv:%#x, rev:%#x }", + ++count, dev->name, &(dev->hpa.start), dev->id.hw_type, + dev->id.hversion, dev->id.sversion, dev->id.hversion_rev); if (dev->num_addrs) { int k; @@ -1080,7 +1077,7 @@ static __init int qemu_print_iodc_data(struct device *lin_dev, void *data) -static int print_one_device(struct device * dev, void * data) +static __init int print_one_device(struct device * dev, void * data) { struct parisc_device * pdev = to_parisc_device(dev); diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S index 6e9cdb269862..0e5ebfe8d9d2 100644 --- a/arch/parisc/kernel/entry.S +++ b/arch/parisc/kernel/entry.S @@ -499,6 +499,10 @@ * Finally, _PAGE_READ goes in the top bit of PL1 (so we * trigger an access rights trap in user space if the user * tries to read an unreadable page */ +#if _PAGE_SPECIAL_BIT == _PAGE_DMB_BIT + /* need to drop DMB bit, as it's used as SPECIAL flag */ + depi 0,_PAGE_SPECIAL_BIT,1,\pte +#endif depd \pte,8,7,\prot /* PAGE_USER indicates the page can be read with user privileges, @@ -529,6 +533,10 @@ * makes the tlb entry for the differently formatted pa11 * insertion instructions */ .macro make_insert_tlb_11 spc,pte,prot +#if _PAGE_SPECIAL_BIT == _PAGE_DMB_BIT + /* need to drop DMB bit, as it's used as SPECIAL flag */ + depi 0,_PAGE_SPECIAL_BIT,1,\pte +#endif zdep \spc,30,15,\prot dep \pte,8,7,\prot extru,= \pte,_PAGE_NO_CACHE_BIT,1,%r0 @@ -554,8 +562,9 @@ extrd,s \pte,63,25,\pte .endm - /* The alias region is an 8MB aligned 16MB to do clear and - * copy user pages at addresses congruent with the user + /* The alias region is comprised of a pair of 4 MB regions + * aligned to 8 MB. It is used to clear/copy/flush user pages + * using kernel virtual addresses congruent with the user * virtual address. * * To use the alias page, you set %r26 up with the to TLB @@ -565,13 +574,8 @@ .macro do_alias spc,tmp,tmp1,va,pte,prot,fault,patype cmpib,COND(<>),n 0,\spc,\fault ldil L%(TMPALIAS_MAP_START),\tmp -#if defined(CONFIG_64BIT) && (TMPALIAS_MAP_START >= 0x80000000) - /* on LP64, ldi will sign extend into the upper 32 bits, - * which is behaviour we don't want */ - depdi 0,31,32,\tmp -#endif copy \va,\tmp1 - depi 0,31,23,\tmp1 + depi_safe 0,31,TMPALIAS_SIZE_BITS+1,\tmp1 cmpb,COND(<>),n \tmp,\tmp1,\fault mfctl %cr19,\tmp /* iir */ /* get the opcode (first six bits) into \tmp */ @@ -604,13 +608,13 @@ * OK, it is in the temp alias region, check whether "from" or "to". * Check "subtle" note in pacache.S re: r23/r26. */ -#ifdef CONFIG_64BIT - extrd,u,*= \va,41,1,%r0 -#else - extrw,u,= \va,9,1,%r0 -#endif + extrw,u,= \va,31-TMPALIAS_SIZE_BITS,1,%r0 or,COND(tr) %r23,%r0,\pte or %r26,%r0,\pte + + /* convert phys addr in \pte (from r23 or r26) to tlb insert format */ + SHRREG \pte,PAGE_SHIFT+PAGE_ADD_SHIFT-5, \pte + depi_safe _PAGE_SIZE_ENCODING_DEFAULT, 31,5, \pte .endm @@ -1288,74 +1292,12 @@ nadtlb_check_alias_20: nadtlb_emulate: /* - * Non access misses can be caused by fdc,fic,pdc,lpa,probe and - * probei instructions. We don't want to fault for these - * instructions (not only does it not make sense, it can cause - * deadlocks, since some flushes are done with the mmap - * semaphore held). If the translation doesn't exist, we can't - * insert a translation, so have to emulate the side effects - * of the instruction. Since we don't insert a translation - * we can get a lot of faults during a flush loop, so it makes - * sense to try to do it here with minimum overhead. We only - * emulate fdc,fic,pdc,probew,prober instructions whose base - * and index registers are not shadowed. We defer everything - * else to the "slow" path. + * Non-access misses can be caused by fdc,fic,pdc,lpa,probe and + * probei instructions. The kernel no longer faults doing flushes. + * Use of lpa and probe instructions is rare. Given the issue + * with shadow registers, we defer everything to the "slow" path. */ - - mfctl %cr19,%r9 /* Get iir */ - - /* PA 2.0 Arch Ref. Book pg 382 has a good description of the insn bits. - Checks for fdc,fdce,pdc,"fic,4f",prober,probeir,probew, probeiw */ - - /* Checks for fdc,fdce,pdc,"fic,4f" only */ - ldi 0x280,%r16 - and %r9,%r16,%r17 - cmpb,<>,n %r16,%r17,nadtlb_probe_check - bb,>=,n %r9,26,nadtlb_nullify /* m bit not set, just nullify */ - BL get_register,%r25 - extrw,u %r9,15,5,%r8 /* Get index register # */ - cmpib,COND(=),n -1,%r1,nadtlb_fault /* have to use slow path */ - copy %r1,%r24 - BL get_register,%r25 - extrw,u %r9,10,5,%r8 /* Get base register # */ - cmpib,COND(=),n -1,%r1,nadtlb_fault /* have to use slow path */ - BL set_register,%r25 - add,l %r1,%r24,%r1 /* doesn't affect c/b bits */ - -nadtlb_nullify: - mfctl %ipsw,%r8 - ldil L%PSW_N,%r9 - or %r8,%r9,%r8 /* Set PSW_N */ - mtctl %r8,%ipsw - - rfir - nop - - /* - When there is no translation for the probe address then we - must nullify the insn and return zero in the target register. - This will indicate to the calling code that it does not have - write/read privileges to this address. - - This should technically work for prober and probew in PA 1.1, - and also probe,r and probe,w in PA 2.0 - - WARNING: USE ONLY NON-SHADOW REGISTERS WITH PROBE INSN! - THE SLOW-PATH EMULATION HAS NOT BEEN WRITTEN YET. - - */ -nadtlb_probe_check: - ldi 0x80,%r16 - and %r9,%r16,%r17 - cmpb,<>,n %r16,%r17,nadtlb_fault /* Must be probe,[rw]*/ - BL get_register,%r25 /* Find the target register */ - extrw,u %r9,31,5,%r8 /* Get target register */ - cmpib,COND(=),n -1,%r1,nadtlb_fault /* have to use slow path */ - BL set_register,%r25 - copy %r0,%r1 /* Write zero to target register */ - b nadtlb_nullify /* Nullify return insn */ - nop - + b,n nadtlb_fault #ifdef CONFIG_64BIT itlb_miss_20w: diff --git a/arch/parisc/kernel/firmware.c b/arch/parisc/kernel/firmware.c index 3370e347dde3..6a7e315bcc2e 100644 --- a/arch/parisc/kernel/firmware.c +++ b/arch/parisc/kernel/firmware.c @@ -83,7 +83,7 @@ extern unsigned long pdc_result2[NUM_PDC_RESULT]; /* Firmware needs to be initially set to narrow to determine the * actual firmware width. */ -int parisc_narrow_firmware __ro_after_init = 1; +int parisc_narrow_firmware __ro_after_init = 2; #endif /* On most currently-supported platforms, IODC I/O calls are 32-bit calls @@ -174,6 +174,11 @@ void set_firmware_width_unlocked(void) void set_firmware_width(void) { unsigned long flags; + + /* already initialized? */ + if (parisc_narrow_firmware != 2) + return; + spin_lock_irqsave(&pdc_lock, flags); set_firmware_width_unlocked(); spin_unlock_irqrestore(&pdc_lock, flags); @@ -324,7 +329,44 @@ int __pdc_cpu_rendezvous(void) return mem_pdc_call(PDC_PROC, 1, 0); } +/** + * pdc_cpu_rendezvous_lock - Lock PDC while transitioning to rendezvous state + */ +void pdc_cpu_rendezvous_lock(void) +{ + spin_lock(&pdc_lock); +} + +/** + * pdc_cpu_rendezvous_unlock - Unlock PDC after reaching rendezvous state + */ +void pdc_cpu_rendezvous_unlock(void) +{ + spin_unlock(&pdc_lock); +} + +/** + * pdc_pat_get_PDC_entrypoint - Get PDC entry point for current CPU + * @retval: -1 on error, 0 on success + */ +int pdc_pat_get_PDC_entrypoint(unsigned long *pdc_entry) +{ + int retval = 0; + unsigned long flags; + + if (!IS_ENABLED(CONFIG_SMP) || !is_pdc_pat()) { + *pdc_entry = MEM_PDC; + return 0; + } + + spin_lock_irqsave(&pdc_lock, flags); + retval = mem_pdc_call(PDC_PAT_CPU, PDC_PAT_CPU_GET_PDC_ENTRYPOINT, + __pa(pdc_result)); + *pdc_entry = pdc_result[0]; + spin_unlock_irqrestore(&pdc_lock, flags); + return retval; +} /** * pdc_chassis_warn - Fetches chassis warnings * @retval: -1 on error, 0 on success diff --git a/arch/parisc/kernel/hardware.c b/arch/parisc/kernel/hardware.c index 17161e72ea29..357d9cdab7ce 100644 --- a/arch/parisc/kernel/hardware.c +++ b/arch/parisc/kernel/hardware.c @@ -23,9 +23,6 @@ * HP PARISC Hardware Database * Access to this database is only possible during bootup * so don't reference this table after starting the init process - * - * NOTE: Product names which are listed here and ends with a '?' - * are guessed. If you know the correct name, please let us know. */ static struct hp_hardware hp_hardware_list[] __initdata = { @@ -212,7 +209,7 @@ static struct hp_hardware hp_hardware_list[] __initdata = { {HPHW_NPROC,0x5DD,0x4,0x81,"Duet W2"}, {HPHW_NPROC,0x5DE,0x4,0x81,"Piccolo W+"}, {HPHW_NPROC,0x5DF,0x4,0x81,"Cantata W2"}, - {HPHW_NPROC,0x5DF,0x0,0x00,"Marcato W+ (rp5470)?"}, + {HPHW_NPROC,0x5DF,0x0,0x00,"Marcato W+ (rp5470)"}, {HPHW_NPROC,0x5E0,0x4,0x91,"Cantata DC- W2"}, {HPHW_NPROC,0x5E1,0x4,0x91,"Crescendo DC- W2"}, {HPHW_NPROC,0x5E2,0x4,0x91,"Crescendo 650 W2"}, @@ -266,11 +263,11 @@ static struct hp_hardware hp_hardware_list[] __initdata = { {HPHW_NPROC,0x888,0x4,0x91,"Storm Peak Fast DC-"}, {HPHW_NPROC,0x889,0x4,0x91,"Storm Peak Fast"}, {HPHW_NPROC,0x88A,0x4,0x91,"Crestone Peak Slow"}, - {HPHW_NPROC,0x88B,0x4,0x91,"Crestone Peak Fast?"}, + {HPHW_NPROC,0x88B,0x4,0x91,"Crestone Peak Fast"}, {HPHW_NPROC,0x88C,0x4,0x91,"Orca Mako+"}, {HPHW_NPROC,0x88D,0x4,0x91,"Rainier/Medel Mako+ Slow"}, {HPHW_NPROC,0x88E,0x4,0x91,"Rainier/Medel Mako+ Fast"}, - {HPHW_NPROC,0x892,0x4,0x91,"Mt. Hamilton Slow Mako+?"}, + {HPHW_NPROC,0x892,0x4,0x91,"Mt. Hamilton Slow Mako+"}, {HPHW_NPROC,0x894,0x4,0x91,"Mt. Hamilton Fast Mako+"}, {HPHW_NPROC,0x895,0x4,0x91,"Storm Peak Slow Mako+"}, {HPHW_NPROC,0x896,0x4,0x91,"Storm Peak Fast Mako+"}, @@ -1198,7 +1195,7 @@ static struct hp_hardware hp_hardware_list[] __initdata = { {HPHW_FIO, 0x004, 0x00340, 0x0, "BARCO CX4500 VME Grphx Cnsl"}, {HPHW_FIO, 0x004, 0x00360, 0x0, "Hughes TOG VME FDDI"}, {HPHW_FIO, 0x076, 0x000AD, 0x0, "Crestone Peak Core RS-232"}, - {HPHW_FIO, 0x077, 0x000AD, 0x0, "Crestone Peak Fast? Core RS-232"}, + {HPHW_FIO, 0x077, 0x000AD, 0x0, "Crestone Peak Fast Core RS-232"}, {HPHW_IOA, 0x185, 0x0000B, 0x00, "Java BC Summit Port"}, {HPHW_IOA, 0x1FF, 0x0000B, 0x00, "Hitachi Ghostview Summit Port"}, {HPHW_IOA, 0x580, 0x0000B, 0x10, "U2-IOA BC Runway Port"}, diff --git a/arch/parisc/kernel/head.S b/arch/parisc/kernel/head.S index b24f77748c22..fd15fd4bbb61 100644 --- a/arch/parisc/kernel/head.S +++ b/arch/parisc/kernel/head.S @@ -22,7 +22,7 @@ #include <linux/init.h> #include <linux/pgtable.h> - .level PA_ASM_LEVEL + .level 1.1 __INITDATA ENTRY(boot_args) @@ -70,6 +70,47 @@ $bss_loop: stw,ma %arg2,4(%r1) stw,ma %arg3,4(%r1) +#if !defined(CONFIG_64BIT) && defined(CONFIG_PA20) + /* This 32-bit kernel was compiled for PA2.0 CPUs. Check current CPU + * and halt kernel if we detect a PA1.x CPU. */ + ldi 32,%r10 + mtctl %r10,%cr11 + .level 2.0 + mfctl,w %cr11,%r10 + .level 1.1 + comib,<>,n 0,%r10,$cpu_ok + + load32 PA(msg1),%arg0 + ldi msg1_end-msg1,%arg1 +$iodc_panic: + copy %arg0, %r10 + copy %arg1, %r11 + load32 PA(init_stack),%sp +#define MEM_CONS 0x3A0 + ldw MEM_CONS+32(%r0),%arg0 // HPA + ldi ENTRY_IO_COUT,%arg1 + ldw MEM_CONS+36(%r0),%arg2 // SPA + ldw MEM_CONS+8(%r0),%arg3 // layers + load32 PA(__bss_start),%r1 + stw %r1,-52(%sp) // arg4 + stw %r0,-56(%sp) // arg5 + stw %r10,-60(%sp) // arg6 = ptr to text + stw %r11,-64(%sp) // arg7 = len + stw %r0,-68(%sp) // arg8 + load32 PA(.iodc_panic_ret), %rp + ldw MEM_CONS+40(%r0),%r1 // ENTRY_IODC + bv,n (%r1) +.iodc_panic_ret: + b . /* wait endless with ... */ + or %r10,%r10,%r10 /* qemu idle sleep */ +msg1: .ascii "Can't boot kernel which was built for PA8x00 CPUs on this machine.\r\n" +msg1_end: + +$cpu_ok: +#endif + + .level PA_ASM_LEVEL + /* Initialize startup VM. Just map first 16/32 MB of memory */ load32 PA(swapper_pg_dir),%r4 mtctl %r4,%cr24 /* Initialize kernel root pointer */ @@ -162,6 +203,15 @@ $pgt_fill_loop: /* FALLTHROUGH */ .procend +#ifdef CONFIG_HOTPLUG_CPU + /* common_stext is far away in another section... jump there */ + load32 PA(common_stext), %rp + bv,n (%rp) + + /* common_stext and smp_slave_stext needs to be in text section */ + .text +#endif + /* ** Code Common to both Monarch and Slave processors. ** Entry: @@ -371,8 +421,6 @@ smp_slave_stext: .procend #endif /* CONFIG_SMP */ -ENDPROC(parisc_kernel_start) - #ifndef CONFIG_64BIT .section .data..ro_after_init diff --git a/arch/parisc/kernel/hpmc.S b/arch/parisc/kernel/hpmc.S index c2981401775c..eb2e4bd67035 100644 --- a/arch/parisc/kernel/hpmc.S +++ b/arch/parisc/kernel/hpmc.S @@ -43,10 +43,8 @@ * IODC requires 7K byte stack. That leaves 1K byte for os_hpmc. */ - __PAGE_ALIGNED_BSS - .align 4096 -hpmc_stack: - .block 16384 + .import toc_stack,data +#define hpmc_stack toc_stack /* re-use the TOC stack */ #define HPMC_IODC_BUF_SIZE 0x8000 diff --git a/arch/parisc/kernel/irq.c b/arch/parisc/kernel/irq.c index eb18e16362f6..b05055f3ba4b 100644 --- a/arch/parisc/kernel/irq.c +++ b/arch/parisc/kernel/irq.c @@ -105,28 +105,12 @@ int cpu_check_affinity(struct irq_data *d, const struct cpumask *dest) if (irqd_is_per_cpu(d)) return -EINVAL; - /* whatever mask they set, we just allow one CPU */ - cpu_dest = cpumask_next_and(d->irq & (num_online_cpus()-1), - dest, cpu_online_mask); + cpu_dest = cpumask_first_and(dest, cpu_online_mask); if (cpu_dest >= nr_cpu_ids) - cpu_dest = cpumask_first_and(dest, cpu_online_mask); + cpu_dest = cpumask_first(cpu_online_mask); return cpu_dest; } - -static int cpu_set_affinity_irq(struct irq_data *d, const struct cpumask *dest, - bool force) -{ - int cpu_dest; - - cpu_dest = cpu_check_affinity(d, dest); - if (cpu_dest < 0) - return -1; - - cpumask_copy(irq_data_get_affinity_mask(d), dest); - - return 0; -} #endif static struct irq_chip cpu_interrupt_type = { @@ -135,9 +119,6 @@ static struct irq_chip cpu_interrupt_type = { .irq_unmask = cpu_unmask_irq, .irq_ack = cpu_ack_irq, .irq_eoi = cpu_eoi_irq, -#ifdef CONFIG_SMP - .irq_set_affinity = cpu_set_affinity_irq, -#endif /* XXX: Needs to be written. We managed without it so far, but * we really ought to write it. */ @@ -334,7 +315,7 @@ unsigned long txn_affinity_addr(unsigned int irq, int cpu) { #ifdef CONFIG_SMP struct irq_data *d = irq_get_irq_data(irq); - cpumask_copy(irq_data_get_affinity_mask(d), cpumask_of(cpu)); + irq_data_update_affinity(d, cpumask_of(cpu)); #endif return per_cpu(cpu_data, cpu).txn_addr; @@ -499,10 +480,12 @@ static void execute_on_irq_stack(void *func, unsigned long param1) *irq_stack_in_use = 1; } +#ifdef CONFIG_SOFTIRQ_ON_OWN_STACK void do_softirq_own_stack(void) { execute_on_irq_stack(__do_softirq, 0); } +#endif #endif /* CONFIG_IRQSTACKS */ /* ONLY called from entry.S:intr_extint() */ @@ -582,7 +565,7 @@ static void claim_cpu_irqs(void) #endif } -void __init init_IRQ(void) +void init_IRQ(void) { local_irq_disable(); /* PARANOID - should already be disabled */ mtctl(~0UL, 23); /* EIRR : clear all pending external intr */ diff --git a/arch/parisc/kernel/jump_label.c b/arch/parisc/kernel/jump_label.c index d2f3cb12e282..e253b134500d 100644 --- a/arch/parisc/kernel/jump_label.c +++ b/arch/parisc/kernel/jump_label.c @@ -42,14 +42,3 @@ void arch_jump_label_transform(struct jump_entry *entry, patch_text(addr, insn); } - -void arch_jump_label_transform_static(struct jump_entry *entry, - enum jump_label_type type) -{ - /* - * We use the architected NOP in arch_static_branch, so there's no - * need to patch an identical NOP over the top of it here. The core - * will call arch_jump_label_transform from a module notifier if the - * NOP needs to be replaced by a branch. - */ -} diff --git a/arch/parisc/kernel/kgdb.c b/arch/parisc/kernel/kgdb.c index c4554ac13eac..ab7620f695be 100644 --- a/arch/parisc/kernel/kgdb.c +++ b/arch/parisc/kernel/kgdb.c @@ -3,6 +3,7 @@ * PA-RISC KGDB support * * Copyright (c) 2019 Sven Schnelle <svens@stackframe.org> + * Copyright (c) 2022 Helge Deller <deller@gmx.de> * */ @@ -207,3 +208,23 @@ int kgdb_arch_handle_exception(int trap, int signo, } return -1; } + +/* KGDB console driver which uses PDC to read chars from keyboard */ + +static void kgdb_pdc_write_char(u8 chr) +{ + /* no need to print char. kgdb will do it. */ +} + +static struct kgdb_io kgdb_pdc_io_ops = { + .name = "kgdb_pdc", + .read_char = pdc_iodc_getc, + .write_char = kgdb_pdc_write_char, +}; + +static int __init kgdb_pdc_init(void) +{ + kgdb_register_io_module(&kgdb_pdc_io_ops); + return 0; +} +early_initcall(kgdb_pdc_init); diff --git a/arch/parisc/kernel/kprobes.c b/arch/parisc/kernel/kprobes.c index e2bdb5a5f93e..6e0b86652f30 100644 --- a/arch/parisc/kernel/kprobes.c +++ b/arch/parisc/kernel/kprobes.c @@ -5,6 +5,7 @@ * PA-RISC kprobes implementation * * Copyright (c) 2019 Sven Schnelle <svens@stackframe.org> + * Copyright (c) 2022 Helge Deller <deller@gmx.de> */ #include <linux/types.h> @@ -25,9 +26,14 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p) if (!p->ainsn.insn) return -ENOMEM; - memcpy(p->ainsn.insn, p->addr, - MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + /* + * Set up new instructions. Second break instruction will + * trigger call of parisc_kprobe_ss_handler(). + */ p->opcode = *p->addr; + p->ainsn.insn[0] = p->opcode; + p->ainsn.insn[1] = PARISC_KPROBES_BREAK_INSN2; + flush_insn_slot(p); return 0; } @@ -73,9 +79,7 @@ static void __kprobes setup_singlestep(struct kprobe *p, { kcb->iaoq[0] = regs->iaoq[0]; kcb->iaoq[1] = regs->iaoq[1]; - regs->iaoq[0] = (unsigned long)p->ainsn.insn; - mtctl(0, 0); - regs->gr[0] |= PSW_R; + instruction_pointer_set(regs, (unsigned long)p->ainsn.insn); } int __kprobes parisc_kprobe_break_handler(struct pt_regs *regs) @@ -148,7 +152,7 @@ int __kprobes parisc_kprobe_ss_handler(struct pt_regs *regs) /* for absolute branch instructions we can copy iaoq_b. for relative * branch instructions we need to calculate the new address based on the * difference between iaoq_f and iaoq_b. We cannot use iaoq_b without - * modificationt because it's based on our ainsn.insn address. + * modifications because it's based on our ainsn.insn address. */ if (p->post_handler) @@ -165,9 +169,8 @@ int __kprobes parisc_kprobe_ss_handler(struct pt_regs *regs) regs->iaoq[0] = kcb->iaoq[1]; break; default: - regs->iaoq[1] = kcb->iaoq[0]; - regs->iaoq[1] += (regs->iaoq[1] - regs->iaoq[0]) + 4; regs->iaoq[0] = kcb->iaoq[1]; + regs->iaoq[1] = regs->iaoq[0] + 4; break; } kcb->kprobe_status = KPROBE_HIT_SSDONE; @@ -191,14 +194,17 @@ static struct kprobe trampoline_p = { static int __kprobes trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs) { - unsigned long orig_ret_address; - - orig_ret_address = __kretprobe_trampoline_handler(regs, NULL); - instruction_pointer_set(regs, orig_ret_address); + __kretprobe_trampoline_handler(regs, NULL); return 1; } +void arch_kretprobe_fixup_return(struct pt_regs *regs, + kprobe_opcode_t *correct_ret_addr) +{ + regs->gr[2] = (unsigned long)correct_ret_addr; +} + void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri, struct pt_regs *regs) { diff --git a/arch/parisc/kernel/pacache.S b/arch/parisc/kernel/pacache.S index b2ba6d633065..9a0018f1f42c 100644 --- a/arch/parisc/kernel/pacache.S +++ b/arch/parisc/kernel/pacache.S @@ -300,7 +300,6 @@ fdoneloop2: fdce,m %arg1(%sr1, %arg0) /* Fdce for one loop */ fdsync: - syncdma sync mtsm %r22 /* restore I-bit */ 89: ALTERNATIVE(88b, 89b, ALT_COND_NO_DCACHE, INSN_NOP) @@ -488,6 +487,8 @@ ENDPROC_CFI(copy_page_asm) * parisc chip designers that there will not ever be a parisc * chip with a larger alias boundary (Never say never :-) ). * + * Yah, what about the PA8800 and PA8900 processors? + * * Subtle: the dtlb miss handlers support the temp alias region by * "knowing" that if a dtlb miss happens within the temp alias * region it must have occurred while in clear_user_page. Since @@ -499,19 +500,10 @@ ENDPROC_CFI(copy_page_asm) * miss on the translation, the dtlb miss handler inserts the * translation into the tlb using these values: * - * %r26 physical page (shifted for tlb insert) of "to" translation - * %r23 physical page (shifted for tlb insert) of "from" translation + * %r26 physical address of "to" translation + * %r23 physical address of "from" translation */ - /* Drop prot bits and convert to page addr for iitlbt and idtlbt */ - #define PAGE_ADD_SHIFT (PAGE_SHIFT-12) - .macro convert_phys_for_tlb_insert20 phys - extrd,u \phys, 56-PAGE_ADD_SHIFT, 32-PAGE_ADD_SHIFT, \phys -#if _PAGE_SIZE_ENCODING_DEFAULT - depdi _PAGE_SIZE_ENCODING_DEFAULT, 63, (63-58), \phys -#endif - .endm - /* * copy_user_page_asm() performs a page copy using mappings * equivalent to the user page mappings. It can be used to @@ -540,24 +532,10 @@ ENTRY_CFI(copy_user_page_asm) sub %r25, %r1, %r23 ldil L%(TMPALIAS_MAP_START), %r28 -#ifdef CONFIG_64BIT -#if (TMPALIAS_MAP_START >= 0x80000000) - depdi 0, 31,32, %r28 /* clear any sign extension */ -#endif - convert_phys_for_tlb_insert20 %r26 /* convert phys addr to tlb insert format */ - convert_phys_for_tlb_insert20 %r23 /* convert phys addr to tlb insert format */ - depd %r24,63,22, %r28 /* Form aliased virtual address 'to' */ - depdi 0, 63,PAGE_SHIFT, %r28 /* Clear any offset bits */ - copy %r28, %r29 - depdi 1, 41,1, %r29 /* Form aliased virtual address 'from' */ -#else - extrw,u %r26, 24,25, %r26 /* convert phys addr to tlb insert format */ - extrw,u %r23, 24,25, %r23 /* convert phys addr to tlb insert format */ - depw %r24, 31,22, %r28 /* Form aliased virtual address 'to' */ - depwi 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ + dep_safe %r24, 31,TMPALIAS_SIZE_BITS, %r28 /* Form aliased virtual address 'to' */ + depi_safe 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ copy %r28, %r29 - depwi 1, 9,1, %r29 /* Form aliased virtual address 'from' */ -#endif + depi_safe 1, 31-TMPALIAS_SIZE_BITS,1, %r29 /* Form aliased virtual address 'from' */ /* Purge any old translations */ @@ -687,18 +665,8 @@ ENTRY_CFI(clear_user_page_asm) tophys_r1 %r26 ldil L%(TMPALIAS_MAP_START), %r28 -#ifdef CONFIG_64BIT -#if (TMPALIAS_MAP_START >= 0x80000000) - depdi 0, 31,32, %r28 /* clear any sign extension */ -#endif - convert_phys_for_tlb_insert20 %r26 /* convert phys addr to tlb insert format */ - depd %r25, 63,22, %r28 /* Form aliased virtual address 'to' */ - depdi 0, 63,PAGE_SHIFT, %r28 /* Clear any offset bits */ -#else - extrw,u %r26, 24,25, %r26 /* convert phys addr to tlb insert format */ - depw %r25, 31,22, %r28 /* Form aliased virtual address 'to' */ - depwi 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ -#endif + dep_safe %r25, 31,TMPALIAS_SIZE_BITS, %r28 /* Form aliased virtual address 'to' */ + depi_safe 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ /* Purge any old translation */ @@ -763,18 +731,8 @@ ENDPROC_CFI(clear_user_page_asm) ENTRY_CFI(flush_dcache_page_asm) ldil L%(TMPALIAS_MAP_START), %r28 -#ifdef CONFIG_64BIT -#if (TMPALIAS_MAP_START >= 0x80000000) - depdi 0, 31,32, %r28 /* clear any sign extension */ -#endif - convert_phys_for_tlb_insert20 %r26 /* convert phys addr to tlb insert format */ - depd %r25, 63,22, %r28 /* Form aliased virtual address 'to' */ - depdi 0, 63,PAGE_SHIFT, %r28 /* Clear any offset bits */ -#else - extrw,u %r26, 24,25, %r26 /* convert phys addr to tlb insert format */ - depw %r25, 31,22, %r28 /* Form aliased virtual address 'to' */ - depwi 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ -#endif + dep_safe %r25, 31,TMPALIAS_SIZE_BITS, %r28 /* Form aliased virtual address 'to' */ + depi_safe 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ /* Purge any old translation */ @@ -822,18 +780,8 @@ ENDPROC_CFI(flush_dcache_page_asm) ENTRY_CFI(purge_dcache_page_asm) ldil L%(TMPALIAS_MAP_START), %r28 -#ifdef CONFIG_64BIT -#if (TMPALIAS_MAP_START >= 0x80000000) - depdi 0, 31,32, %r28 /* clear any sign extension */ -#endif - convert_phys_for_tlb_insert20 %r26 /* convert phys addr to tlb insert format */ - depd %r25, 63,22, %r28 /* Form aliased virtual address 'to' */ - depdi 0, 63,PAGE_SHIFT, %r28 /* Clear any offset bits */ -#else - extrw,u %r26, 24,25, %r26 /* convert phys addr to tlb insert format */ - depw %r25, 31,22, %r28 /* Form aliased virtual address 'to' */ - depwi 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ -#endif + dep_safe %r25, 31,TMPALIAS_SIZE_BITS, %r28 /* Form aliased virtual address 'to' */ + depi_safe 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ /* Purge any old translation */ @@ -881,18 +829,8 @@ ENDPROC_CFI(purge_dcache_page_asm) ENTRY_CFI(flush_icache_page_asm) ldil L%(TMPALIAS_MAP_START), %r28 -#ifdef CONFIG_64BIT -#if (TMPALIAS_MAP_START >= 0x80000000) - depdi 0, 31,32, %r28 /* clear any sign extension */ -#endif - convert_phys_for_tlb_insert20 %r26 /* convert phys addr to tlb insert format */ - depd %r25, 63,22, %r28 /* Form aliased virtual address 'to' */ - depdi 0, 63,PAGE_SHIFT, %r28 /* Clear any offset bits */ -#else - extrw,u %r26, 24,25, %r26 /* convert phys addr to tlb insert format */ - depw %r25, 31,22, %r28 /* Form aliased virtual address 'to' */ - depwi 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ -#endif + dep_safe %r25, 31,TMPALIAS_SIZE_BITS, %r28 /* Form aliased virtual address 'to' */ + depi_safe 0, 31,PAGE_SHIFT, %r28 /* Clear any offset bits */ /* Purge any old translation. Note that the FIC instruction * may use either the instruction or data TLB. Given that we @@ -1098,7 +1036,6 @@ ENTRY_CFI(flush_kernel_dcache_range_asm) sync 89: ALTERNATIVE(88b, 89b, ALT_COND_NO_DCACHE, INSN_NOP) - syncdma bv %r0(%r2) nop ENDPROC_CFI(flush_kernel_dcache_range_asm) @@ -1140,7 +1077,6 @@ ENTRY_CFI(purge_kernel_dcache_range_asm) sync 89: ALTERNATIVE(88b, 89b, ALT_COND_NO_DCACHE, INSN_NOP) - syncdma bv %r0(%r2) nop ENDPROC_CFI(purge_kernel_dcache_range_asm) @@ -1264,7 +1200,7 @@ ENTRY_CFI(flush_kernel_icache_range_asm) nop ENDPROC_CFI(flush_kernel_icache_range_asm) - __INIT + .text /* align should cover use of rfi in disable_sr_hashing_asm and * srdis_done. diff --git a/arch/parisc/kernel/patch.c b/arch/parisc/kernel/patch.c index 80a0ab372802..e59574f65e64 100644 --- a/arch/parisc/kernel/patch.c +++ b/arch/parisc/kernel/patch.c @@ -40,10 +40,7 @@ static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags, *need_unmap = 1; set_fixmap(fixmap, page_to_phys(page)); - if (flags) - raw_spin_lock_irqsave(&patch_lock, *flags); - else - __acquire(&patch_lock); + raw_spin_lock_irqsave(&patch_lock, *flags); return (void *) (__fix_to_virt(fixmap) + (uintaddr & ~PAGE_MASK)); } @@ -52,10 +49,7 @@ static void __kprobes patch_unmap(int fixmap, unsigned long *flags) { clear_fixmap(fixmap); - if (flags) - raw_spin_unlock_irqrestore(&patch_lock, *flags); - else - __release(&patch_lock); + raw_spin_unlock_irqrestore(&patch_lock, *flags); } void __kprobes __patch_text_multiple(void *addr, u32 *insn, unsigned int len) @@ -67,8 +61,9 @@ void __kprobes __patch_text_multiple(void *addr, u32 *insn, unsigned int len) int mapped; /* Make sure we don't have any aliases in cache */ - flush_kernel_vmap_range(addr, len); - flush_icache_range(start, end); + flush_kernel_dcache_range_asm(start, end); + flush_kernel_icache_range_asm(start, end); + flush_tlb_kernel_range(start, end); p = fixmap = patch_map(addr, FIX_TEXT_POKE0, &flags, &mapped); @@ -81,8 +76,10 @@ void __kprobes __patch_text_multiple(void *addr, u32 *insn, unsigned int len) * We're crossing a page boundary, so * need to remap */ - flush_kernel_vmap_range((void *)fixmap, - (p-fixmap) * sizeof(*p)); + flush_kernel_dcache_range_asm((unsigned long)fixmap, + (unsigned long)p); + flush_tlb_kernel_range((unsigned long)fixmap, + (unsigned long)p); if (mapped) patch_unmap(FIX_TEXT_POKE0, &flags); p = fixmap = patch_map(addr, FIX_TEXT_POKE0, &flags, @@ -90,10 +87,10 @@ void __kprobes __patch_text_multiple(void *addr, u32 *insn, unsigned int len) } } - flush_kernel_vmap_range((void *)fixmap, (p-fixmap) * sizeof(*p)); + flush_kernel_dcache_range_asm((unsigned long)fixmap, (unsigned long)p); + flush_tlb_kernel_range((unsigned long)fixmap, (unsigned long)p); if (mapped) patch_unmap(FIX_TEXT_POKE0, &flags); - flush_icache_range(start, end); } void __kprobes __patch_text(void *addr, u32 insn) diff --git a/arch/parisc/kernel/pci-dma.c b/arch/parisc/kernel/pci-dma.c index 36a57aa38e87..ba87f791323b 100644 --- a/arch/parisc/kernel/pci-dma.c +++ b/arch/parisc/kernel/pci-dma.c @@ -36,8 +36,8 @@ #include <asm/tlbflush.h> /* for purge_tlb_*() macros */ static struct proc_dir_entry * proc_gsc_root __read_mostly = NULL; -static unsigned long pcxl_used_bytes __read_mostly = 0; -static unsigned long pcxl_used_pages __read_mostly = 0; +static unsigned long pcxl_used_bytes __read_mostly; +static unsigned long pcxl_used_pages __read_mostly; extern unsigned long pcxl_dma_start; /* Start of pcxl dma mapping area */ static DEFINE_SPINLOCK(pcxl_res_lock); @@ -91,7 +91,7 @@ static inline int map_pte_uncached(pte_t * pte, printk(KERN_ERR "map_pte_uncached: page already exists\n"); purge_tlb_start(flags); set_pte(pte, __mk_pte(*paddr_ptr, PAGE_KERNEL_UNC)); - pdtlb_kernel(orig_vaddr); + pdtlb(SR_KERNEL, orig_vaddr); purge_tlb_end(flags); vaddr += PAGE_SIZE; orig_vaddr += PAGE_SIZE; @@ -175,7 +175,7 @@ static inline void unmap_uncached_pte(pmd_t * pmd, unsigned long vaddr, pte_clear(&init_mm, vaddr, pte); purge_tlb_start(flags); - pdtlb_kernel(orig_vaddr); + pdtlb(SR_KERNEL, orig_vaddr); purge_tlb_end(flags); vaddr += PAGE_SIZE; orig_vaddr += PAGE_SIZE; diff --git a/arch/parisc/kernel/pdc_cons.c b/arch/parisc/kernel/pdc_cons.c index 2661cdd256ae..7d0989f523d0 100644 --- a/arch/parisc/kernel/pdc_cons.c +++ b/arch/parisc/kernel/pdc_cons.c @@ -1,46 +1,18 @@ // SPDX-License-Identifier: GPL-2.0-or-later /* - * PDC Console support - ie use firmware to dump text via boot console + * PDC early console support - use PDC firmware to dump text via boot console * - * Copyright (C) 1999-2003 Matthew Wilcox <willy at parisc-linux.org> - * Copyright (C) 2000 Martin K Petersen <mkp at mkp.net> - * Copyright (C) 2000 John Marvin <jsm at parisc-linux.org> - * Copyright (C) 2000-2003 Paul Bame <bame at parisc-linux.org> - * Copyright (C) 2000 Philipp Rumpf <prumpf with tux.org> - * Copyright (C) 2000 Michael Ang <mang with subcarrier.org> - * Copyright (C) 2000 Grant Grundler <grundler with parisc-linux.org> - * Copyright (C) 2001-2002 Ryan Bradetich <rbrad at parisc-linux.org> - * Copyright (C) 2001 Helge Deller <deller at parisc-linux.org> - * Copyright (C) 2001 Thomas Bogendoerfer <tsbogend at parisc-linux.org> - * Copyright (C) 2002 Randolph Chung <tausq with parisc-linux.org> - * Copyright (C) 2010 Guy Martin <gmsoft at tuxicoman.be> + * Copyright (C) 2001-2022 Helge Deller <deller@gmx.de> */ -/* - * The PDC console is a simple console, which can be used for debugging - * boot related problems on HP PA-RISC machines. It is also useful when no - * other console works. - * - * This code uses the ROM (=PDC) based functions to read and write characters - * from and to PDC's boot path. - */ - -/* Define EARLY_BOOTUP_DEBUG to debug kernel related boot problems. - * On production kernels EARLY_BOOTUP_DEBUG should be undefined. */ -#define EARLY_BOOTUP_DEBUG - - -#include <linux/kernel.h> #include <linux/console.h> -#include <linux/string.h> #include <linux/init.h> -#include <linux/major.h> -#include <linux/tty.h> +#include <linux/serial_core.h> +#include <linux/kgdb.h> #include <asm/page.h> /* for PAGE0 */ #include <asm/pdc.h> /* for iodc_call() proto and friends */ static DEFINE_SPINLOCK(pdc_console_lock); -static struct console pdc_cons; static void pdc_console_write(struct console *co, const char *s, unsigned count) { @@ -54,7 +26,8 @@ static void pdc_console_write(struct console *co, const char *s, unsigned count) spin_unlock_irqrestore(&pdc_console_lock, flags); } -int pdc_console_poll_key(struct console *co) +#ifdef CONFIG_KGDB +static int kgdb_pdc_read_char(void) { int c; unsigned long flags; @@ -63,201 +36,40 @@ int pdc_console_poll_key(struct console *co) c = pdc_iodc_getc(); spin_unlock_irqrestore(&pdc_console_lock, flags); - return c; -} - -static int pdc_console_setup(struct console *co, char *options) -{ - return 0; -} - -#if defined(CONFIG_PDC_CONSOLE) -#include <linux/vt_kern.h> -#include <linux/tty_flip.h> - -#define PDC_CONS_POLL_DELAY (30 * HZ / 1000) - -static void pdc_console_poll(struct timer_list *unused); -static DEFINE_TIMER(pdc_console_timer, pdc_console_poll); -static struct tty_port tty_port; - -static int pdc_console_tty_open(struct tty_struct *tty, struct file *filp) -{ - tty_port_tty_set(&tty_port, tty); - mod_timer(&pdc_console_timer, jiffies + PDC_CONS_POLL_DELAY); - - return 0; + return (c <= 0) ? NO_POLL_CHAR : c; } -static void pdc_console_tty_close(struct tty_struct *tty, struct file *filp) +static void kgdb_pdc_write_char(u8 chr) { - if (tty->count == 1) { - del_timer_sync(&pdc_console_timer); - tty_port_tty_set(&tty_port, NULL); - } + if (PAGE0->mem_cons.cl_class != CL_DUPLEX) + pdc_console_write(NULL, &chr, 1); } -static int pdc_console_tty_write(struct tty_struct *tty, const unsigned char *buf, int count) -{ - pdc_console_write(NULL, buf, count); - return count; -} - -static unsigned int pdc_console_tty_write_room(struct tty_struct *tty) -{ - return 32768; /* no limit, no buffer used */ -} - -static const struct tty_operations pdc_console_tty_ops = { - .open = pdc_console_tty_open, - .close = pdc_console_tty_close, - .write = pdc_console_tty_write, - .write_room = pdc_console_tty_write_room, +static struct kgdb_io kgdb_pdc_io_ops = { + .name = "kgdb_pdc", + .read_char = kgdb_pdc_read_char, + .write_char = kgdb_pdc_write_char, }; - -static void pdc_console_poll(struct timer_list *unused) -{ - int data, count = 0; - - while (1) { - data = pdc_console_poll_key(NULL); - if (data == -1) - break; - tty_insert_flip_char(&tty_port, data & 0xFF, TTY_NORMAL); - count ++; - } - - if (count) - tty_flip_buffer_push(&tty_port); - - if (pdc_cons.flags & CON_ENABLED) - mod_timer(&pdc_console_timer, jiffies + PDC_CONS_POLL_DELAY); -} - -static struct tty_driver *pdc_console_tty_driver; - -static int __init pdc_console_tty_driver_init(void) -{ - struct tty_driver *driver; - int err; - - /* Check if the console driver is still registered. - * It is unregistered if the pdc console was not selected as the - * primary console. */ - - struct console *tmp; - - console_lock(); - for_each_console(tmp) - if (tmp == &pdc_cons) - break; - console_unlock(); - - if (!tmp) { - printk(KERN_INFO "PDC console driver not registered anymore, not creating %s\n", pdc_cons.name); - return -ENODEV; - } - - printk(KERN_INFO "The PDC console driver is still registered, removing CON_BOOT flag\n"); - pdc_cons.flags &= ~CON_BOOT; - - driver = tty_alloc_driver(1, TTY_DRIVER_REAL_RAW | - TTY_DRIVER_RESET_TERMIOS); - if (IS_ERR(driver)) - return PTR_ERR(driver); - - tty_port_init(&tty_port); - - driver->driver_name = "pdc_cons"; - driver->name = "ttyB"; - driver->major = MUX_MAJOR; - driver->minor_start = 0; - driver->type = TTY_DRIVER_TYPE_SYSTEM; - driver->init_termios = tty_std_termios; - tty_set_operations(driver, &pdc_console_tty_ops); - tty_port_link_device(&tty_port, driver, 0); - - err = tty_register_driver(driver); - if (err) { - printk(KERN_ERR "Unable to register the PDC console TTY driver\n"); - tty_port_destroy(&tty_port); - tty_driver_kref_put(driver); - return err; - } - - pdc_console_tty_driver = driver; - - return 0; -} -device_initcall(pdc_console_tty_driver_init); - -static struct tty_driver * pdc_console_device (struct console *c, int *index) -{ - *index = c->index; - return pdc_console_tty_driver; -} -#else -#define pdc_console_device NULL #endif -static struct console pdc_cons = { - .name = "ttyB", - .write = pdc_console_write, - .device = pdc_console_device, - .setup = pdc_console_setup, - .flags = CON_BOOT | CON_PRINTBUFFER, - .index = -1, -}; - -static int pdc_console_initialized; - -static void pdc_console_init_force(void) +static int __init pdc_earlycon_setup(struct earlycon_device *device, + const char *opt) { - if (pdc_console_initialized) - return; - ++pdc_console_initialized; - + struct console *earlycon_console; + /* If the console is duplex then copy the COUT parameters to CIN. */ if (PAGE0->mem_cons.cl_class == CL_DUPLEX) memcpy(&PAGE0->mem_kbd, &PAGE0->mem_cons, sizeof(PAGE0->mem_cons)); - /* register the pdc console */ - register_console(&pdc_cons); -} + earlycon_console = device->con; + earlycon_console->write = pdc_console_write; + device->port.iotype = UPIO_MEM32BE; -void __init pdc_console_init(void) -{ -#if defined(EARLY_BOOTUP_DEBUG) || defined(CONFIG_PDC_CONSOLE) - pdc_console_init_force(); +#ifdef CONFIG_KGDB + kgdb_register_io_module(&kgdb_pdc_io_ops); #endif -#ifdef EARLY_BOOTUP_DEBUG - printk(KERN_INFO "Initialized PDC Console for debugging.\n"); -#endif -} - - -/* - * Used for emergencies. Currently only used if an HPMC occurs. If an - * HPMC occurs, it is possible that the current console may not be - * properly initialised after the PDC IO reset. This routine unregisters - * all of the current consoles, reinitializes the pdc console and - * registers it. - */ - -void pdc_console_restart(void) -{ - struct console *console; - - if (pdc_console_initialized) - return; - /* If we've already seen the output, don't bother to print it again */ - if (console_drivers != NULL) - pdc_cons.flags &= ~CON_PRINTBUFFER; - - while ((console = console_drivers) != NULL) - unregister_console(console_drivers); - - /* force registering the pdc console */ - pdc_console_init_force(); + return 0; } + +EARLYCON_DECLARE(pdc, pdc_earlycon_setup); diff --git a/arch/parisc/kernel/process.c b/arch/parisc/kernel/process.c index ea3d83b6fb62..c4f8374c7018 100644 --- a/arch/parisc/kernel/process.c +++ b/arch/parisc/kernel/process.c @@ -26,6 +26,7 @@ #include <linux/module.h> #include <linux/personality.h> #include <linux/ptrace.h> +#include <linux/reboot.h> #include <linux/sched.h> #include <linux/sched/debug.h> #include <linux/sched/task.h> @@ -38,6 +39,7 @@ #include <linux/rcupdate.h> #include <linux/random.h> #include <linux/nmi.h> +#include <linux/sched/hotplug.h> #include <asm/io.h> #include <asm/asm-offsets.h> @@ -46,6 +48,7 @@ #include <asm/pdc_chassis.h> #include <asm/unwind.h> #include <asm/sections.h> +#include <asm/cacheflush.h> #define COMMAND_GLOBAL F_EXTEND(0xfffe0030) #define CMD_RESET 5 /* reset any module */ @@ -114,8 +117,7 @@ void machine_power_off(void) pdc_chassis_send_status(PDC_CHASSIS_DIRECT_SHUTDOWN); /* ipmi_poweroff may have been installed. */ - if (pm_power_off) - pm_power_off(); + do_kernel_power_off(); /* It seems we have no way to power the system off via * software. The user has to press the button himself. */ @@ -144,10 +146,6 @@ void flush_thread(void) */ } -void release_thread(struct task_struct *dead_task) -{ -} - /* * Idle thread support * @@ -158,10 +156,29 @@ void release_thread(struct task_struct *dead_task) int running_on_qemu __ro_after_init; EXPORT_SYMBOL(running_on_qemu); -void __cpuidle arch_cpu_idle_dead(void) +/* + * Called from the idle thread for the CPU which has been shutdown. + */ +void arch_cpu_idle_dead(void) { - /* nop on real hardware, qemu will offline CPU. */ - asm volatile("or %%r31,%%r31,%%r31\n":::); +#ifdef CONFIG_HOTPLUG_CPU + idle_task_exit(); + + local_irq_disable(); + + /* Tell __cpu_die() that this CPU is now safe to dispose of. */ + (void)cpu_report_death(); + + /* Ensure that the cache lines are written out. */ + flush_cache_all_local(); + flush_tlb_all_local(NULL); + + /* Let PDC firmware put CPU into firmware idle loop. */ + __pdc_cpu_rendezvous(); + + pr_warn("PDC does not provide rendezvous function.\n"); +#endif + while (1); } void __cpuidle arch_cpu_idle(void) @@ -185,9 +202,11 @@ arch_initcall(parisc_idle_init); * Copy architecture-specific thread state */ int -copy_thread(unsigned long clone_flags, unsigned long usp, - unsigned long kthread_arg, struct task_struct *p, unsigned long tls) +copy_thread(struct task_struct *p, const struct kernel_clone_args *args) { + unsigned long clone_flags = args->flags; + unsigned long usp = args->stack; + unsigned long tls = args->tls; struct pt_regs *cregs = &(p->thread.regs); void *stack = task_stack_page(p); @@ -197,10 +216,10 @@ copy_thread(unsigned long clone_flags, unsigned long usp, extern void * const ret_from_kernel_thread; extern void * const child_return; - if (unlikely(p->flags & (PF_KTHREAD | PF_IO_WORKER))) { + if (unlikely(args->fn)) { /* kernel thread */ memset(cregs, 0, sizeof(struct pt_regs)); - if (!usp) /* idle thread */ + if (args->idle) /* idle thread */ return 0; /* Must exit via ret_from_kernel_thread in order * to call schedule_tail() @@ -212,12 +231,12 @@ copy_thread(unsigned long clone_flags, unsigned long usp, * ret_from_kernel_thread. */ #ifdef CONFIG_64BIT - cregs->gr[27] = ((unsigned long *)usp)[3]; - cregs->gr[26] = ((unsigned long *)usp)[2]; + cregs->gr[27] = ((unsigned long *)args->fn)[3]; + cregs->gr[26] = ((unsigned long *)args->fn)[2]; #else - cregs->gr[26] = usp; + cregs->gr[26] = (unsigned long) args->fn; #endif - cregs->gr[25] = kthread_arg; + cregs->gr[25] = (unsigned long) args->fn_arg; } else { /* user thread */ /* usp must be word aligned. This also prevents users from @@ -263,30 +282,9 @@ __get_wchan(struct task_struct *p) return 0; } -#ifdef CONFIG_64BIT -void *dereference_function_descriptor(void *ptr) -{ - Elf64_Fdesc *desc = ptr; - void *p; - - if (!get_kernel_nofault(p, (void *)&desc->addr)) - ptr = p; - return ptr; -} - -void *dereference_kernel_function_descriptor(void *ptr) -{ - if (ptr < (void *)__start_opd || - ptr >= (void *)__end_opd) - return ptr; - - return dereference_function_descriptor(ptr); -} -#endif - static inline unsigned long brk_rnd(void) { - return (get_random_int() & BRK_RND_MASK) << PAGE_SHIFT; + return (get_random_u32() & BRK_RND_MASK) << PAGE_SHIFT; } unsigned long arch_randomize_brk(struct mm_struct *mm) diff --git a/arch/parisc/kernel/processor.c b/arch/parisc/kernel/processor.c index 1b6129e7d776..dddaaa6e7a82 100644 --- a/arch/parisc/kernel/processor.c +++ b/arch/parisc/kernel/processor.c @@ -19,6 +19,7 @@ #include <linux/random.h> #include <linux/slab.h> #include <linux/cpu.h> +#include <asm/topology.h> #include <asm/param.h> #include <asm/cache.h> #include <asm/hardware.h> /* for register_parisc_driver() stuff */ @@ -170,6 +171,7 @@ static int __init processor_probe(struct parisc_device *dev) p->cpu_num = cpu_info.cpu_num; p->cpu_loc = cpu_info.cpu_loc; + set_cpu_possible(cpuid, true); store_cpu_topology(cpuid); #ifdef CONFIG_SMP @@ -317,7 +319,7 @@ void __init collect_boot_cpu_data(void) * * o Enable CPU profiling hooks. */ -int __init init_per_cpu(int cpunum) +int init_per_cpu(int cpunum) { int ret; struct pdc_coproc_cfg coproc_cfg; @@ -325,8 +327,6 @@ int __init init_per_cpu(int cpunum) set_firmware_width(); ret = pdc_coproc_cfg(&coproc_cfg); - store_cpu_topology(cpunum); - if(ret >= 0 && coproc_cfg.ccr_functional) { mtctl(coproc_cfg.ccr_functional, 10); /* 10 == Coprocessor Control Reg */ @@ -390,7 +390,7 @@ show_cpuinfo (struct seq_file *m, void *v) boot_cpu_data.cpu_hz / 1000000, boot_cpu_data.cpu_hz % 1000000 ); -#ifdef CONFIG_PARISC_CPU_TOPOLOGY +#ifdef CONFIG_GENERIC_ARCH_TOPOLOGY seq_printf(m, "physical id\t: %d\n", topology_physical_package_id(cpu)); seq_printf(m, "siblings\t: %d\n", @@ -418,8 +418,7 @@ show_cpuinfo (struct seq_file *m, void *v) } seq_printf(m, " (0x%02lx)\n", boot_cpu_data.pdc.capabilities); - seq_printf(m, "model\t\t: %s\n" - "model name\t: %s\n", + seq_printf(m, "model\t\t: %s - %s\n", boot_cpu_data.pdc.sys_model_name, cpuinfo->dev ? cpuinfo->dev->name : "Unknown"); @@ -460,5 +459,13 @@ static struct parisc_driver cpu_driver __refdata = { */ void __init processor_init(void) { + unsigned int cpu; + + reset_cpu_topology(); + + /* reset possible mask. We will mark those which are possible. */ + for_each_possible_cpu(cpu) + set_cpu_possible(cpu, false); + register_parisc_driver(&cpu_driver); } diff --git a/arch/parisc/kernel/ptrace.c b/arch/parisc/kernel/ptrace.c index 65de6c4c9354..96ef6a6b66e5 100644 --- a/arch/parisc/kernel/ptrace.c +++ b/arch/parisc/kernel/ptrace.c @@ -15,7 +15,6 @@ #include <linux/elf.h> #include <linux/errno.h> #include <linux/ptrace.h> -#include <linux/tracehook.h> #include <linux/user.h> #include <linux/personality.h> #include <linux/regset.h> @@ -316,7 +315,7 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request, long do_syscall_trace_enter(struct pt_regs *regs) { if (test_thread_flag(TIF_SYSCALL_TRACE)) { - int rc = tracehook_report_syscall_entry(regs); + int rc = ptrace_report_syscall_entry(regs); /* * As tracesys_next does not set %r28 to -ENOSYS @@ -327,7 +326,7 @@ long do_syscall_trace_enter(struct pt_regs *regs) if (rc) { /* * A nonzero return code from - * tracehook_report_syscall_entry() tells us + * ptrace_report_syscall_entry() tells us * to prevent the syscall execution. Skip * the syscall call and the syscall restart handling. * @@ -381,7 +380,7 @@ void do_syscall_trace_exit(struct pt_regs *regs) #endif if (stepping || test_thread_flag(TIF_SYSCALL_TRACE)) - tracehook_report_syscall_exit(regs, stepping); + ptrace_report_syscall_exit(regs, stepping); } diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c index cceb09855e03..375f38d6e1a4 100644 --- a/arch/parisc/kernel/setup.c +++ b/arch/parisc/kernel/setup.c @@ -48,6 +48,7 @@ struct proc_dir_entry * proc_mckinley_root __read_mostly = NULL; void __init setup_cmdline(char **cmdline_p) { extern unsigned int boot_args[]; + char *p; /* Collect stuff passed in from the boot loader */ @@ -56,9 +57,23 @@ void __init setup_cmdline(char **cmdline_p) /* called from hpux boot loader */ boot_command_line[0] = '\0'; } else { - strlcpy(boot_command_line, (char *)__va(boot_args[1]), + strscpy(boot_command_line, (char *)__va(boot_args[1]), COMMAND_LINE_SIZE); + /* autodetect console type (if not done by palo yet) */ + p = boot_command_line; + if (!str_has_prefix(p, "console=") && !strstr(p, " console=")) { + strlcat(p, " console=", COMMAND_LINE_SIZE); + if (PAGE0->mem_cons.cl_class == CL_DUPLEX) + strlcat(p, "ttyS0", COMMAND_LINE_SIZE); + else + strlcat(p, "tty0", COMMAND_LINE_SIZE); + } + + /* default to use early console */ + if (!strstr(p, "earlycon")) + strlcat(p, " earlycon=pdc", COMMAND_LINE_SIZE); + #ifdef CONFIG_BLK_DEV_INITRD if (boot_args[2] != 0) /* did palo pass us a ramdisk? */ { @@ -68,7 +83,7 @@ void __init setup_cmdline(char **cmdline_p) #endif } - strcpy(command_line, boot_command_line); + strscpy(command_line, boot_command_line, COMMAND_LINE_SIZE); *cmdline_p = command_line; } @@ -128,8 +143,6 @@ void __init setup_arch(char **cmdline_p) if (__pa((unsigned long) &_end) >= KERNEL_INITIAL_SIZE) panic("KERNEL_INITIAL_ORDER too small!"); - pdc_console_init(); - #ifdef CONFIG_64BIT if(parisc_narrow_firmware) { printk(KERN_INFO "Kernel is using PDC in 32-bit mode.\n"); @@ -150,6 +163,8 @@ void __init setup_arch(char **cmdline_p) #ifdef CONFIG_PA11 dma_ops_init(); #endif + + clear_sched_clock_stable(); } /* diff --git a/arch/parisc/kernel/signal.c b/arch/parisc/kernel/signal.c index 46b1050640b8..8bc0ddaa6219 100644 --- a/arch/parisc/kernel/signal.c +++ b/arch/parisc/kernel/signal.c @@ -1,16 +1,13 @@ // SPDX-License-Identifier: GPL-2.0 /* - * linux/arch/parisc/kernel/signal.c: Architecture-specific signal - * handling support. + * PA-RISC architecture-specific signal handling support. * * Copyright (C) 2000 David Huggins-Daines <dhd@debian.org> * Copyright (C) 2000 Linuxcare, Inc. + * Copyright (C) 2000-2022 Helge Deller <deller@gmx.de> + * Copyright (C) 2022 John David Anglin <dave.anglin@bell.net> * * Based on the ia64, i386, and alpha versions. - * - * Like the IA-64, we are a recent enough port (we are *starting* - * with glibc2.2) that we do not need to support the old non-realtime - * Linux signals. Therefore we don't. */ #include <linux/sched.h> @@ -22,7 +19,7 @@ #include <linux/errno.h> #include <linux/wait.h> #include <linux/ptrace.h> -#include <linux/tracehook.h> +#include <linux/resume_user_mode.h> #include <linux/unistd.h> #include <linux/stddef.h> #include <linux/compat.h> @@ -32,6 +29,7 @@ #include <linux/uaccess.h> #include <asm/cacheflush.h> #include <asm/asm-offsets.h> +#include <asm/vdso.h> #ifdef CONFIG_COMPAT #include "signal32.h" @@ -59,14 +57,6 @@ * Do a signal return - restore sigcontext. */ -/* Trampoline for calling rt_sigreturn() */ -#define INSN_LDI_R25_0 0x34190000 /* ldi 0,%r25 (in_syscall=0) */ -#define INSN_LDI_R25_1 0x34190002 /* ldi 1,%r25 (in_syscall=1) */ -#define INSN_LDI_R20 0x3414015a /* ldi __NR_rt_sigreturn,%r20 */ -#define INSN_BLE_SR2_R0 0xe4008200 /* be,l 0x100(%sr2,%r0),%sr0,%r31 */ -/* For debugging */ -#define INSN_DIE_HORRIBLY 0x68000ccc /* stw %r0,0x666(%sr0,%r0) */ - static long restore_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs) { @@ -77,9 +67,9 @@ restore_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs) err |= __copy_from_user(regs->iaoq, sc->sc_iaoq, sizeof(regs->iaoq)); err |= __copy_from_user(regs->iasq, sc->sc_iasq, sizeof(regs->iasq)); err |= __get_user(regs->sar, &sc->sc_sar); - DBG(2,"restore_sigcontext: iaoq is %#lx / %#lx\n", - regs->iaoq[0],regs->iaoq[1]); - DBG(2,"restore_sigcontext: r28 is %ld\n", regs->gr[28]); + DBG(2, "%s: iaoq is %#lx / %#lx\n", + __func__, regs->iaoq[0], regs->iaoq[1]); + DBG(2, "%s: r28 is %ld\n", __func__, regs->gr[28]); return err; } @@ -102,7 +92,7 @@ sys_rt_sigreturn(struct pt_regs *regs, int in_syscall) /* Unwind the user stack to get the rt_sigframe structure. */ frame = (struct rt_sigframe __user *) (usp - sigframe_size); - DBG(2,"sys_rt_sigreturn: frame is %p\n", frame); + DBG(2, "%s: frame is %p pid %d\n", __func__, frame, task_pid_nr(current)); regs->orig_r28 = 1; /* no restarts for sigreturn */ @@ -110,7 +100,6 @@ sys_rt_sigreturn(struct pt_regs *regs, int in_syscall) compat_frame = (struct compat_rt_sigframe __user *)frame; if (is_compat_task()) { - DBG(2,"sys_rt_sigreturn: ELF32 process.\n"); if (get_compat_sigset(&set, &compat_frame->uc.uc_sigmask)) goto give_sigsegv; } else @@ -125,25 +114,25 @@ sys_rt_sigreturn(struct pt_regs *regs, int in_syscall) /* Good thing we saved the old gr[30], eh? */ #ifdef CONFIG_64BIT if (is_compat_task()) { - DBG(1,"sys_rt_sigreturn: compat_frame->uc.uc_mcontext 0x%p\n", - &compat_frame->uc.uc_mcontext); + DBG(1, "%s: compat_frame->uc.uc_mcontext 0x%p\n", + __func__, &compat_frame->uc.uc_mcontext); // FIXME: Load upper half from register file if (restore_sigcontext32(&compat_frame->uc.uc_mcontext, &compat_frame->regs, regs)) goto give_sigsegv; - DBG(1,"sys_rt_sigreturn: usp %#08lx stack 0x%p\n", - usp, &compat_frame->uc.uc_stack); + DBG(1, "%s: usp %#08lx stack 0x%p\n", + __func__, usp, &compat_frame->uc.uc_stack); if (compat_restore_altstack(&compat_frame->uc.uc_stack)) goto give_sigsegv; } else #endif { - DBG(1,"sys_rt_sigreturn: frame->uc.uc_mcontext 0x%p\n", - &frame->uc.uc_mcontext); + DBG(1, "%s: frame->uc.uc_mcontext 0x%p\n", + __func__, &frame->uc.uc_mcontext); if (restore_sigcontext(&frame->uc.uc_mcontext, regs)) goto give_sigsegv; - DBG(1,"sys_rt_sigreturn: usp %#08lx stack 0x%p\n", - usp, &frame->uc.uc_stack); + DBG(1, "%s: usp %#08lx stack 0x%p\n", + __func__, usp, &frame->uc.uc_stack); if (restore_altstack(&frame->uc.uc_stack)) goto give_sigsegv; } @@ -155,14 +144,11 @@ sys_rt_sigreturn(struct pt_regs *regs, int in_syscall) */ if (in_syscall) regs->gr[31] = regs->iaoq[0]; -#if DEBUG_SIG - DBG(1,"sys_rt_sigreturn: returning to %#lx, DUMPING REGS:\n", regs->iaoq[0]); - show_regs(regs); -#endif + return; give_sigsegv: - DBG(1,"sys_rt_sigreturn: Sending SIGSEGV\n"); + DBG(1, "%s: Sending SIGSEGV\n", __func__); force_sig(SIGSEGV); return; } @@ -177,15 +163,15 @@ get_sigframe(struct k_sigaction *ka, unsigned long sp, size_t frame_size) /*FIXME: ELF32 vs. ELF64 has different frame_size, but since we don't use the parameter it doesn't matter */ - DBG(1,"get_sigframe: ka = %#lx, sp = %#lx, frame_size = %#lx\n", - (unsigned long)ka, sp, frame_size); + DBG(1, "%s: ka = %#lx, sp = %#lx, frame_size = %zu\n", + __func__, (unsigned long)ka, sp, frame_size); /* Align alternate stack and reserve 64 bytes for the signal handler's frame marker. */ if ((ka->sa.sa_flags & SA_ONSTACK) != 0 && ! sas_ss_flags(sp)) sp = (current->sas_ss_sp + 0x7f) & ~0x3f; /* Stacks grow up! */ - DBG(1,"get_sigframe: Returning sp = %#lx\n", (unsigned long)sp); + DBG(1, "%s: Returning sp = %#lx\n", __func__, (unsigned long)sp); return (void __user *) sp; /* Stacks grow up. Fun. */ } @@ -205,20 +191,20 @@ setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, int in_sysc err |= __put_user(regs->gr[31]+4, &sc->sc_iaoq[1]); err |= __put_user(regs->sr[3], &sc->sc_iasq[0]); err |= __put_user(regs->sr[3], &sc->sc_iasq[1]); - DBG(1,"setup_sigcontext: iaoq %#lx / %#lx (in syscall)\n", - regs->gr[31], regs->gr[31]+4); + DBG(1, "%s: iaoq %#lx / %#lx (in syscall)\n", + __func__, regs->gr[31], regs->gr[31]+4); } else { err |= __copy_to_user(sc->sc_iaoq, regs->iaoq, sizeof(regs->iaoq)); err |= __copy_to_user(sc->sc_iasq, regs->iasq, sizeof(regs->iasq)); - DBG(1,"setup_sigcontext: iaoq %#lx / %#lx (not in syscall)\n", - regs->iaoq[0], regs->iaoq[1]); + DBG(1, "%s: iaoq %#lx / %#lx (not in syscall)\n", + __func__, regs->iaoq[0], regs->iaoq[1]); } err |= __put_user(flags, &sc->sc_flags); err |= __copy_to_user(sc->sc_gr, regs->gr, sizeof(regs->gr)); err |= __copy_to_user(sc->sc_fr, regs->fr, sizeof(regs->fr)); err |= __put_user(regs->sar, &sc->sc_sar); - DBG(1,"setup_sigcontext: r28 is %ld\n", regs->gr[28]); + DBG(1, "%s: r28 is %ld\n", __func__, regs->gr[28]); return err; } @@ -230,7 +216,7 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, struct rt_sigframe __user *frame; unsigned long rp, usp; unsigned long haddr, sigframe_size; - unsigned long start, end; + unsigned long start; int err = 0; #ifdef CONFIG_64BIT struct compat_rt_sigframe __user * compat_frame; @@ -247,11 +233,10 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, #endif frame = get_sigframe(&ksig->ka, usp, sigframe_size); - DBG(1,"SETUP_RT_FRAME: START\n"); - DBG(1,"setup_rt_frame: frame %p info %p\n", frame, ksig->info); + DBG(1, "%s: frame %p info %p\n", __func__, frame, &ksig->info); start = (unsigned long) frame; - if (start >= user_addr_max() - sigframe_size) + if (start >= TASK_SIZE_MAX - sigframe_size) return -EFAULT; #ifdef CONFIG_64BIT @@ -259,11 +244,12 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, compat_frame = (struct compat_rt_sigframe __user *)frame; if (is_compat_task()) { - DBG(1,"setup_rt_frame: frame->info = 0x%p\n", &compat_frame->info); + DBG(1, "%s: frame->info = 0x%p\n", __func__, &compat_frame->info); err |= copy_siginfo_to_user32(&compat_frame->info, &ksig->info); err |= __compat_save_altstack( &compat_frame->uc.uc_stack, regs->gr[30]); - DBG(1,"setup_rt_frame: frame->uc = 0x%p\n", &compat_frame->uc); - DBG(1,"setup_rt_frame: frame->uc.uc_mcontext = 0x%p\n", &compat_frame->uc.uc_mcontext); + DBG(1, "%s: frame->uc = 0x%p\n", __func__, &compat_frame->uc); + DBG(1, "%s: frame->uc.uc_mcontext = 0x%p\n", + __func__, &compat_frame->uc.uc_mcontext); err |= setup_sigcontext32(&compat_frame->uc.uc_mcontext, &compat_frame->regs, regs, in_syscall); err |= put_compat_sigset(&compat_frame->uc.uc_sigmask, set, @@ -271,11 +257,12 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, } else #endif { - DBG(1,"setup_rt_frame: frame->info = 0x%p\n", &frame->info); + DBG(1, "%s: frame->info = 0x%p\n", __func__, &frame->info); err |= copy_siginfo_to_user(&frame->info, &ksig->info); err |= __save_altstack(&frame->uc.uc_stack, regs->gr[30]); - DBG(1,"setup_rt_frame: frame->uc = 0x%p\n", &frame->uc); - DBG(1,"setup_rt_frame: frame->uc.uc_mcontext = 0x%p\n", &frame->uc.uc_mcontext); + DBG(1, "%s: frame->uc = 0x%p\n", __func__, &frame->uc); + DBG(1, "%s: frame->uc.uc_mcontext = 0x%p\n", + __func__, &frame->uc.uc_mcontext); err |= setup_sigcontext(&frame->uc.uc_mcontext, regs, in_syscall); /* FIXME: Should probably be converted as well for the compat case */ err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); @@ -284,32 +271,15 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, if (err) return -EFAULT; - /* Set up to return from userspace. If provided, use a stub - already in userspace. The first words of tramp are used to - save the previous sigrestartblock trampoline that might be - on the stack. We start the sigreturn trampoline at - SIGRESTARTBLOCK_TRAMP+X. */ - err |= __put_user(in_syscall ? INSN_LDI_R25_1 : INSN_LDI_R25_0, - &frame->tramp[SIGRESTARTBLOCK_TRAMP+0]); - err |= __put_user(INSN_LDI_R20, - &frame->tramp[SIGRESTARTBLOCK_TRAMP+1]); - err |= __put_user(INSN_BLE_SR2_R0, - &frame->tramp[SIGRESTARTBLOCK_TRAMP+2]); - err |= __put_user(INSN_NOP, &frame->tramp[SIGRESTARTBLOCK_TRAMP+3]); - - start = (unsigned long) &frame->tramp[0]; - end = (unsigned long) &frame->tramp[TRAMP_SIZE]; - flush_user_dcache_range_asm(start, end); - flush_user_icache_range_asm(start, end); - - /* TRAMP Words 0-4, Length 5 = SIGRESTARTBLOCK_TRAMP - * TRAMP Words 5-9, Length 4 = SIGRETURN_TRAMP - * So the SIGRETURN_TRAMP is at the end of SIGRESTARTBLOCK_TRAMP - */ - rp = (unsigned long) &frame->tramp[SIGRESTARTBLOCK_TRAMP]; +#ifdef CONFIG_64BIT + if (!is_compat_task()) + rp = VDSO64_SYMBOL(current, sigtramp_rt); + else +#endif + rp = VDSO32_SYMBOL(current, sigtramp_rt); - if (err) - return -EFAULT; + if (in_syscall) + rp += 4*4; /* skip 4 instructions and start at ldi 1,%r25 */ haddr = A(ksig->ka.sa.sa_handler); /* The sa_handler may be a pointer to a function descriptor */ @@ -340,8 +310,8 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, haddr = fdesc.addr; regs->gr[19] = fdesc.gp; - DBG(1,"setup_rt_frame: 64 bit signal, exe=%#lx, r19=%#lx, in_syscall=%d\n", - haddr, regs->gr[19], in_syscall); + DBG(1, "%s: 64 bit signal, exe=%#lx, r19=%#lx, in_syscall=%d\n", + __func__, haddr, regs->gr[19], in_syscall); } #endif @@ -351,7 +321,7 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, regs->gr[31] = haddr; #ifdef CONFIG_64BIT if (!test_thread_flag(TIF_32BIT)) - sigframe_size |= 1; + sigframe_size |= 1; /* XXX ???? */ #endif } else { unsigned long psw = USER_PSW; @@ -373,11 +343,11 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, } regs->gr[0] = psw; - regs->iaoq[0] = haddr | 3; + regs->iaoq[0] = haddr | PRIV_USER; regs->iaoq[1] = regs->iaoq[0] + 4; } - regs->gr[2] = rp; /* userland return pointer */ + regs->gr[2] = rp; /* userland return pointer */ regs->gr[26] = ksig->sig; /* signal number */ #ifdef CONFIG_64BIT @@ -391,15 +361,15 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, regs->gr[24] = A(&frame->uc); /* ucontext pointer */ } - DBG(1,"setup_rt_frame: making sigreturn frame: %#lx + %#lx = %#lx\n", + DBG(1, "%s: making sigreturn frame: %#lx + %#lx = %#lx\n", __func__, regs->gr[30], sigframe_size, regs->gr[30] + sigframe_size); /* Raise the user stack pointer to make a proper call frame. */ regs->gr[30] = (A(frame) + sigframe_size); - DBG(1,"setup_rt_frame: sig deliver (%s,%d) frame=0x%p sp=%#lx iaoq=%#lx/%#lx rp=%#lx\n", - current->comm, current->pid, frame, regs->gr[30], + DBG(1, "%s: sig deliver (%s,%d) frame=0x%p sp=%#lx iaoq=%#lx/%#lx rp=%#lx\n", + __func__, current->comm, current->pid, frame, regs->gr[30], regs->iaoq[0], regs->iaoq[1], rp); return 0; @@ -415,8 +385,8 @@ handle_signal(struct ksignal *ksig, struct pt_regs *regs, int in_syscall) int ret; sigset_t *oldset = sigmask_to_save(); - DBG(1,"handle_signal: sig=%ld, ka=%p, info=%p, oldset=%p, regs=%p\n", - ksig->sig, ksig->ka, ksig->info, oldset, regs); + DBG(1, "%s: sig=%d, ka=%p, info=%p, oldset=%p, regs=%p\n", + __func__, ksig->sig, &ksig->ka, &ksig->info, oldset, regs); /* Set up the stack frame */ ret = setup_rt_frame(ksig, oldset, regs, in_syscall); @@ -424,8 +394,8 @@ handle_signal(struct ksignal *ksig, struct pt_regs *regs, int in_syscall) signal_setup_done(ret, ksig, test_thread_flag(TIF_SINGLESTEP) || test_thread_flag(TIF_BLOCKSTEP)); - DBG(1,KERN_DEBUG "do_signal: Exit (success), regs->gr[28] = %ld\n", - regs->gr[28]); + DBG(1, "%s: Exit (success), regs->gr[28] = %ld\n", + __func__, regs->gr[28]); } /* @@ -483,21 +453,27 @@ syscall_restart(struct pt_regs *regs, struct k_sigaction *ka) if (regs->orig_r28) return; regs->orig_r28 = 1; /* no more restarts */ + + DBG(1, "%s: orig_r28 = %ld pid %d r20 %ld\n", + __func__, regs->orig_r28, task_pid_nr(current), regs->gr[20]); + /* Check the return code */ switch (regs->gr[28]) { case -ERESTART_RESTARTBLOCK: case -ERESTARTNOHAND: - DBG(1,"ERESTARTNOHAND: returning -EINTR\n"); + DBG(1, "%s: ERESTARTNOHAND: returning -EINTR\n", __func__); regs->gr[28] = -EINTR; break; case -ERESTARTSYS: if (!(ka->sa.sa_flags & SA_RESTART)) { - DBG(1,"ERESTARTSYS: putting -EINTR\n"); + DBG(1, "%s: ERESTARTSYS: putting -EINTR pid %d\n", + __func__, task_pid_nr(current)); regs->gr[28] = -EINTR; break; } fallthrough; case -ERESTARTNOINTR: + DBG(1, "%s: %ld\n", __func__, regs->gr[28]); check_syscallno_in_delay_branch(regs); break; } @@ -509,50 +485,52 @@ insert_restart_trampoline(struct pt_regs *regs) if (regs->orig_r28) return; regs->orig_r28 = 1; /* no more restarts */ - switch(regs->gr[28]) { + + DBG(2, "%s: gr28 = %ld pid %d\n", + __func__, regs->gr[28], task_pid_nr(current)); + + switch (regs->gr[28]) { case -ERESTART_RESTARTBLOCK: { /* Restart the system call - no handlers present */ unsigned int *usp = (unsigned int *)regs->gr[30]; - unsigned long start = (unsigned long) &usp[2]; - unsigned long end = (unsigned long) &usp[5]; + unsigned long rp; long err = 0; /* check that we don't exceed the stack */ - if (A(&usp[0]) >= user_addr_max() - 5 * sizeof(int)) + if (A(&usp[0]) >= TASK_SIZE_MAX - 5 * sizeof(int)) return; - /* Setup a trampoline to restart the syscall - * with __NR_restart_syscall + /* Call trampoline in vdso to restart the syscall + * with __NR_restart_syscall. + * Original return addresses are on stack like this: * * 0: <return address (orig r31)> * 4: <2nd half for 64-bit> - * 8: ldw 0(%sp), %r31 - * 12: be 0x100(%sr2, %r0) - * 16: ldi __NR_restart_syscall, %r20 */ #ifdef CONFIG_64BIT - err |= put_user(regs->gr[31] >> 32, &usp[0]); - err |= put_user(regs->gr[31] & 0xffffffff, &usp[1]); - err |= put_user(0x0fc010df, &usp[2]); -#else - err |= put_user(regs->gr[31], &usp[0]); - err |= put_user(0x0fc0109f, &usp[2]); + if (!is_compat_task()) { + err |= put_user(regs->gr[31] >> 32, &usp[0]); + err |= put_user(regs->gr[31] & 0xffffffff, &usp[1]); + rp = VDSO64_SYMBOL(current, restart_syscall); + } else #endif - err |= put_user(0xe0008200, &usp[3]); - err |= put_user(0x34140000, &usp[4]); - + { + err |= put_user(regs->gr[31], &usp[0]); + rp = VDSO32_SYMBOL(current, restart_syscall); + } WARN_ON(err); - /* flush data/instruction cache for new insns */ - flush_user_dcache_range_asm(start, end); - flush_user_icache_range_asm(start, end); - - regs->gr[31] = regs->gr[30] + 8; + regs->gr[31] = rp; + DBG(1, "%s: ERESTART_RESTARTBLOCK\n", __func__); return; } + case -EINTR: + /* ok, was handled before and should be returned. */ + break; case -ERESTARTNOHAND: case -ERESTARTSYS: case -ERESTARTNOINTR: + DBG(1, "%s: Type %ld\n", __func__, regs->gr[28]); check_syscallno_in_delay_branch(regs); return; default: @@ -567,30 +545,35 @@ insert_restart_trampoline(struct pt_regs *regs) * registers). As noted below, the syscall number gets restored for * us due to the magic of delayed branching. */ -asmlinkage void -do_signal(struct pt_regs *regs, long in_syscall) +static void do_signal(struct pt_regs *regs, long in_syscall) { struct ksignal ksig; + int restart_syscall; + bool has_handler; - DBG(1,"\ndo_signal: regs=0x%p, sr7 %#lx, in_syscall=%d\n", - regs, regs->sr[7], in_syscall); + has_handler = get_signal(&ksig); - if (get_signal(&ksig)) { - DBG(3,"do_signal: signr = %d, regs->gr[28] = %ld\n", signr, regs->gr[28]); + restart_syscall = 0; + if (in_syscall) + restart_syscall = 1; + + if (has_handler) { /* Restart a system call if necessary. */ - if (in_syscall) + if (restart_syscall) syscall_restart(regs, &ksig.ka); handle_signal(&ksig, regs, in_syscall); + DBG(1, "%s: Handled signal pid %d\n", + __func__, task_pid_nr(current)); return; } - /* Did we come from a system call? */ - if (in_syscall) + /* Do we need to restart the system call? */ + if (restart_syscall) insert_restart_trampoline(regs); - DBG(1,"do_signal: Exit (not delivered), regs->gr[28] = %ld\n", - regs->gr[28]); + DBG(1, "%s: Exit (not delivered), regs->gr[28] = %ld orig_r28 = %ld pid %d\n", + __func__, regs->gr[28], regs->orig_r28, task_pid_nr(current)); restore_saved_sigmask(); } @@ -602,5 +585,5 @@ void do_notify_resume(struct pt_regs *regs, long in_syscall) do_signal(regs, in_syscall); if (test_thread_flag(TIF_NOTIFY_RESUME)) - tracehook_notify_resume(regs); + resume_user_mode_work(regs); } diff --git a/arch/parisc/kernel/signal32.h b/arch/parisc/kernel/signal32.h index f166250f2d06..c03eb1ed4c53 100644 --- a/arch/parisc/kernel/signal32.h +++ b/arch/parisc/kernel/signal32.h @@ -36,21 +36,12 @@ struct compat_regfile { compat_int_t rf_sar; }; -#define COMPAT_SIGRETURN_TRAMP 4 -#define COMPAT_SIGRESTARTBLOCK_TRAMP 5 -#define COMPAT_TRAMP_SIZE (COMPAT_SIGRETURN_TRAMP + \ - COMPAT_SIGRESTARTBLOCK_TRAMP) - struct compat_rt_sigframe { - /* XXX: Must match trampoline size in arch/parisc/kernel/signal.c - Secondary to that it must protect the ERESTART_RESTARTBLOCK - trampoline we left on the stack (we were bad and didn't - change sp so we could run really fast.) */ - compat_uint_t tramp[COMPAT_TRAMP_SIZE]; - compat_siginfo_t info; - struct compat_ucontext uc; - /* Hidden location of truncated registers, *must* be last. */ - struct compat_regfile regs; + unsigned int tramp[2]; /* holds original return address */ + compat_siginfo_t info; + struct compat_ucontext uc; + /* Hidden location of truncated registers, *must* be last. */ + struct compat_regfile regs; }; /* diff --git a/arch/parisc/kernel/smp.c b/arch/parisc/kernel/smp.c index a32a882a2d58..7dbd92cafae3 100644 --- a/arch/parisc/kernel/smp.c +++ b/arch/parisc/kernel/smp.c @@ -30,6 +30,7 @@ #include <linux/ftrace.h> #include <linux/cpu.h> #include <linux/kgdb.h> +#include <linux/sched/hotplug.h> #include <linux/atomic.h> #include <asm/current.h> @@ -60,8 +61,6 @@ volatile struct task_struct *smp_init_current_idle_task; /* track which CPU is booting */ static volatile int cpu_now_booting; -static int parisc_max_cpus = 1; - static DEFINE_PER_CPU(spinlock_t, ipi_lock); enum ipi_message_type { @@ -269,7 +268,7 @@ void arch_send_call_function_single_ipi(int cpu) /* * Called by secondaries to update state and initialize CPU registers. */ -static void __init +static void smp_cpu_init(int cpunum) { extern void init_IRQ(void); /* arch/parisc/kernel/irq.c */ @@ -309,7 +308,7 @@ smp_cpu_init(int cpunum) * Slaves start using C here. Indirectly called from smp_slave_stext. * Do what start_kernel() and main() do for boot strap processor (aka monarch) */ -void __init smp_callin(unsigned long pdce_proc) +void smp_callin(unsigned long pdce_proc) { int slave_id = cpu_now_booting; @@ -334,11 +333,28 @@ void __init smp_callin(unsigned long pdce_proc) /* * Bring one cpu online. */ -int smp_boot_one_cpu(int cpuid, struct task_struct *idle) +static int smp_boot_one_cpu(int cpuid, struct task_struct *idle) { const struct cpuinfo_parisc *p = &per_cpu(cpu_data, cpuid); long timeout; +#ifdef CONFIG_HOTPLUG_CPU + int i; + + /* reset irq statistics for this CPU */ + memset(&per_cpu(irq_stat, cpuid), 0, sizeof(irq_cpustat_t)); + for (i = 0; i < NR_IRQS; i++) { + struct irq_desc *desc = irq_to_desc(i); + + if (desc && desc->kstat_irqs) + *per_cpu_ptr(desc->kstat_irqs, cpuid) = 0; + } +#endif + + /* wait until last booting CPU has started. */ + while (cpu_now_booting) + ; + /* Let _start know what logical CPU we're booting ** (offset into init_tasks[],cpu_data[]) */ @@ -374,7 +390,6 @@ int smp_boot_one_cpu(int cpuid, struct task_struct *idle) if(cpu_online(cpuid)) { /* Which implies Slave has started up */ cpu_now_booting = 0; - smp_init_current_idle_task = NULL; goto alive ; } udelay(100); @@ -415,30 +430,86 @@ void __init smp_prepare_cpus(unsigned int max_cpus) spin_lock_init(&per_cpu(ipi_lock, cpu)); init_cpu_present(cpumask_of(0)); - - parisc_max_cpus = max_cpus; - if (!max_cpus) - printk(KERN_INFO "SMP mode deactivated.\n"); } -void smp_cpus_done(unsigned int cpu_max) +void __init smp_cpus_done(unsigned int cpu_max) { - return; } int __cpu_up(unsigned int cpu, struct task_struct *tidle) { - if (cpu != 0 && cpu < parisc_max_cpus && smp_boot_one_cpu(cpu, tidle)) - return -ENOSYS; + if (cpu_online(cpu)) + return 0; + + if (num_online_cpus() < setup_max_cpus && smp_boot_one_cpu(cpu, tidle)) + return -EIO; - return cpu_online(cpu) ? 0 : -ENOSYS; + return cpu_online(cpu) ? 0 : -EIO; } -#ifdef CONFIG_PROC_FS -int setup_profiling_timer(unsigned int multiplier) +/* + * __cpu_disable runs on the processor to be shutdown. + */ +int __cpu_disable(void) { - return -EINVAL; -} +#ifdef CONFIG_HOTPLUG_CPU + unsigned int cpu = smp_processor_id(); + + remove_cpu_topology(cpu); + + /* + * Take this CPU offline. Once we clear this, we can't return, + * and we must not schedule until we're ready to give up the cpu. + */ + set_cpu_online(cpu, false); + + /* Find a new timesync master */ + if (cpu == time_keeper_id) { + time_keeper_id = cpumask_first(cpu_online_mask); + pr_info("CPU %d is now promoted to time-keeper master\n", time_keeper_id); + } + + disable_percpu_irq(IPI_IRQ); + + irq_migrate_all_off_this_cpu(); + + flush_cache_all_local(); + flush_tlb_all_local(NULL); + + /* disable all irqs, including timer irq */ + local_irq_disable(); + + /* wait for next timer irq ... */ + mdelay(1000/HZ+100); + + /* ... and then clear all pending external irqs */ + set_eiem(0); + mtctl(~0UL, CR_EIRR); + mfctl(CR_EIRR); + mtctl(0, CR_EIRR); #endif + return 0; +} + +/* + * called on the thread which is asking for a CPU to be shutdown - + * waits until shutdown has completed, or it is timed out. + */ +void __cpu_die(unsigned int cpu) +{ + pdc_cpu_rendezvous_lock(); + + if (!cpu_wait_death(cpu, 5)) { + pr_crit("CPU%u: cpu didn't die\n", cpu); + return; + } + pr_info("CPU%u: is shutting down\n", cpu); + + /* set task's state to interruptible sleep */ + set_current_state(TASK_INTERRUPTIBLE); + schedule_timeout((IS_ENABLED(CONFIG_64BIT) ? 8:2) * HZ); + + pdc_cpu_rendezvous_unlock(); +} diff --git a/arch/parisc/kernel/sys_parisc.c b/arch/parisc/kernel/sys_parisc.c index 2b34294517a1..848b0702005d 100644 --- a/arch/parisc/kernel/sys_parisc.c +++ b/arch/parisc/kernel/sys_parisc.c @@ -239,14 +239,14 @@ static unsigned long mmap_rnd(void) unsigned long rnd = 0; if (current->flags & PF_RANDOMIZE) - rnd = get_random_int() & MMAP_RND_MASK; + rnd = get_random_u32() & MMAP_RND_MASK; return rnd << PAGE_SHIFT; } unsigned long arch_mmap_rnd(void) { - return (get_random_int() & MMAP_RND_MASK) << PAGE_SHIFT; + return (get_random_u32() & MMAP_RND_MASK) << PAGE_SHIFT; } static unsigned long mmap_legacy_base(void) diff --git a/arch/parisc/kernel/syscall.S b/arch/parisc/kernel/syscall.S index d2497b339d13..1373e5129868 100644 --- a/arch/parisc/kernel/syscall.S +++ b/arch/parisc/kernel/syscall.S @@ -50,6 +50,22 @@ registers). .level PA_ASM_LEVEL + .macro lws_pagefault_disable reg1,reg2 + mfctl %cr30, \reg2 + ldo TASK_PAGEFAULT_DISABLED(\reg2), \reg2 + ldw 0(%sr2,\reg2), \reg1 + ldo 1(\reg1), \reg1 + stw \reg1, 0(%sr2,\reg2) + .endm + + .macro lws_pagefault_enable reg1,reg2 + mfctl %cr30, \reg2 + ldo TASK_PAGEFAULT_DISABLED(\reg2), \reg2 + ldw 0(%sr2,\reg2), \reg1 + ldo -1(\reg1), \reg1 + stw \reg1, 0(%sr2,\reg2) + .endm + .text .import syscall_exit,code @@ -74,7 +90,7 @@ ENTRY(linux_gateway_page) /* ADDRESS 0xb0 to 0xb8, lws uses two insns for entry */ /* Light-weight-syscall entry must always be located at 0xb0 */ /* WARNING: Keep this number updated with table size changes */ -#define __NR_lws_entries (3) +#define __NR_lws_entries (5) lws_entry: gate lws_start, %r0 /* increase privilege */ @@ -472,7 +488,7 @@ lws_start: extrd,u %r1,PSW_W_BIT,1,%r1 /* sp must be aligned on 4, so deposit the W bit setting into * the bottom of sp temporarily */ - or,ev %r1,%r30,%r30 + or,od %r1,%r30,%r30 /* Clip LWS number to a 32-bit value for 32-bit processes */ depdi 0, 31, 32, %r20 @@ -490,8 +506,34 @@ lws_start: /* Jump to lws, lws table pointers already relocated */ be,n 0(%sr2,%r21) +lws_exit_noerror: + lws_pagefault_enable %r1,%r21 + stw,ma %r20, 0(%sr2,%r20) + ssm PSW_SM_I, %r0 + b lws_exit + copy %r0, %r21 + +lws_wouldblock: + ssm PSW_SM_I, %r0 + ldo 2(%r0), %r28 + b lws_exit + ldo -EAGAIN(%r0), %r21 + +lws_pagefault: + lws_pagefault_enable %r1,%r21 + stw,ma %r20, 0(%sr2,%r20) + ssm PSW_SM_I, %r0 + ldo 3(%r0),%r28 + b lws_exit + ldo -EAGAIN(%r0),%r21 + +lws_fault: + ldo 1(%r0),%r28 + b lws_exit + ldo -EFAULT(%r0),%r21 + lws_exit_nosys: - ldo -ENOSYS(%r0),%r21 /* set errno */ + ldo -ENOSYS(%r0),%r21 /* Fall through: Return to userspace */ lws_exit: @@ -518,27 +560,19 @@ lws_exit: %r28 - Return prev through this register. %r21 - Kernel error code - If debugging is DISabled: - - %r21 has the following meanings: - + %r21 returns the following error codes: EAGAIN - CAS is busy, ldcw failed, try again. EFAULT - Read or write failed. - If debugging is enabled: - - EDEADLOCK - CAS called recursively. - EAGAIN && r28 == 1 - CAS is busy. Lock contended. - EAGAIN && r28 == 2 - CAS is busy. ldcw failed. - EFAULT - Read or write failed. + If EAGAIN is returned, %r28 indicates the busy reason: + r28 == 1 - CAS is busy. lock contended. + r28 == 2 - CAS is busy. ldcw failed. + r28 == 3 - CAS is busy. page fault. Scratch: r20, r28, r1 ****************************************************/ - /* Do not enable LWS debugging */ -#define ENABLE_LWS_DEBUG 0 - /* ELF64 Process entry path */ lws_compare_and_swap64: #ifdef CONFIG_64BIT @@ -551,59 +585,45 @@ lws_compare_and_swap64: b,n lws_exit_nosys #endif - /* ELF32 Process entry path */ + /* ELF32/ELF64 Process entry path */ lws_compare_and_swap32: #ifdef CONFIG_64BIT - /* Clip all the input registers */ + /* Wide mode user process? */ + bb,<,n %sp, 31, lws_compare_and_swap + + /* Clip all the input registers for 32-bit processes */ depdi 0, 31, 32, %r26 depdi 0, 31, 32, %r25 depdi 0, 31, 32, %r24 #endif lws_compare_and_swap: - /* Load start of lock table */ - ldil L%lws_lock_start, %r20 - ldo R%lws_lock_start(%r20), %r28 + /* Trigger memory reference interruptions without writing to memory */ +1: ldw 0(%r26), %r28 +2: stbys,e %r0, 0(%r26) + + /* Calculate 8-bit hash index from virtual address */ + extru_safe %r26, 27, 8, %r20 - /* Extract eight bits from r26 and hash lock (Bits 3-11) */ - extru_safe %r26, 28, 8, %r20 + /* Load start of lock table */ + ldil L%lws_lock_start, %r28 + ldo R%lws_lock_start(%r28), %r28 - /* Find lock to use, the hash is either one of 0 to - 15, multiplied by 16 (keep it 16-byte aligned) + /* Find lock to use, the hash index is one of 0 to + 255, multiplied by 16 (keep it 16-byte aligned) and add to the lock table offset. */ shlw %r20, 4, %r20 add %r20, %r28, %r20 -# if ENABLE_LWS_DEBUG - /* - DEBUG, check for deadlock! - If the thread register values are the same - then we were the one that locked it last and - this is a recurisve call that will deadlock. - We *must* giveup this call and fail. - */ - ldw 4(%sr2,%r20), %r28 /* Load thread register */ - /* WARNING: If cr27 cycles to the same value we have problems */ - mfctl %cr27, %r21 /* Get current thread register */ - cmpb,<>,n %r21, %r28, cas_lock /* Called recursive? */ - b lws_exit /* Return error! */ - ldo -EDEADLOCK(%r0), %r21 -cas_lock: - cmpb,=,n %r0, %r28, cas_nocontend /* Is nobody using it? */ - ldo 1(%r0), %r28 /* 1st case */ - b lws_exit /* Contended... */ - ldo -EAGAIN(%r0), %r21 /* Spin in userspace */ -cas_nocontend: -# endif -/* ENABLE_LWS_DEBUG */ - - /* COW breaks can cause contention on UP systems */ - LDCW 0(%sr2,%r20), %r28 /* Try to acquire the lock */ - cmpb,<>,n %r0, %r28, cas_action /* Did we get it? */ -cas_wouldblock: - ldo 2(%r0), %r28 /* 2nd case */ - b lws_exit /* Contended... */ - ldo -EAGAIN(%r0), %r21 /* Spin in userspace */ + rsm PSW_SM_I, %r0 /* Disable interrupts */ + + /* Try to acquire the lock */ + LDCW 0(%sr2,%r20), %r28 + comclr,<> %r0, %r28, %r0 + b,n lws_wouldblock + + /* Disable page faults to prevent sleeping in critical region */ + lws_pagefault_disable %r21,%r28 /* prev = *addr; @@ -613,59 +633,35 @@ cas_wouldblock: */ /* NOTES: - This all works becuse intr_do_signal + This all works because intr_do_signal and schedule both check the return iasq and see that we are on the kernel page so this process is never scheduled off or is ever sent any signal of any sort, - thus it is wholly atomic from usrspaces + thus it is wholly atomic from usrspace's perspective */ -cas_action: -#if defined CONFIG_SMP && ENABLE_LWS_DEBUG - /* DEBUG */ - mfctl %cr27, %r1 - stw %r1, 4(%sr2,%r20) -#endif /* The load and store could fail */ -1: ldw 0(%r26), %r28 +3: ldw 0(%r26), %r28 sub,<> %r28, %r25, %r0 -2: stw %r24, 0(%r26) - /* Free lock */ - stw,ma %r20, 0(%sr2,%r20) -#if ENABLE_LWS_DEBUG - /* Clear thread register indicator */ - stw %r0, 4(%sr2,%r20) -#endif - /* Return to userspace, set no error */ - b lws_exit - copy %r0, %r21 +4: stw %r24, 0(%r26) + b,n lws_exit_noerror -3: - /* Error occurred on load or store */ - /* Free lock */ - stw,ma %r20, 0(%sr2,%r20) -#if ENABLE_LWS_DEBUG - stw %r0, 4(%sr2,%r20) -#endif - b lws_exit - ldo -EFAULT(%r0),%r21 /* set errno */ - nop - nop - nop - nop + /* A fault occurred on load or stbys,e store */ +5: b,n lws_fault + ASM_EXCEPTIONTABLE_ENTRY(1b-linux_gateway_page, 5b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(2b-linux_gateway_page, 5b-linux_gateway_page) - /* Two exception table entries, one for the load, - the other for the store. Either return -EFAULT. - Each of the entries must be relocated. */ - ASM_EXCEPTIONTABLE_ENTRY(1b-linux_gateway_page, 3b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(2b-linux_gateway_page, 3b-linux_gateway_page) + /* A page fault occurred in critical region */ +6: b,n lws_pagefault + ASM_EXCEPTIONTABLE_ENTRY(3b-linux_gateway_page, 6b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(4b-linux_gateway_page, 6b-linux_gateway_page) /*************************************************** New CAS implementation which uses pointers and variable size information. The value pointed by old and new MUST NOT change - while performing CAS. The lock only protect the value at %r26. + while performing CAS. The lock only protects the value at %r26. %r26 - Address to examine %r25 - Pointer to the value to check (old) @@ -674,25 +670,32 @@ cas_action: %r28 - Return non-zero on failure %r21 - Kernel error code - %r21 has the following meanings: - + %r21 returns the following error codes: EAGAIN - CAS is busy, ldcw failed, try again. EFAULT - Read or write failed. + If EAGAIN is returned, %r28 indicates the busy reason: + r28 == 1 - CAS is busy. lock contended. + r28 == 2 - CAS is busy. ldcw failed. + r28 == 3 - CAS is busy. page fault. + Scratch: r20, r22, r28, r29, r1, fr4 (32bit for 64bit CAS only) ****************************************************/ - /* ELF32 Process entry path */ lws_compare_and_swap_2: #ifdef CONFIG_64BIT - /* Clip the input registers. We don't need to clip %r23 as we - only use it for word operations */ + /* Wide mode user process? */ + bb,<,n %sp, 31, cas2_begin + + /* Clip the input registers for 32-bit processes. We don't + need to clip %r23 as we only use it for word operations */ depdi 0, 31, 32, %r26 depdi 0, 31, 32, %r25 depdi 0, 31, 32, %r24 #endif +cas2_begin: /* Check the validity of the size pointer */ subi,>>= 3, %r23, %r0 b,n lws_exit_nosys @@ -703,69 +706,77 @@ lws_compare_and_swap_2: blr %r29, %r0 nop - /* 8bit load */ -4: ldb 0(%r25), %r25 + /* 8-bit load */ +1: ldb 0(%r25), %r25 b cas2_lock_start -5: ldb 0(%r24), %r24 +2: ldb 0(%r24), %r24 nop nop nop nop nop - /* 16bit load */ -6: ldh 0(%r25), %r25 + /* 16-bit load */ +3: ldh 0(%r25), %r25 b cas2_lock_start -7: ldh 0(%r24), %r24 +4: ldh 0(%r24), %r24 nop nop nop nop nop - /* 32bit load */ -8: ldw 0(%r25), %r25 + /* 32-bit load */ +5: ldw 0(%r25), %r25 b cas2_lock_start -9: ldw 0(%r24), %r24 +6: ldw 0(%r24), %r24 nop nop nop nop nop - /* 64bit load */ + /* 64-bit load */ #ifdef CONFIG_64BIT -10: ldd 0(%r25), %r25 -11: ldd 0(%r24), %r24 +7: ldd 0(%r25), %r25 +8: ldd 0(%r24), %r24 #else /* Load old value into r22/r23 - high/low */ -10: ldw 0(%r25), %r22 -11: ldw 4(%r25), %r23 +7: ldw 0(%r25), %r22 +8: ldw 4(%r25), %r23 /* Load new value into fr4 for atomic store later */ -12: flddx 0(%r24), %fr4 +9: flddx 0(%r24), %fr4 #endif cas2_lock_start: - /* Load start of lock table */ - ldil L%lws_lock_start, %r20 - ldo R%lws_lock_start(%r20), %r28 + /* Trigger memory reference interruptions without writing to memory */ + copy %r26, %r28 + depi_safe 0, 31, 2, %r28 +10: ldw 0(%r28), %r1 +11: stbys,e %r0, 0(%r28) - /* Extract eight bits from r26 and hash lock (Bits 3-11) */ - extru_safe %r26, 28, 8, %r20 + /* Calculate 8-bit hash index from virtual address */ + extru_safe %r26, 27, 8, %r20 + + /* Load start of lock table */ + ldil L%lws_lock_start, %r28 + ldo R%lws_lock_start(%r28), %r28 - /* Find lock to use, the hash is either one of 0 to - 15, multiplied by 16 (keep it 16-byte aligned) + /* Find lock to use, the hash index is one of 0 to + 255, multiplied by 16 (keep it 16-byte aligned) and add to the lock table offset. */ shlw %r20, 4, %r20 add %r20, %r28, %r20 - /* COW breaks can cause contention on UP systems */ - LDCW 0(%sr2,%r20), %r28 /* Try to acquire the lock */ - cmpb,<>,n %r0, %r28, cas2_action /* Did we get it? */ -cas2_wouldblock: - ldo 2(%r0), %r28 /* 2nd case */ - b lws_exit /* Contended... */ - ldo -EAGAIN(%r0), %r21 /* Spin in userspace */ + rsm PSW_SM_I, %r0 /* Disable interrupts */ + + /* Try to acquire the lock */ + LDCW 0(%sr2,%r20), %r28 + comclr,<> %r0, %r28, %r0 + b,n lws_wouldblock + + /* Disable page faults to prevent sleeping in critical region */ + lws_pagefault_disable %r21,%r28 /* prev = *addr; @@ -775,112 +786,493 @@ cas2_wouldblock: */ /* NOTES: - This all works becuse intr_do_signal + This all works because intr_do_signal and schedule both check the return iasq and see that we are on the kernel page so this process is never scheduled off or is ever sent any signal of any sort, - thus it is wholly atomic from usrspaces + thus it is wholly atomic from usrspace's perspective */ -cas2_action: + /* Jump to the correct function */ blr %r29, %r0 /* Set %r28 as non-zero for now */ ldo 1(%r0),%r28 - /* 8bit CAS */ -13: ldb 0(%r26), %r29 + /* 8-bit CAS */ +12: ldb 0(%r26), %r29 sub,= %r29, %r25, %r0 - b,n cas2_end -14: stb %r24, 0(%r26) - b cas2_end + b,n lws_exit_noerror +13: stb %r24, 0(%r26) + b lws_exit_noerror copy %r0, %r28 nop nop - /* 16bit CAS */ -15: ldh 0(%r26), %r29 + /* 16-bit CAS */ +14: ldh 0(%r26), %r29 sub,= %r29, %r25, %r0 - b,n cas2_end -16: sth %r24, 0(%r26) - b cas2_end + b,n lws_exit_noerror +15: sth %r24, 0(%r26) + b lws_exit_noerror copy %r0, %r28 nop nop - /* 32bit CAS */ -17: ldw 0(%r26), %r29 + /* 32-bit CAS */ +16: ldw 0(%r26), %r29 sub,= %r29, %r25, %r0 - b,n cas2_end -18: stw %r24, 0(%r26) - b cas2_end + b,n lws_exit_noerror +17: stw %r24, 0(%r26) + b lws_exit_noerror copy %r0, %r28 nop nop - /* 64bit CAS */ + /* 64-bit CAS */ #ifdef CONFIG_64BIT -19: ldd 0(%r26), %r29 +18: ldd 0(%r26), %r29 sub,*= %r29, %r25, %r0 - b,n cas2_end -20: std %r24, 0(%r26) + b,n lws_exit_noerror +19: std %r24, 0(%r26) copy %r0, %r28 #else /* Compare first word */ -19: ldw 0(%r26), %r29 +18: ldw 0(%r26), %r29 sub,= %r29, %r22, %r0 - b,n cas2_end + b,n lws_exit_noerror /* Compare second word */ -20: ldw 4(%r26), %r29 +19: ldw 4(%r26), %r29 sub,= %r29, %r23, %r0 - b,n cas2_end + b,n lws_exit_noerror /* Perform the store */ -21: fstdx %fr4, 0(%r26) +20: fstdx %fr4, 0(%r26) copy %r0, %r28 #endif + b lws_exit_noerror + copy %r0, %r28 -cas2_end: - /* Free lock */ - stw,ma %r20, 0(%sr2,%r20) - /* Return to userspace, set no error */ - b lws_exit - copy %r0, %r21 + /* A fault occurred on load or stbys,e store */ +30: b,n lws_fault + ASM_EXCEPTIONTABLE_ENTRY(1b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(2b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(3b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(4b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(5b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(6b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(7b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(8b-linux_gateway_page, 30b-linux_gateway_page) +#ifndef CONFIG_64BIT + ASM_EXCEPTIONTABLE_ENTRY(9b-linux_gateway_page, 30b-linux_gateway_page) +#endif -22: - /* Error occurred on load or store */ - /* Free lock */ - stw,ma %r20, 0(%sr2,%r20) + ASM_EXCEPTIONTABLE_ENTRY(10b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(11b-linux_gateway_page, 30b-linux_gateway_page) + + /* A page fault occurred in critical region */ +31: b,n lws_pagefault + ASM_EXCEPTIONTABLE_ENTRY(12b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(13b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(14b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(15b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(16b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(17b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(18b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(19b-linux_gateway_page, 31b-linux_gateway_page) +#ifndef CONFIG_64BIT + ASM_EXCEPTIONTABLE_ENTRY(20b-linux_gateway_page, 31b-linux_gateway_page) +#endif + + + /*************************************************** + LWS atomic exchange. + + %r26 - Exchange address + %r25 - Size of the variable (0/1/2/3 for 8/16/32/64 bit) + %r24 - Address of new value + %r23 - Address of old value + %r28 - Return non-zero on failure + %r21 - Kernel error code + + %r21 returns the following error codes: + EAGAIN - CAS is busy, ldcw failed, try again. + EFAULT - Read or write failed. + + If EAGAIN is returned, %r28 indicates the busy reason: + r28 == 1 - CAS is busy. lock contended. + r28 == 2 - CAS is busy. ldcw failed. + r28 == 3 - CAS is busy. page fault. + + Scratch: r20, r1 + + ****************************************************/ + +lws_atomic_xchg: +#ifdef CONFIG_64BIT + /* Wide mode user process? */ + bb,<,n %sp, 31, atomic_xchg_begin + + /* Clip the input registers for 32-bit processes. We don't + need to clip %r23 as we only use it for word operations */ + depdi 0, 31, 32, %r26 + depdi 0, 31, 32, %r25 + depdi 0, 31, 32, %r24 + depdi 0, 31, 32, %r23 +#endif + +atomic_xchg_begin: + /* Check the validity of the size pointer */ + subi,>>= 3, %r25, %r0 + b,n lws_exit_nosys + + /* Jump to the functions which will load the old and new values into + registers depending on the their size */ + shlw %r25, 2, %r1 + blr %r1, %r0 + nop + + /* Perform exception checks */ + + /* 8-bit exchange */ +1: ldb 0(%r24), %r20 + copy %r23, %r20 + depi_safe 0, 31, 2, %r20 + b atomic_xchg_start +2: stbys,e %r0, 0(%r20) + nop + nop + nop + + /* 16-bit exchange */ +3: ldh 0(%r24), %r20 + copy %r23, %r20 + depi_safe 0, 31, 2, %r20 + b atomic_xchg_start +4: stbys,e %r0, 0(%r20) + nop + nop + nop + + /* 32-bit exchange */ +5: ldw 0(%r24), %r20 + b atomic_xchg_start +6: stbys,e %r0, 0(%r23) + nop + nop + nop + nop + nop + + /* 64-bit exchange */ +#ifdef CONFIG_64BIT +7: ldd 0(%r24), %r20 +8: stdby,e %r0, 0(%r23) +#else +7: ldw 0(%r24), %r20 +8: ldw 4(%r24), %r20 + copy %r23, %r20 + depi_safe 0, 31, 2, %r20 +9: stbys,e %r0, 0(%r20) +10: stbys,e %r0, 4(%r20) +#endif + +atomic_xchg_start: + /* Trigger memory reference interruptions without writing to memory */ + copy %r26, %r28 + depi_safe 0, 31, 2, %r28 +11: ldw 0(%r28), %r1 +12: stbys,e %r0, 0(%r28) + + /* Calculate 8-bit hash index from virtual address */ + extru_safe %r26, 27, 8, %r20 + + /* Load start of lock table */ + ldil L%lws_lock_start, %r28 + ldo R%lws_lock_start(%r28), %r28 + + /* Find lock to use, the hash index is one of 0 to + 255, multiplied by 16 (keep it 16-byte aligned) + and add to the lock table offset. */ + shlw %r20, 4, %r20 + add %r20, %r28, %r20 + + rsm PSW_SM_I, %r0 /* Disable interrupts */ + + /* Try to acquire the lock */ + LDCW 0(%sr2,%r20), %r28 + comclr,<> %r0, %r28, %r0 + b,n lws_wouldblock + + /* Disable page faults to prevent sleeping in critical region */ + lws_pagefault_disable %r21,%r28 + + /* NOTES: + This all works because intr_do_signal + and schedule both check the return iasq + and see that we are on the kernel page + so this process is never scheduled off + or is ever sent any signal of any sort, + thus it is wholly atomic from userspace's + perspective + */ + + /* Jump to the correct function */ + blr %r1, %r0 + /* Set %r28 as non-zero for now */ ldo 1(%r0),%r28 - b lws_exit - ldo -EFAULT(%r0),%r21 /* set errno */ - nop - nop - nop - - /* Exception table entries, for the load and store, return EFAULT. - Each of the entries must be relocated. */ - ASM_EXCEPTIONTABLE_ENTRY(4b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(5b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(6b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(7b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(8b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(9b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(10b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(11b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(13b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(14b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(15b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(16b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(17b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(18b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(19b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(20b-linux_gateway_page, 22b-linux_gateway_page) + + /* 8-bit exchange */ +14: ldb 0(%r26), %r1 +15: stb %r1, 0(%r23) +15: ldb 0(%r24), %r1 +17: stb %r1, 0(%r26) + b lws_exit_noerror + copy %r0, %r28 + nop + nop + + /* 16-bit exchange */ +18: ldh 0(%r26), %r1 +19: sth %r1, 0(%r23) +20: ldh 0(%r24), %r1 +21: sth %r1, 0(%r26) + b lws_exit_noerror + copy %r0, %r28 + nop + nop + + /* 32-bit exchange */ +22: ldw 0(%r26), %r1 +23: stw %r1, 0(%r23) +24: ldw 0(%r24), %r1 +25: stw %r1, 0(%r26) + b lws_exit_noerror + copy %r0, %r28 + nop + nop + + /* 64-bit exchange */ +#ifdef CONFIG_64BIT +26: ldd 0(%r26), %r1 +27: std %r1, 0(%r23) +28: ldd 0(%r24), %r1 +29: std %r1, 0(%r26) +#else +26: flddx 0(%r26), %fr4 +27: fstdx %fr4, 0(%r23) +28: flddx 0(%r24), %fr4 +29: fstdx %fr4, 0(%r26) +#endif + b lws_exit_noerror + copy %r0, %r28 + + /* A fault occurred on load or stbys,e store */ +30: b,n lws_fault + ASM_EXCEPTIONTABLE_ENTRY(1b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(2b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(3b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(4b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(5b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(6b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(7b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(8b-linux_gateway_page, 30b-linux_gateway_page) +#ifndef CONFIG_64BIT + ASM_EXCEPTIONTABLE_ENTRY(9b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(10b-linux_gateway_page, 30b-linux_gateway_page) +#endif + + ASM_EXCEPTIONTABLE_ENTRY(11b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(12b-linux_gateway_page, 30b-linux_gateway_page) + + /* A page fault occurred in critical region */ +31: b,n lws_pagefault + ASM_EXCEPTIONTABLE_ENTRY(14b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(15b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(16b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(17b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(18b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(19b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(20b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(21b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(22b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(23b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(24b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(25b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(26b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(27b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(28b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(29b-linux_gateway_page, 31b-linux_gateway_page) + + /*************************************************** + LWS atomic store. + + %r26 - Address to store + %r25 - Size of the variable (0/1/2/3 for 8/16/32/64 bit) + %r24 - Address of value to store + %r28 - Return non-zero on failure + %r21 - Kernel error code + + %r21 returns the following error codes: + EAGAIN - CAS is busy, ldcw failed, try again. + EFAULT - Read or write failed. + + If EAGAIN is returned, %r28 indicates the busy reason: + r28 == 1 - CAS is busy. lock contended. + r28 == 2 - CAS is busy. ldcw failed. + r28 == 3 - CAS is busy. page fault. + + Scratch: r20, r1 + + ****************************************************/ + +lws_atomic_store: +#ifdef CONFIG_64BIT + /* Wide mode user process? */ + bb,<,n %sp, 31, atomic_store_begin + + /* Clip the input registers for 32-bit processes. We don't + need to clip %r23 as we only use it for word operations */ + depdi 0, 31, 32, %r26 + depdi 0, 31, 32, %r25 + depdi 0, 31, 32, %r24 +#endif + +atomic_store_begin: + /* Check the validity of the size pointer */ + subi,>>= 3, %r25, %r0 + b,n lws_exit_nosys + + shlw %r25, 1, %r1 + blr %r1, %r0 + nop + + /* Perform exception checks */ + + /* 8-bit store */ +1: ldb 0(%r24), %r20 + b,n atomic_store_start + nop + nop + + /* 16-bit store */ +2: ldh 0(%r24), %r20 + b,n atomic_store_start + nop + nop + + /* 32-bit store */ +3: ldw 0(%r24), %r20 + b,n atomic_store_start + nop + nop + + /* 64-bit store */ +#ifdef CONFIG_64BIT +4: ldd 0(%r24), %r20 +#else +4: ldw 0(%r24), %r20 +5: ldw 4(%r24), %r20 +#endif + +atomic_store_start: + /* Trigger memory reference interruptions without writing to memory */ + copy %r26, %r28 + depi_safe 0, 31, 2, %r28 +6: ldw 0(%r28), %r1 +7: stbys,e %r0, 0(%r28) + + /* Calculate 8-bit hash index from virtual address */ + extru_safe %r26, 27, 8, %r20 + + /* Load start of lock table */ + ldil L%lws_lock_start, %r28 + ldo R%lws_lock_start(%r28), %r28 + + /* Find lock to use, the hash index is one of 0 to + 255, multiplied by 16 (keep it 16-byte aligned) + and add to the lock table offset. */ + shlw %r20, 4, %r20 + add %r20, %r28, %r20 + + rsm PSW_SM_I, %r0 /* Disable interrupts */ + + /* Try to acquire the lock */ + LDCW 0(%sr2,%r20), %r28 + comclr,<> %r0, %r28, %r0 + b,n lws_wouldblock + + /* Disable page faults to prevent sleeping in critical region */ + lws_pagefault_disable %r21,%r28 + + /* NOTES: + This all works because intr_do_signal + and schedule both check the return iasq + and see that we are on the kernel page + so this process is never scheduled off + or is ever sent any signal of any sort, + thus it is wholly atomic from userspace's + perspective + */ + + /* Jump to the correct function */ + blr %r1, %r0 + /* Set %r28 as non-zero for now */ + ldo 1(%r0),%r28 + + /* 8-bit store */ +9: ldb 0(%r24), %r1 +10: stb %r1, 0(%r26) + b lws_exit_noerror + copy %r0, %r28 + + /* 16-bit store */ +11: ldh 0(%r24), %r1 +12: sth %r1, 0(%r26) + b lws_exit_noerror + copy %r0, %r28 + + /* 32-bit store */ +13: ldw 0(%r24), %r1 +14: stw %r1, 0(%r26) + b lws_exit_noerror + copy %r0, %r28 + + /* 64-bit store */ +#ifdef CONFIG_64BIT +15: ldd 0(%r24), %r1 +16: std %r1, 0(%r26) +#else +15: flddx 0(%r24), %fr4 +16: fstdx %fr4, 0(%r26) +#endif + b lws_exit_noerror + copy %r0, %r28 + + /* A fault occurred on load or stbys,e store */ +30: b,n lws_fault + ASM_EXCEPTIONTABLE_ENTRY(1b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(2b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(3b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(4b-linux_gateway_page, 30b-linux_gateway_page) #ifndef CONFIG_64BIT - ASM_EXCEPTIONTABLE_ENTRY(12b-linux_gateway_page, 22b-linux_gateway_page) - ASM_EXCEPTIONTABLE_ENTRY(21b-linux_gateway_page, 22b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(5b-linux_gateway_page, 30b-linux_gateway_page) #endif + ASM_EXCEPTIONTABLE_ENTRY(6b-linux_gateway_page, 30b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(7b-linux_gateway_page, 30b-linux_gateway_page) + + /* A page fault occurred in critical region */ +31: b,n lws_pagefault + ASM_EXCEPTIONTABLE_ENTRY(9b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(10b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(11b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(12b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(13b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(14b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(15b-linux_gateway_page, 31b-linux_gateway_page) + ASM_EXCEPTIONTABLE_ENTRY(16b-linux_gateway_page, 31b-linux_gateway_page) + /* Make sure nothing else is placed on this page */ .align PAGE_SIZE END(linux_gateway_page) @@ -899,7 +1291,9 @@ ENTRY(end_linux_gateway_page) ENTRY(lws_table) LWS_ENTRY(compare_and_swap32) /* 0 - ELF32 Atomic 32bit CAS */ LWS_ENTRY(compare_and_swap64) /* 1 - ELF64 Atomic 32bit CAS */ - LWS_ENTRY(compare_and_swap_2) /* 2 - ELF32 Atomic 64bit CAS */ + LWS_ENTRY(compare_and_swap_2) /* 2 - Atomic 64bit CAS */ + LWS_ENTRY(atomic_xchg) /* 3 - Atomic Exchange */ + LWS_ENTRY(atomic_store) /* 4 - Atomic Store */ END(lws_table) /* End of lws table */ diff --git a/arch/parisc/kernel/syscalls/Makefile b/arch/parisc/kernel/syscalls/Makefile index d63f18dd058d..8440c16dfb22 100644 --- a/arch/parisc/kernel/syscalls/Makefile +++ b/arch/parisc/kernel/syscalls/Makefile @@ -2,8 +2,7 @@ kapi := arch/$(SRCARCH)/include/generated/asm uapi := arch/$(SRCARCH)/include/generated/uapi/asm -_dummy := $(shell [ -d '$(uapi)' ] || mkdir -p '$(uapi)') \ - $(shell [ -d '$(kapi)' ] || mkdir -p '$(kapi)') +$(shell mkdir -p $(uapi) $(kapi)) syscall := $(src)/syscall.tbl syshdr := $(srctree)/scripts/syscallhdr.sh diff --git a/arch/parisc/kernel/syscalls/syscall.tbl b/arch/parisc/kernel/syscalls/syscall.tbl index 358c00000755..8a99c998da9b 100644 --- a/arch/parisc/kernel/syscalls/syscall.tbl +++ b/arch/parisc/kernel/syscalls/syscall.tbl @@ -413,7 +413,7 @@ 412 32 utimensat_time64 sys_utimensat sys_utimensat 413 32 pselect6_time64 sys_pselect6 compat_sys_pselect6_time64 414 32 ppoll_time64 sys_ppoll compat_sys_ppoll_time64 -416 32 io_pgetevents_time64 sys_io_pgetevents sys_io_pgetevents +416 32 io_pgetevents_time64 sys_io_pgetevents compat_sys_io_pgetevents_time64 417 32 recvmmsg_time64 sys_recvmmsg compat_sys_recvmmsg_time64 418 32 mq_timedsend_time64 sys_mq_timedsend sys_mq_timedsend 419 32 mq_timedreceive_time64 sys_mq_timedreceive sys_mq_timedreceive @@ -447,3 +447,4 @@ # 447 reserved for memfd_secret 448 common process_mrelease sys_process_mrelease 449 common futex_waitv sys_futex_waitv +450 common set_mempolicy_home_node sys_set_mempolicy_home_node diff --git a/arch/parisc/kernel/time.c b/arch/parisc/kernel/time.c index 061119a56fbe..9714fbd7c42d 100644 --- a/arch/parisc/kernel/time.c +++ b/arch/parisc/kernel/time.c @@ -40,6 +40,8 @@ #include <linux/timex.h> +int time_keeper_id __read_mostly; /* CPU used for timekeeping. */ + static unsigned long clocktick __ro_after_init; /* timer cycles per tick */ /* @@ -84,7 +86,7 @@ irqreturn_t __irq_entry timer_interrupt(int irq, void *dev_id) cpuinfo->it_value = next_tick; /* Go do system house keeping. */ - if (cpu != 0) + if (IS_ENABLED(CONFIG_SMP) && (cpu != time_keeper_id)) ticks_elapsed = 0; legacy_timer_tick(ticks_elapsed); @@ -150,7 +152,7 @@ static struct clocksource clocksource_cr16 = { .flags = CLOCK_SOURCE_IS_CONTINUOUS, }; -void __init start_cpu_itimer(void) +void start_cpu_itimer(void) { unsigned int cpu = smp_processor_id(); unsigned long next_tick = mfctl(16) + clocktick; @@ -249,13 +251,9 @@ void __init time_init(void) static int __init init_cr16_clocksource(void) { /* - * The cr16 interval timers are not syncronized across CPUs, even if - * they share the same socket. + * The cr16 interval timers are not synchronized across CPUs. */ if (num_online_cpus() > 1 && !running_on_qemu) { - /* mark sched_clock unstable */ - clear_sched_clock_stable(); - clocksource_cr16.name = "cr16_unstable"; clocksource_cr16.flags = CLOCK_SOURCE_UNSTABLE; clocksource_cr16.rating = 0; diff --git a/arch/parisc/kernel/toc.c b/arch/parisc/kernel/toc.c index 18327611cf8f..e4b48d07afbd 100644 --- a/arch/parisc/kernel/toc.c +++ b/arch/parisc/kernel/toc.c @@ -9,8 +9,11 @@ #include <asm/pdc.h> #include <asm/pdc_chassis.h> +#include <asm/ldcw.h> +#include <asm/processor.h> -unsigned int __aligned(16) toc_lock = 1; +static unsigned int __aligned(16) toc_lock = 1; +DEFINE_PER_CPU_PAGE_ALIGNED(char [16384], toc_stack) __visible; static void toc20_to_pt_regs(struct pt_regs *regs, struct pdc_toc_pim_20 *toc) { @@ -63,7 +66,8 @@ void notrace __noreturn __cold toc_intr(struct pt_regs *regs) struct pdc_toc_pim_20 pim_data20; struct pdc_toc_pim_11 pim_data11; - nmi_enter(); + /* verify we wrote regs to the correct stack */ + BUG_ON(regs != (struct pt_regs *)&per_cpu(toc_stack, raw_smp_processor_id())); if (boot_cpu_data.cpu_type >= pcxu) { if (pdc_pim_toc20(&pim_data20)) @@ -76,14 +80,25 @@ void notrace __noreturn __cold toc_intr(struct pt_regs *regs) } #ifdef CONFIG_KGDB + nmi_enter(); + if (atomic_read(&kgdb_active) != -1) kgdb_nmicallback(raw_smp_processor_id(), regs); kgdb_handle_exception(9, SIGTRAP, 0, regs); #endif + + /* serialize output, otherwise all CPUs write backtrace at once */ + while (__ldcw(&toc_lock) == 0) + ; /* wait */ show_regs(regs); + toc_lock = 1; /* release lock for next CPU */ + + if (raw_smp_processor_id() != 0) + while (1) ; /* all but monarch CPU will wait endless. */ /* give other CPUs time to show their backtrace */ mdelay(2000); + machine_restart("TOC"); /* should never reach this */ diff --git a/arch/parisc/kernel/toc_asm.S b/arch/parisc/kernel/toc_asm.S index e94ba8044190..570f5cef526e 100644 --- a/arch/parisc/kernel/toc_asm.S +++ b/arch/parisc/kernel/toc_asm.S @@ -5,34 +5,25 @@ .level 1.1 #include <asm/assembly.h> -#include <asm/psw.h> #include <linux/threads.h> #include <linux/linkage.h> .text .import toc_intr,code - .import toc_lock,data + .import toc_stack,data .align 16 ENTRY_CFI(toc_handler) - /* - * synchronize CPUs and obtain offset - * for stack setup. - */ - load32 PA(toc_lock),%r1 -0: ldcw,co 0(%r1),%r2 - cmpib,= 0,%r2,0b - nop - addi 1,%r2,%r4 - stw %r4,0(%r1) - addi -1,%r2,%r4 - load32 PA(toc_stack),%sp - /* - * deposit CPU number into stack address, - * so every CPU will have its own stack. - */ - SHLREG %r4,14,%r4 + +#ifdef CONFIG_SMP + /* get per-cpu toc_stack address. */ + mfctl %cr30, %r1 + tophys %r1,%r2 /* task_struct */ + LDREG TASK_TI_CPU(%r2),%r4 /* cpu */ + load32 PA(__per_cpu_offset),%r1 + LDREGX %r4(%r1),%r4 add %r4,%sp,%sp +#endif /* * setup pt_regs on stack and save the @@ -82,7 +73,3 @@ ENDPROC_CFI(toc_handler) */ SYM_DATA(toc_handler_csum, .long 0) SYM_DATA(toc_handler_size, .long . - toc_handler) - - __PAGE_ALIGNED_BSS - .align 64 -SYM_DATA(toc_stack, .block 16384*NR_CPUS) diff --git a/arch/parisc/kernel/topology.c b/arch/parisc/kernel/topology.c index 0a10e4ddc528..b9d845e314f8 100644 --- a/arch/parisc/kernel/topology.c +++ b/arch/parisc/kernel/topology.c @@ -13,56 +13,21 @@ #include <linux/percpu.h> #include <linux/sched.h> #include <linux/sched/topology.h> +#include <linux/cpu.h> #include <asm/topology.h> +#include <asm/sections.h> - /* - * cpu topology table - */ -struct cputopo_parisc cpu_topology[NR_CPUS] __read_mostly; -EXPORT_SYMBOL_GPL(cpu_topology); - -const struct cpumask *cpu_coregroup_mask(int cpu) -{ - return &cpu_topology[cpu].core_sibling; -} - -static void update_siblings_masks(unsigned int cpuid) -{ - struct cputopo_parisc *cpu_topo, *cpuid_topo = &cpu_topology[cpuid]; - int cpu; - - /* update core and thread sibling masks */ - for_each_possible_cpu(cpu) { - cpu_topo = &cpu_topology[cpu]; - - if (cpuid_topo->socket_id != cpu_topo->socket_id) - continue; - - cpumask_set_cpu(cpuid, &cpu_topo->core_sibling); - if (cpu != cpuid) - cpumask_set_cpu(cpu, &cpuid_topo->core_sibling); - - if (cpuid_topo->core_id != cpu_topo->core_id) - continue; - - cpumask_set_cpu(cpuid, &cpu_topo->thread_sibling); - if (cpu != cpuid) - cpumask_set_cpu(cpu, &cpuid_topo->thread_sibling); - } - smp_wmb(); -} - -static int dualcores_found __initdata; +static DEFINE_PER_CPU(struct cpu, cpu_devices); /* * store_cpu_topology is called at boot when only one cpu is running * and with the mutex cpu_hotplug.lock locked, when several cpus have booted, * which prevents simultaneous write access to cpu_topology array */ -void __init store_cpu_topology(unsigned int cpuid) +void store_cpu_topology(unsigned int cpuid) { - struct cputopo_parisc *cpuid_topo = &cpu_topology[cpuid]; + struct cpu_topology *cpuid_topo = &cpu_topology[cpuid]; struct cpuinfo_parisc *p; int max_socket = -1; unsigned long cpu; @@ -71,6 +36,12 @@ void __init store_cpu_topology(unsigned int cpuid) if (cpuid_topo->core_id != -1) return; +#ifdef CONFIG_HOTPLUG_CPU + per_cpu(cpu_devices, cpuid).hotpluggable = 1; +#endif + if (register_cpu(&per_cpu(cpu_devices, cpuid), cpuid)) + pr_warn("Failed to register CPU%d device", cpuid); + /* create cpu topology mapping */ cpuid_topo->thread_id = -1; cpuid_topo->core_id = 0; @@ -86,57 +57,31 @@ void __init store_cpu_topology(unsigned int cpuid) cpuid_topo->core_id = cpu_topology[cpu].core_id; if (p->cpu_loc) { cpuid_topo->core_id++; - cpuid_topo->socket_id = cpu_topology[cpu].socket_id; - dualcores_found = 1; + cpuid_topo->package_id = cpu_topology[cpu].package_id; continue; } } - if (cpuid_topo->socket_id == -1) - max_socket = max(max_socket, cpu_topology[cpu].socket_id); + if (cpuid_topo->package_id == -1) + max_socket = max(max_socket, cpu_topology[cpu].package_id); } - if (cpuid_topo->socket_id == -1) - cpuid_topo->socket_id = max_socket + 1; + if (cpuid_topo->package_id == -1) + cpuid_topo->package_id = max_socket + 1; update_siblings_masks(cpuid); - pr_info("CPU%u: thread %d, cpu %d, socket %d\n", - cpuid, cpu_topology[cpuid].thread_id, + pr_info("CPU%u: cpu core %d of socket %d\n", + cpuid, cpu_topology[cpuid].core_id, - cpu_topology[cpuid].socket_id); + cpu_topology[cpuid].package_id); } -static struct sched_domain_topology_level parisc_mc_topology[] = { -#ifdef CONFIG_SCHED_MC - { cpu_coregroup_mask, cpu_core_flags, SD_INIT_NAME(MC) }, -#endif - - { cpu_cpu_mask, SD_INIT_NAME(DIE) }, - { NULL, }, -}; - /* * init_cpu_topology is called at boot when only one cpu is running * which prevent simultaneous write access to cpu_topology array */ void __init init_cpu_topology(void) { - unsigned int cpu; - - /* init core mask and capacity */ - for_each_possible_cpu(cpu) { - struct cputopo_parisc *cpu_topo = &(cpu_topology[cpu]); - - cpu_topo->thread_id = -1; - cpu_topo->core_id = -1; - cpu_topo->socket_id = -1; - cpumask_clear(&cpu_topo->core_sibling); - cpumask_clear(&cpu_topo->thread_sibling); - } - smp_wmb(); - - /* Set scheduler topology descriptor */ - if (dualcores_found) - set_sched_topology(parisc_mc_topology); + reset_cpu_topology(); } diff --git a/arch/parisc/kernel/traps.c b/arch/parisc/kernel/traps.c index b11fb26ce299..f9696fbf646c 100644 --- a/arch/parisc/kernel/traps.c +++ b/arch/parisc/kernel/traps.c @@ -239,13 +239,6 @@ void die_if_kernel(char *str, struct pt_regs *regs, long err) /* unlock the pdc lock if necessary */ pdc_emergency_unlock(); - /* maybe the kernel hasn't booted very far yet and hasn't been able - * to initialize the serial or STI console. In that case we should - * re-enable the pdc console, so that the user will be able to - * identify the problem. */ - if (!console_drivers) - pdc_console_restart(); - if (err) printk(KERN_CRIT "%s (pid %d): %s (code %ld)\n", current->comm, task_pid_nr(current), str, err); @@ -269,7 +262,7 @@ void die_if_kernel(char *str, struct pt_regs *regs, long err) panic("Fatal exception"); oops_exit(); - do_exit(SIGSEGV); + make_task_dead(SIGSEGV); } /* gdb uses break 4,8 */ @@ -302,7 +295,10 @@ static void handle_break(struct pt_regs *regs) parisc_kprobe_break_handler(regs); return; } - + if (unlikely(iir == PARISC_KPROBES_BREAK_INSN2)) { + parisc_kprobe_ss_handler(regs); + return; + } #endif #ifdef CONFIG_KGDB @@ -426,10 +422,6 @@ void parisc_terminate(char *msg, struct pt_regs *regs, int code, unsigned long o /* unlock the pdc lock if necessary */ pdc_emergency_unlock(); - /* restart pdc console if necessary */ - if (!console_drivers) - pdc_console_restart(); - /* Not all paths will gutter the processor... */ switch(code){ @@ -466,7 +458,7 @@ void parisc_terminate(char *msg, struct pt_regs *regs, int code, unsigned long o * panic notifiers, and we should call panic * directly from the location that we wish. * e.g. We should not call panic from - * parisc_terminate, but rather the oter way around. + * parisc_terminate, but rather the other way around. * This hack works, prints the panic message twice, * and it enables reboot timers! */ @@ -479,9 +471,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs) unsigned long fault_space = 0; int si_code; - if (code == 1) - pdc_console_restart(); /* switch back to pdc if HPMC */ - else if (!irqs_disabled_flags(regs->gr[0])) + if (!irqs_disabled_flags(regs->gr[0])) local_irq_enable(); /* Security check: @@ -539,11 +529,6 @@ void notrace handle_interruption(int code, struct pt_regs *regs) /* Recovery counter trap */ regs->gr[0] &= ~PSW_R; -#ifdef CONFIG_KPROBES - if (parisc_kprobe_ss_handler(regs)) - return; -#endif - #ifdef CONFIG_KGDB if (kgdb_single_step) { kgdb_handle_exception(0, SIGTRAP, 0, regs); @@ -662,6 +647,8 @@ void notrace handle_interruption(int code, struct pt_regs *regs) by hand. Technically we need to emulate: fdc,fdce,pdc,"fic,4f",prober,probeir,probew, probeiw */ + if (code == 17 && handle_nadtlb_fault(regs)) + return; fault_address = regs->ior; fault_space = regs->isr; break; @@ -730,6 +717,8 @@ void notrace handle_interruption(int code, struct pt_regs *regs) } mmap_read_unlock(current->mm); } + /* CPU could not fetch instruction, so clear stale IIR value. */ + regs->iir = 0xbaadf00d; fallthrough; case 27: /* Data memory protection ID trap */ @@ -783,7 +772,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs) * unless pagefault_disable() was called before. */ - if (fault_space == 0 && !faulthandler_disabled()) + if (faulthandler_disabled() || fault_space == 0) { /* Clean up and return if in exception table. */ if (fixup_exception(regs)) diff --git a/arch/parisc/kernel/unaligned.c b/arch/parisc/kernel/unaligned.c index 237d20dd5622..e8a4d77cff53 100644 --- a/arch/parisc/kernel/unaligned.c +++ b/arch/parisc/kernel/unaligned.c @@ -3,14 +3,11 @@ * Unaligned memory access handler * * Copyright (C) 2001 Randolph Chung <tausq@debian.org> + * Copyright (C) 2022 Helge Deller <deller@gmx.de> * Significantly tweaked by LaMont Jones <lamont@debian.org> */ -#include <linux/jiffies.h> -#include <linux/kernel.h> -#include <linux/module.h> #include <linux/sched/signal.h> -#include <linux/sched/debug.h> #include <linux/signal.h> #include <linux/ratelimit.h> #include <linux/uaccess.h> @@ -25,18 +22,7 @@ #define DPRINTF(fmt, args...) #endif -#ifdef CONFIG_64BIT -#define RFMT "%016lx" -#else -#define RFMT "%08lx" -#endif - -#define FIXUP_BRANCH(lbl) \ - "\tldil L%%" #lbl ", %%r1\n" \ - "\tldo R%%" #lbl "(%%r1), %%r1\n" \ - "\tbv,n %%r0(%%r1)\n" -/* If you use FIXUP_BRANCH, then you must list this clobber */ -#define FIXUP_BRANCH_CLOBBER "r1" +#define RFMT "%#08lx" /* 1111 1100 0000 0000 0001 0011 1100 0000 */ #define OPCODE1(a,b,c) ((a)<<26|(b)<<12|(c)<<6) @@ -107,44 +93,37 @@ #define R1(i) (((i)>>21)&0x1f) #define R2(i) (((i)>>16)&0x1f) #define R3(i) ((i)&0x1f) -#define FR3(i) ((((i)<<1)&0x1f)|(((i)>>6)&1)) +#define FR3(i) ((((i)&0x1f)<<1)|(((i)>>6)&1)) #define IM(i,n) (((i)>>1&((1<<(n-1))-1))|((i)&1?((0-1L)<<(n-1)):0)) #define IM5_2(i) IM((i)>>16,5) #define IM5_3(i) IM((i),5) #define IM14(i) IM((i),14) #define ERR_NOTHANDLED -1 -#define ERR_PAGEFAULT -2 int unaligned_enabled __read_mostly = 1; static int emulate_ldh(struct pt_regs *regs, int toreg) { unsigned long saddr = regs->ior; - unsigned long val = 0; - int ret; + unsigned long val = 0, temp1; + ASM_EXCEPTIONTABLE_VAR(ret); DPRINTF("load " RFMT ":" RFMT " to r%d for 2 bytes\n", regs->isr, regs->ior, toreg); __asm__ __volatile__ ( " mtsp %4, %%sr1\n" -"1: ldbs 0(%%sr1,%3), %%r20\n" +"1: ldbs 0(%%sr1,%3), %2\n" "2: ldbs 1(%%sr1,%3), %0\n" -" depw %%r20, 23, 24, %0\n" -" copy %%r0, %1\n" +" depw %2, 23, 24, %0\n" "3: \n" -" .section .fixup,\"ax\"\n" -"4: ldi -2, %1\n" - FIXUP_BRANCH(3b) -" .previous\n" - ASM_EXCEPTIONTABLE_ENTRY(1b, 4b) - ASM_EXCEPTIONTABLE_ENTRY(2b, 4b) - : "=r" (val), "=r" (ret) - : "0" (val), "r" (saddr), "r" (regs->isr) - : "r20", FIXUP_BRANCH_CLOBBER ); + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + : "+r" (val), "+r" (ret), "=&r" (temp1) + : "r" (saddr), "r" (regs->isr) ); - DPRINTF("val = 0x" RFMT "\n", val); + DPRINTF("val = " RFMT "\n", val); if (toreg) regs->gr[toreg] = val; @@ -155,34 +134,28 @@ static int emulate_ldh(struct pt_regs *regs, int toreg) static int emulate_ldw(struct pt_regs *regs, int toreg, int flop) { unsigned long saddr = regs->ior; - unsigned long val = 0; - int ret; + unsigned long val = 0, temp1, temp2; + ASM_EXCEPTIONTABLE_VAR(ret); DPRINTF("load " RFMT ":" RFMT " to r%d for 4 bytes\n", regs->isr, regs->ior, toreg); __asm__ __volatile__ ( -" zdep %3,28,2,%%r19\n" /* r19=(ofs&3)*8 */ -" mtsp %4, %%sr1\n" -" depw %%r0,31,2,%3\n" -"1: ldw 0(%%sr1,%3),%0\n" -"2: ldw 4(%%sr1,%3),%%r20\n" -" subi 32,%%r19,%%r19\n" -" mtctl %%r19,11\n" -" vshd %0,%%r20,%0\n" -" copy %%r0, %1\n" +" zdep %4,28,2,%2\n" /* r19=(ofs&3)*8 */ +" mtsp %5, %%sr1\n" +" depw %%r0,31,2,%4\n" +"1: ldw 0(%%sr1,%4),%0\n" +"2: ldw 4(%%sr1,%4),%3\n" +" subi 32,%2,%2\n" +" mtctl %2,11\n" +" vshd %0,%3,%0\n" "3: \n" -" .section .fixup,\"ax\"\n" -"4: ldi -2, %1\n" - FIXUP_BRANCH(3b) -" .previous\n" - ASM_EXCEPTIONTABLE_ENTRY(1b, 4b) - ASM_EXCEPTIONTABLE_ENTRY(2b, 4b) - : "=r" (val), "=r" (ret) - : "0" (val), "r" (saddr), "r" (regs->isr) - : "r19", "r20", FIXUP_BRANCH_CLOBBER ); + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + : "+r" (val), "+r" (ret), "=&r" (temp1), "=&r" (temp2) + : "r" (saddr), "r" (regs->isr) ); - DPRINTF("val = 0x" RFMT "\n", val); + DPRINTF("val = " RFMT "\n", val); if (flop) ((__u32*)(regs->fr))[toreg] = val; @@ -195,16 +168,15 @@ static int emulate_ldd(struct pt_regs *regs, int toreg, int flop) { unsigned long saddr = regs->ior; __u64 val = 0; - int ret; + ASM_EXCEPTIONTABLE_VAR(ret); DPRINTF("load " RFMT ":" RFMT " to r%d for 8 bytes\n", regs->isr, regs->ior, toreg); -#ifdef CONFIG_PA20 -#ifndef CONFIG_64BIT - if (!flop) - return -1; -#endif + if (!IS_ENABLED(CONFIG_64BIT) && !flop) + return ERR_NOTHANDLED; + +#ifdef CONFIG_64BIT __asm__ __volatile__ ( " depd,z %3,60,3,%%r19\n" /* r19=(ofs&7)*8 */ " mtsp %4, %%sr1\n" @@ -214,44 +186,32 @@ static int emulate_ldd(struct pt_regs *regs, int toreg, int flop) " subi 64,%%r19,%%r19\n" " mtsar %%r19\n" " shrpd %0,%%r20,%%sar,%0\n" -" copy %%r0, %1\n" "3: \n" -" .section .fixup,\"ax\"\n" -"4: ldi -2, %1\n" - FIXUP_BRANCH(3b) -" .previous\n" - ASM_EXCEPTIONTABLE_ENTRY(1b,4b) - ASM_EXCEPTIONTABLE_ENTRY(2b,4b) - : "=r" (val), "=r" (ret) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + : "=r" (val), "+r" (ret) : "0" (val), "r" (saddr), "r" (regs->isr) - : "r19", "r20", FIXUP_BRANCH_CLOBBER ); + : "r19", "r20" ); #else { - unsigned long valh=0,vall=0; + unsigned long shift, temp1; __asm__ __volatile__ ( -" zdep %5,29,2,%%r19\n" /* r19=(ofs&3)*8 */ -" mtsp %6, %%sr1\n" -" dep %%r0,31,2,%5\n" -"1: ldw 0(%%sr1,%5),%0\n" -"2: ldw 4(%%sr1,%5),%1\n" -"3: ldw 8(%%sr1,%5),%%r20\n" -" subi 32,%%r19,%%r19\n" -" mtsar %%r19\n" -" vshd %0,%1,%0\n" -" vshd %1,%%r20,%1\n" -" copy %%r0, %2\n" +" zdep %2,29,2,%3\n" /* r19=(ofs&3)*8 */ +" mtsp %5, %%sr1\n" +" dep %%r0,31,2,%2\n" +"1: ldw 0(%%sr1,%2),%0\n" +"2: ldw 4(%%sr1,%2),%R0\n" +"3: ldw 8(%%sr1,%2),%4\n" +" subi 32,%3,%3\n" +" mtsar %3\n" +" vshd %0,%R0,%0\n" +" vshd %R0,%4,%R0\n" "4: \n" -" .section .fixup,\"ax\"\n" -"5: ldi -2, %2\n" - FIXUP_BRANCH(4b) -" .previous\n" - ASM_EXCEPTIONTABLE_ENTRY(1b,5b) - ASM_EXCEPTIONTABLE_ENTRY(2b,5b) - ASM_EXCEPTIONTABLE_ENTRY(3b,5b) - : "=r" (valh), "=r" (vall), "=r" (ret) - : "0" (valh), "1" (vall), "r" (saddr), "r" (regs->isr) - : "r19", "r20", FIXUP_BRANCH_CLOBBER ); - val=((__u64)valh<<32)|(__u64)vall; + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 4b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 4b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 4b) + : "+r" (val), "+r" (ret), "+r" (saddr), "=&r" (shift), "=&r" (temp1) + : "r" (regs->isr) ); } #endif @@ -267,31 +227,25 @@ static int emulate_ldd(struct pt_regs *regs, int toreg, int flop) static int emulate_sth(struct pt_regs *regs, int frreg) { - unsigned long val = regs->gr[frreg]; - int ret; + unsigned long val = regs->gr[frreg], temp1; + ASM_EXCEPTIONTABLE_VAR(ret); if (!frreg) val = 0; - DPRINTF("store r%d (0x" RFMT ") to " RFMT ":" RFMT " for 2 bytes\n", frreg, + DPRINTF("store r%d (" RFMT ") to " RFMT ":" RFMT " for 2 bytes\n", frreg, val, regs->isr, regs->ior); __asm__ __volatile__ ( -" mtsp %3, %%sr1\n" -" extrw,u %1, 23, 8, %%r19\n" -"1: stb %1, 1(%%sr1, %2)\n" -"2: stb %%r19, 0(%%sr1, %2)\n" -" copy %%r0, %0\n" +" mtsp %4, %%sr1\n" +" extrw,u %2, 23, 8, %1\n" +"1: stb %1, 0(%%sr1, %3)\n" +"2: stb %2, 1(%%sr1, %3)\n" "3: \n" -" .section .fixup,\"ax\"\n" -"4: ldi -2, %0\n" - FIXUP_BRANCH(3b) -" .previous\n" - ASM_EXCEPTIONTABLE_ENTRY(1b,4b) - ASM_EXCEPTIONTABLE_ENTRY(2b,4b) - : "=r" (ret) - : "r" (val), "r" (regs->ior), "r" (regs->isr) - : "r19", FIXUP_BRANCH_CLOBBER ); + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + : "+r" (ret), "=&r" (temp1) + : "r" (val), "r" (regs->ior), "r" (regs->isr) ); return ret; } @@ -299,7 +253,7 @@ static int emulate_sth(struct pt_regs *regs, int frreg) static int emulate_stw(struct pt_regs *regs, int frreg, int flop) { unsigned long val; - int ret; + ASM_EXCEPTIONTABLE_VAR(ret); if (flop) val = ((__u32*)(regs->fr))[frreg]; @@ -308,7 +262,7 @@ static int emulate_stw(struct pt_regs *regs, int frreg, int flop) else val = 0; - DPRINTF("store r%d (0x" RFMT ") to " RFMT ":" RFMT " for 4 bytes\n", frreg, + DPRINTF("store r%d (" RFMT ") to " RFMT ":" RFMT " for 4 bytes\n", frreg, val, regs->isr, regs->ior); @@ -328,24 +282,19 @@ static int emulate_stw(struct pt_regs *regs, int frreg, int flop) " or %%r1, %%r21, %%r21\n" " stw %%r20,0(%%sr1,%2)\n" " stw %%r21,4(%%sr1,%2)\n" -" copy %%r0, %0\n" "3: \n" -" .section .fixup,\"ax\"\n" -"4: ldi -2, %0\n" - FIXUP_BRANCH(3b) -" .previous\n" - ASM_EXCEPTIONTABLE_ENTRY(1b,4b) - ASM_EXCEPTIONTABLE_ENTRY(2b,4b) - : "=r" (ret) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + : "+r" (ret) : "r" (val), "r" (regs->ior), "r" (regs->isr) - : "r19", "r20", "r21", "r22", "r1", FIXUP_BRANCH_CLOBBER ); + : "r19", "r20", "r21", "r22", "r1" ); - return 0; + return ret; } static int emulate_std(struct pt_regs *regs, int frreg, int flop) { __u64 val; - int ret; + ASM_EXCEPTIONTABLE_VAR(ret); if (flop) val = regs->fr[frreg]; @@ -357,11 +306,10 @@ static int emulate_std(struct pt_regs *regs, int frreg, int flop) DPRINTF("store r%d (0x%016llx) to " RFMT ":" RFMT " for 8 bytes\n", frreg, val, regs->isr, regs->ior); -#ifdef CONFIG_PA20 -#ifndef CONFIG_64BIT - if (!flop) - return -1; -#endif + if (!IS_ENABLED(CONFIG_64BIT) && !flop) + return ERR_NOTHANDLED; + +#ifdef CONFIG_64BIT __asm__ __volatile__ ( " mtsp %3, %%sr1\n" " depd,z %2, 60, 3, %%r19\n" @@ -378,26 +326,21 @@ static int emulate_std(struct pt_regs *regs, int frreg, int flop) " or %%r1, %%r21, %%r21\n" "3: std %%r20,0(%%sr1,%2)\n" "4: std %%r21,8(%%sr1,%2)\n" -" copy %%r0, %0\n" "5: \n" -" .section .fixup,\"ax\"\n" -"6: ldi -2, %0\n" - FIXUP_BRANCH(5b) -" .previous\n" - ASM_EXCEPTIONTABLE_ENTRY(1b,6b) - ASM_EXCEPTIONTABLE_ENTRY(2b,6b) - ASM_EXCEPTIONTABLE_ENTRY(3b,6b) - ASM_EXCEPTIONTABLE_ENTRY(4b,6b) - : "=r" (ret) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 5b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 5b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 5b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(4b, 5b) + : "+r" (ret) : "r" (val), "r" (regs->ior), "r" (regs->isr) - : "r19", "r20", "r21", "r22", "r1", FIXUP_BRANCH_CLOBBER ); + : "r19", "r20", "r21", "r22", "r1" ); #else { unsigned long valh=(val>>32),vall=(val&0xffffffffl); __asm__ __volatile__ ( " mtsp %4, %%sr1\n" " zdep %2, 29, 2, %%r19\n" -" dep %%r0, 31, 2, %2\n" +" dep %%r0, 31, 2, %3\n" " mtsar %%r19\n" " zvdepi -2, 32, %%r19\n" "1: ldw 0(%%sr1,%3),%%r20\n" @@ -409,23 +352,18 @@ static int emulate_std(struct pt_regs *regs, int frreg, int flop) " andcm %%r21, %%r19, %%r21\n" " or %1, %%r20, %1\n" " or %2, %%r21, %2\n" -"3: stw %1,0(%%sr1,%1)\n" +"3: stw %1,0(%%sr1,%3)\n" "4: stw %%r1,4(%%sr1,%3)\n" "5: stw %2,8(%%sr1,%3)\n" -" copy %%r0, %0\n" "6: \n" -" .section .fixup,\"ax\"\n" -"7: ldi -2, %0\n" - FIXUP_BRANCH(6b) -" .previous\n" - ASM_EXCEPTIONTABLE_ENTRY(1b,7b) - ASM_EXCEPTIONTABLE_ENTRY(2b,7b) - ASM_EXCEPTIONTABLE_ENTRY(3b,7b) - ASM_EXCEPTIONTABLE_ENTRY(4b,7b) - ASM_EXCEPTIONTABLE_ENTRY(5b,7b) - : "=r" (ret) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 6b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 6b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 6b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(4b, 6b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(5b, 6b) + : "+r" (ret) : "r" (valh), "r" (vall), "r" (regs->ior), "r" (regs->isr) - : "r19", "r20", "r21", "r1", FIXUP_BRANCH_CLOBBER ); + : "r19", "r20", "r21", "r1" ); } #endif @@ -438,7 +376,6 @@ void handle_unaligned(struct pt_regs *regs) unsigned long newbase = R1(regs->iir)?regs->gr[R1(regs->iir)]:0; int modify = 0; int ret = ERR_NOTHANDLED; - register int flop=0; /* true if this is a flop */ __inc_irq_stat(irq_unaligned_count); @@ -450,10 +387,10 @@ void handle_unaligned(struct pt_regs *regs) if (!(current->thread.flags & PARISC_UAC_NOPRINT) && __ratelimit(&ratelimit)) { - char buf[256]; - sprintf(buf, "%s(%d): unaligned access to 0x" RFMT " at ip=0x" RFMT "\n", - current->comm, task_pid_nr(current), regs->ior, regs->iaoq[0]); - printk(KERN_WARNING "%s", buf); + printk(KERN_WARNING "%s(%d): unaligned access to " RFMT + " at ip " RFMT " (iir " RFMT ")\n", + current->comm, task_pid_nr(current), regs->ior, + regs->iaoq[0], regs->iir); #ifdef DEBUG_UNALIGNED show_regs(regs); #endif @@ -547,7 +484,7 @@ void handle_unaligned(struct pt_regs *regs) ret = emulate_stw(regs, R2(regs->iir),0); break; -#ifdef CONFIG_PA20 +#ifdef CONFIG_64BIT case OPCODE_LDD_I: case OPCODE_LDDA_I: case OPCODE_LDD_S: @@ -565,13 +502,11 @@ void handle_unaligned(struct pt_regs *regs) case OPCODE_FLDWS: case OPCODE_FLDWXR: case OPCODE_FLDWSR: - flop=1; ret = emulate_ldw(regs,FR3(regs->iir),1); break; case OPCODE_FLDDX: case OPCODE_FLDDS: - flop=1; ret = emulate_ldd(regs,R3(regs->iir),1); break; @@ -579,13 +514,11 @@ void handle_unaligned(struct pt_regs *regs) case OPCODE_FSTWS: case OPCODE_FSTWXR: case OPCODE_FSTWSR: - flop=1; ret = emulate_stw(regs,FR3(regs->iir),1); break; case OPCODE_FSTDX: case OPCODE_FSTDS: - flop=1; ret = emulate_std(regs,R3(regs->iir),1); break; @@ -596,37 +529,33 @@ void handle_unaligned(struct pt_regs *regs) ret = ERR_NOTHANDLED; /* "undefined", but lets kill them. */ break; } -#ifdef CONFIG_PA20 switch (regs->iir & OPCODE2_MASK) { case OPCODE_FLDD_L: - flop=1; ret = emulate_ldd(regs,R2(regs->iir),1); break; case OPCODE_FSTD_L: - flop=1; ret = emulate_std(regs, R2(regs->iir),1); break; +#ifdef CONFIG_64BIT case OPCODE_LDD_L: ret = emulate_ldd(regs, R2(regs->iir),0); break; case OPCODE_STD_L: ret = emulate_std(regs, R2(regs->iir),0); break; - } #endif + } switch (regs->iir & OPCODE3_MASK) { case OPCODE_FLDW_L: - flop=1; - ret = emulate_ldw(regs, R2(regs->iir),0); + ret = emulate_ldw(regs, R2(regs->iir), 1); break; case OPCODE_LDW_M: - ret = emulate_ldw(regs, R2(regs->iir),1); + ret = emulate_ldw(regs, R2(regs->iir), 0); break; case OPCODE_FSTW_L: - flop=1; ret = emulate_stw(regs, R2(regs->iir),1); break; case OPCODE_STW_M: @@ -673,7 +602,7 @@ void handle_unaligned(struct pt_regs *regs) printk(KERN_CRIT "Unaligned handler failed, ret = %d\n", ret); die_if_kernel("Unaligned data reference", regs, 28); - if (ret == ERR_PAGEFAULT) + if (ret == -EFAULT) { force_sig_fault(SIGSEGV, SEGV_MAPERR, (void __user *)regs->ior); diff --git a/arch/parisc/kernel/vdso.c b/arch/parisc/kernel/vdso.c new file mode 100644 index 000000000000..47e5960a2f96 --- /dev/null +++ b/arch/parisc/kernel/vdso.c @@ -0,0 +1,122 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (c) 2022 Helge Deller <deller@gmx.de> + * + * based on arch/s390/kernel/vdso.c which is + * Copyright IBM Corp. 2008 + * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com) + */ + +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/slab.h> +#include <linux/elf.h> +#include <linux/timekeeper_internal.h> +#include <linux/compat.h> +#include <linux/nsproxy.h> +#include <linux/time_namespace.h> +#include <linux/random.h> + +#include <asm/pgtable.h> +#include <asm/page.h> +#include <asm/sections.h> +#include <asm/vdso.h> +#include <asm/cacheflush.h> + +extern char vdso32_start, vdso32_end; +extern char vdso64_start, vdso64_end; + +static int vdso_mremap(const struct vm_special_mapping *sm, + struct vm_area_struct *vma) +{ + current->mm->context.vdso_base = vma->vm_start; + return 0; +} + +#ifdef CONFIG_64BIT +static struct vm_special_mapping vdso64_mapping = { + .name = "[vdso]", + .mremap = vdso_mremap, +}; +#endif + +static struct vm_special_mapping vdso32_mapping = { + .name = "[vdso]", + .mremap = vdso_mremap, +}; + +/* + * This is called from binfmt_elf, we create the special vma for the + * vDSO and insert it into the mm struct tree + */ +int arch_setup_additional_pages(struct linux_binprm *bprm, + int executable_stack) +{ + + unsigned long vdso_text_start, vdso_text_len, map_base; + struct vm_special_mapping *vdso_mapping; + struct mm_struct *mm = current->mm; + struct vm_area_struct *vma; + int rc; + + if (mmap_write_lock_killable(mm)) + return -EINTR; + +#ifdef CONFIG_64BIT + if (!is_compat_task()) { + vdso_text_len = &vdso64_end - &vdso64_start; + vdso_mapping = &vdso64_mapping; + } else +#endif + { + vdso_text_len = &vdso32_end - &vdso32_start; + vdso_mapping = &vdso32_mapping; + } + + map_base = mm->mmap_base; + if (current->flags & PF_RANDOMIZE) + map_base -= prandom_u32_max(0x20) * PAGE_SIZE; + + vdso_text_start = get_unmapped_area(NULL, map_base, vdso_text_len, 0, 0); + + /* VM_MAYWRITE for COW so gdb can set breakpoints */ + vma = _install_special_mapping(mm, vdso_text_start, vdso_text_len, + VM_READ|VM_EXEC| + VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC, + vdso_mapping); + if (IS_ERR(vma)) { + do_munmap(mm, vdso_text_start, PAGE_SIZE, NULL); + rc = PTR_ERR(vma); + } else { + current->mm->context.vdso_base = vdso_text_start; + rc = 0; + } + + mmap_write_unlock(mm); + return rc; +} + +static struct page ** __init vdso_setup_pages(void *start, void *end) +{ + int pages = (end - start) >> PAGE_SHIFT; + struct page **pagelist; + int i; + + pagelist = kcalloc(pages + 1, sizeof(struct page *), GFP_KERNEL); + if (!pagelist) + panic("%s: Cannot allocate page list for VDSO", __func__); + for (i = 0; i < pages; i++) + pagelist[i] = virt_to_page(start + i * PAGE_SIZE); + return pagelist; +} + +static int __init vdso_init(void) +{ +#ifdef CONFIG_64BIT + vdso64_mapping.pages = vdso_setup_pages(&vdso64_start, &vdso64_end); +#endif + if (IS_ENABLED(CONFIG_COMPAT) || !IS_ENABLED(CONFIG_64BIT)) + vdso32_mapping.pages = vdso_setup_pages(&vdso32_start, &vdso32_end); + return 0; +} +arch_initcall(vdso_init); diff --git a/arch/parisc/kernel/vdso32/Makefile b/arch/parisc/kernel/vdso32/Makefile new file mode 100644 index 000000000000..85b1c6d261d1 --- /dev/null +++ b/arch/parisc/kernel/vdso32/Makefile @@ -0,0 +1,53 @@ +# List of files in the vdso, has to be asm only for now + +obj-vdso32 = note.o sigtramp.o restart_syscall.o + +# Build rules + +targets := $(obj-vdso32) vdso32.so +obj-vdso32 := $(addprefix $(obj)/, $(obj-vdso32)) + +ccflags-y := -shared -fno-common -fbuiltin -mno-fast-indirect-calls -O2 -mno-long-calls +# -march=1.1 -mschedule=7100LC +ccflags-y += -nostdlib -Wl,-soname=linux-vdso32.so.1 \ + $(call ld-option, -Wl$(comma)--hash-style=sysv) +asflags-y := -D__VDSO32__ -s + +KBUILD_AFLAGS += -DBUILD_VDSO +KBUILD_CFLAGS += -DBUILD_VDSO -DDISABLE_BRANCH_PROFILING + +VDSO_LIBGCC := $(shell $(CROSS32CC) -print-libgcc-file-name) + +obj-y += vdso32_wrapper.o +extra-y += vdso32.lds +CPPFLAGS_vdso32.lds += -P -C # -U$(ARCH) + +$(obj)/vdso32_wrapper.o : $(obj)/vdso32.so FORCE + +# Force dependency (incbin is bad) +# link rule for the .so file, .lds has to be first +$(obj)/vdso32.so: $(src)/vdso32.lds $(obj-vdso32) $(obj-cvdso32) $(VDSO_LIBGCC) + $(call if_changed,vdso32ld) + +# assembly rules for the .S files +$(obj-vdso32): %.o: %.S FORCE + $(call if_changed_dep,vdso32as) + +$(obj-cvdso32): %.o: %.c FORCE + $(call if_changed_dep,vdso32cc) + +# actual build commands +quiet_cmd_vdso32ld = VDSO32L $@ + cmd_vdso32ld = $(CROSS32CC) $(c_flags) -Wl,-T $^ -o $@ +quiet_cmd_vdso32as = VDSO32A $@ + cmd_vdso32as = $(CROSS32CC) $(a_flags) -c -o $@ $< +quiet_cmd_vdso32cc = VDSO32C $@ + cmd_vdso32cc = $(CROSS32CC) $(c_flags) -c -fPIC -mno-fast-indirect-calls -o $@ $< + +# Generate VDSO offsets using helper script +gen-vdsosym := $(srctree)/$(src)/gen_vdso_offsets.sh +quiet_cmd_vdsosym = VDSOSYM $@ + cmd_vdsosym = $(NM) $< | $(gen-vdsosym) | LC_ALL=C sort > $@ + +include/generated/vdso32-offsets.h: $(obj)/vdso32.so FORCE + $(call if_changed,vdsosym) diff --git a/arch/parisc/kernel/vdso32/gen_vdso_offsets.sh b/arch/parisc/kernel/vdso32/gen_vdso_offsets.sh new file mode 100755 index 000000000000..da39d6cff7f1 --- /dev/null +++ b/arch/parisc/kernel/vdso32/gen_vdso_offsets.sh @@ -0,0 +1,15 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 + +# +# Match symbols in the DSO that look like VDSO_*; produce a header file +# of constant offsets into the shared object. +# +# Doing this inside the Makefile will break the $(filter-out) function, +# causing Kbuild to rebuild the vdso-offsets header file every time. +# +# Inspired by arm64 version. +# + +LC_ALL=C +sed -n 's/\([0-9a-f]*\) . __kernel_\(.*\)/\#define vdso32_offset_\2\t0x\1/p' diff --git a/arch/parisc/kernel/vdso32/note.S b/arch/parisc/kernel/vdso32/note.S new file mode 100644 index 000000000000..bb350918bebd --- /dev/null +++ b/arch/parisc/kernel/vdso32/note.S @@ -0,0 +1,26 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text. + * Here we can supply some information useful to userland. + */ + +#include <linux/uts.h> +#include <linux/version.h> + +#define ASM_ELF_NOTE_BEGIN(name, flags, vendor, type) \ + .section name, flags; \ + .balign 4; \ + .long 1f - 0f; /* name length */ \ + .long 3f - 2f; /* data length */ \ + .long type; /* note type */ \ +0: .asciz vendor; /* vendor name */ \ +1: .balign 4; \ +2: + +#define ASM_ELF_NOTE_END \ +3: .balign 4; /* pad out section */ \ + .previous + + ASM_ELF_NOTE_BEGIN(".note.kernel-version", "a", UTS_SYSNAME, 0) + .long LINUX_VERSION_CODE + ASM_ELF_NOTE_END diff --git a/arch/parisc/kernel/vdso32/restart_syscall.S b/arch/parisc/kernel/vdso32/restart_syscall.S new file mode 100644 index 000000000000..7e82008d7e40 --- /dev/null +++ b/arch/parisc/kernel/vdso32/restart_syscall.S @@ -0,0 +1,32 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Syscall restart trampoline for 32 and 64 bits processes. + * + * Copyright (C) 2018-2022 Helge Deller <deller@gmx.de> + * Copyright (C) 2022 John David Anglin <dave.anglin@bell.net> + */ + +#include <asm/unistd.h> +#include <asm/vdso.h> + +#include <linux/linkage.h> + + .text + +ENTRY_CFI(__kernel_restart_syscall) + /* + * Setup a trampoline to restart the syscall + * with __NR_restart_syscall + */ + + /* load return pointer */ +#if defined(__VDSO64__) + ldd 0(%sp), %r31 +#elif defined(__VDSO32__) + ldw 0(%sp), %r31 +#endif + + be 0x100(%sr2, %r0) + ldi __NR_restart_syscall, %r20 + +ENDPROC_CFI(__kernel_restart_syscall) diff --git a/arch/parisc/kernel/vdso32/sigtramp.S b/arch/parisc/kernel/vdso32/sigtramp.S new file mode 100644 index 000000000000..192da7077869 --- /dev/null +++ b/arch/parisc/kernel/vdso32/sigtramp.S @@ -0,0 +1,195 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Signal trampolines for 32 bit processes. + * + * Copyright (C) 2006 Randolph Chung <tausq@debian.org> + * Copyright (C) 2018-2022 Helge Deller <deller@gmx.de> + * Copyright (C) 2022 John David Anglin <dave.anglin@bell.net> + */ +#include <asm/unistd.h> +#include <linux/linkage.h> +#include <generated/asm-offsets.h> + + .text + +/* Gdb expects the trampoline is on the stack and the pc is offset from + a 64-byte boundary by 0, 4 or 5 instructions. Since the vdso trampoline + is not on the stack, we need a new variant with different offsets and + data to tell gdb where to find the signal context on the stack. + + Here we put the offset to the context data at the start of the trampoline + region and offset the first trampoline by 2 instructions. Please do + not change the trampoline as the code in gdb depends on the following + instruction sequence exactly. + */ + .align 64 + .word SIGFRAME_CONTEXT_REGS32 + +/* The nop here is a hack. The dwarf2 unwind routines subtract 1 from + the return address to get an address in the middle of the presumed + call instruction. Since we don't have a call here, we artifically + extend the range covered by the unwind info by adding a nop before + the real start. + */ + nop + + .globl __kernel_sigtramp_rt + .type __kernel_sigtramp_rt, @function +__kernel_sigtramp_rt: + .proc + .callinfo FRAME=ASM_SIGFRAME_SIZE32,CALLS,SAVE_RP + .entry + +.Lsigrt_start = . - 4 +0: ldi 0, %r25 /* (in_syscall=0) */ + ldi __NR_rt_sigreturn, %r20 + ble 0x100(%sr2, %r0) + nop + +1: ldi 1, %r25 /* (in_syscall=1) */ + ldi __NR_rt_sigreturn, %r20 + ble 0x100(%sr2, %r0) + nop +.Lsigrt_end: + .exit + .procend + .size __kernel_sigtramp_rt,.-__kernel_sigtramp_rt + + + .section .eh_frame,"a",@progbits + +/* This is where the mcontext_t struct can be found on the stack. */ +#define PTREGS SIGFRAME_CONTEXT_REGS32 /* 32-bit process offset is -672 */ + +/* Register REGNO can be found at offset OFS of the mcontext_t structure. */ + .macro rsave regno,ofs + .byte 0x05 /* DW_CFA_offset_extended */ + .uleb128 \regno; /* regno */ + .uleb128 \ofs /* factored offset */ + .endm + +.Lcie: + .long .Lcie_end - .Lcie_start +.Lcie_start: + .long 0 /* CIE ID */ + .byte 1 /* Version number */ + .stringz "zRS" /* NUL-terminated augmentation string */ + .uleb128 4 /* Code alignment factor */ + .sleb128 4 /* Data alignment factor */ + .byte 89 /* Return address register column, iaoq[0] */ + .uleb128 1 /* Augmentation value length */ + .byte 0x1b /* DW_EH_PE_pcrel | DW_EH_PE_sdata4. */ + .byte 0x0f /* DW_CFA_def_cfa_expresion */ + .uleb128 9f - 1f /* length */ +1: + .byte 0x8e /* DW_OP_breg30 */ + .sleb128 PTREGS +9: + .balign 4 +.Lcie_end: + + .long .Lfde0_end - .Lfde0_start +.Lfde0_start: + .long .Lfde0_start - .Lcie /* CIE pointer. */ + .long .Lsigrt_start - . /* PC start, length */ + .long .Lsigrt_end - .Lsigrt_start + .uleb128 0 /* Augmentation */ + + /* General registers */ + rsave 1, 2 + rsave 2, 3 + rsave 3, 4 + rsave 4, 5 + rsave 5, 6 + rsave 6, 7 + rsave 7, 8 + rsave 8, 9 + rsave 9, 10 + rsave 10, 11 + rsave 11, 12 + rsave 12, 13 + rsave 13, 14 + rsave 14, 15 + rsave 15, 16 + rsave 16, 17 + rsave 17, 18 + rsave 18, 19 + rsave 19, 20 + rsave 20, 21 + rsave 21, 22 + rsave 22, 23 + rsave 23, 24 + rsave 24, 25 + rsave 25, 26 + rsave 26, 27 + rsave 27, 28 + rsave 28, 29 + rsave 29, 30 + rsave 30, 31 + rsave 31, 32 + + /* Floating-point registers */ + rsave 32, 42 + rsave 33, 43 + rsave 34, 44 + rsave 35, 45 + rsave 36, 46 + rsave 37, 47 + rsave 38, 48 + rsave 39, 49 + rsave 40, 50 + rsave 41, 51 + rsave 42, 52 + rsave 43, 53 + rsave 44, 54 + rsave 45, 55 + rsave 46, 56 + rsave 47, 57 + rsave 48, 58 + rsave 49, 59 + rsave 50, 60 + rsave 51, 61 + rsave 52, 62 + rsave 53, 63 + rsave 54, 64 + rsave 55, 65 + rsave 56, 66 + rsave 57, 67 + rsave 58, 68 + rsave 59, 69 + rsave 60, 70 + rsave 61, 71 + rsave 62, 72 + rsave 63, 73 + rsave 64, 74 + rsave 65, 75 + rsave 66, 76 + rsave 67, 77 + rsave 68, 78 + rsave 69, 79 + rsave 70, 80 + rsave 71, 81 + rsave 72, 82 + rsave 73, 83 + rsave 74, 84 + rsave 75, 85 + rsave 76, 86 + rsave 77, 87 + rsave 78, 88 + rsave 79, 89 + rsave 80, 90 + rsave 81, 91 + rsave 82, 92 + rsave 83, 93 + rsave 84, 94 + rsave 85, 95 + rsave 86, 96 + rsave 87, 97 + + /* SAR register */ + rsave 88, 102 + + /* iaoq[0] return address register */ + rsave 89, 100 + .balign 4 +.Lfde0_end: diff --git a/arch/parisc/kernel/vdso32/vdso32.lds.S b/arch/parisc/kernel/vdso32/vdso32.lds.S new file mode 100644 index 000000000000..d4aff3af5262 --- /dev/null +++ b/arch/parisc/kernel/vdso32/vdso32.lds.S @@ -0,0 +1,111 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * This is the infamous ld script for the 32 bits vdso library + */ +#include <asm/vdso.h> +#include <asm/page.h> + +/* Default link addresses for the vDSOs */ +OUTPUT_FORMAT("elf32-hppa-linux") +OUTPUT_ARCH(hppa) +ENTRY(_start) + +SECTIONS +{ + . = VDSO_LBASE + SIZEOF_HEADERS; + .hash : { *(.hash) } :text + .gnu.hash : { *(.gnu.hash) } + .dynsym : { *(.dynsym) } + .dynstr : { *(.dynstr) } + .gnu.version : { *(.gnu.version) } + .gnu.version_d : { *(.gnu.version_d) } + .gnu.version_r : { *(.gnu.version_r) } + + .note : { *(.note.*) } :text :note + + . = ALIGN (16); + .text : + { + *(.text .stub .text.* .gnu.linkonce.t.*) + } + PROVIDE (__etext = .); + PROVIDE (_etext = .); + PROVIDE (etext = .); + + /* Other stuff is appended to the text segment: */ + .rodata : { *(.rodata .rodata.* .gnu.linkonce.r.*) } + .rodata1 : { *(.rodata1) } + .rodata2 : { *(.data.rel.ro) } + + .eh_frame_hdr : { *(.eh_frame_hdr) } :text :eh_frame_hdr + .eh_frame : { KEEP (*(.eh_frame)) } :text + .gcc_except_table : { *(.gcc_except_table) } + .fixup : { *(.fixup) } + + .dynamic : { *(.dynamic) } :text :dynamic + .plt : { *(.plt) } + .got : { *(.got) } + + _end = .; + __end = .; + PROVIDE (end = .); + + + /* Stabs debugging sections are here too + */ + .stab 0 : { *(.stab) } + .stabstr 0 : { *(.stabstr) } + .stab.excl 0 : { *(.stab.excl) } + .stab.exclstr 0 : { *(.stab.exclstr) } + .stab.index 0 : { *(.stab.index) } + .stab.indexstr 0 : { *(.stab.indexstr) } + .comment 0 : { *(.comment) } + .debug 0 : { *(.debug) } + .line 0 : { *(.line) } + + .debug_srcinfo 0 : { *(.debug_srcinfo) } + .debug_sfnames 0 : { *(.debug_sfnames) } + + .debug_aranges 0 : { *(.debug_aranges) } + .debug_pubnames 0 : { *(.debug_pubnames) } + + .debug_info 0 : { *(.debug_info .gnu.linkonce.wi.*) } + .debug_abbrev 0 : { *(.debug_abbrev) } + .debug_line 0 : { *(.debug_line) } + .debug_frame 0 : { *(.debug_frame) } + .debug_str 0 : { *(.debug_str) } + .debug_loc 0 : { *(.debug_loc) } + .debug_macinfo 0 : { *(.debug_macinfo) } + + .debug_weaknames 0 : { *(.debug_weaknames) } + .debug_funcnames 0 : { *(.debug_funcnames) } + .debug_typenames 0 : { *(.debug_typenames) } + .debug_varnames 0 : { *(.debug_varnames) } + + /DISCARD/ : { *(.note.GNU-stack) } + /DISCARD/ : { *(.data .data.* .gnu.linkonce.d.* .sdata*) } + /DISCARD/ : { *(.bss .sbss .dynbss .dynsbss) } +} + + +PHDRS +{ + text PT_LOAD FILEHDR PHDRS FLAGS(5); /* PF_R|PF_X */ + note PT_NOTE FLAGS(4); /* PF_R */ + dynamic PT_DYNAMIC FLAGS(4); /* PF_R */ + eh_frame_hdr PT_GNU_EH_FRAME; +} + + +/* + * This controls what symbols we export from the DSO. + */ +VERSION +{ + VDSO_VERSION_STRING { + global: + __kernel_sigtramp_rt32; + __kernel_restart_syscall32; + local: *; + }; +} diff --git a/arch/parisc/kernel/vdso32/vdso32_wrapper.S b/arch/parisc/kernel/vdso32/vdso32_wrapper.S new file mode 100644 index 000000000000..5ac06093e8ec --- /dev/null +++ b/arch/parisc/kernel/vdso32/vdso32_wrapper.S @@ -0,0 +1,14 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#include <linux/linkage.h> +#include <asm/page.h> + + __PAGE_ALIGNED_DATA + + .globl vdso32_start, vdso32_end + .balign PAGE_SIZE +vdso32_start: + .incbin "arch/parisc/kernel/vdso32/vdso32.so" + .balign PAGE_SIZE +vdso32_end: + + .previous diff --git a/arch/parisc/kernel/vdso64/Makefile b/arch/parisc/kernel/vdso64/Makefile new file mode 100644 index 000000000000..a30f5ec5eb4b --- /dev/null +++ b/arch/parisc/kernel/vdso64/Makefile @@ -0,0 +1,48 @@ +# List of files in the vdso, has to be asm only for now + +obj-vdso64 = note.o sigtramp.o restart_syscall.o + +# Build rules + +targets := $(obj-vdso64) vdso64.so +obj-vdso64 := $(addprefix $(obj)/, $(obj-vdso64)) + + +ccflags-y := -shared -fno-common -fno-builtin +ccflags-y += -nostdlib -Wl,-soname=linux-vdso64.so.1 \ + $(call ld-option, -Wl$(comma)--hash-style=sysv) +asflags-y := -D__VDSO64__ -s + +KBUILD_AFLAGS += -DBUILD_VDSO +KBUILD_CFLAGS += -DBUILD_VDSO -DDISABLE_BRANCH_PROFILING + +VDSO_LIBGCC := $(shell $(CC) -print-libgcc-file-name) + +obj-y += vdso64_wrapper.o +extra-y += vdso64.lds +CPPFLAGS_vdso64.lds += -P -C -U$(ARCH) + +$(obj)/vdso64_wrapper.o : $(obj)/vdso64.so FORCE + +# Force dependency (incbin is bad) +# link rule for the .so file, .lds has to be first +$(obj)/vdso64.so: $(src)/vdso64.lds $(obj-vdso64) $(VDSO_LIBGCC) + $(call if_changed,vdso64ld) + +# assembly rules for the .S files +$(obj-vdso64): %.o: %.S FORCE + $(call if_changed_dep,vdso64as) + +# actual build commands +quiet_cmd_vdso64ld = VDSO64L $@ + cmd_vdso64ld = $(CC) $(c_flags) -Wl,-T $^ -o $@ +quiet_cmd_vdso64as = VDSO64A $@ + cmd_vdso64as = $(CC) $(a_flags) -c -o $@ $< + +# Generate VDSO offsets using helper script +gen-vdsosym := $(srctree)/$(src)/gen_vdso_offsets.sh +quiet_cmd_vdsosym = VDSOSYM $@ + cmd_vdsosym = $(NM) $< | $(gen-vdsosym) | LC_ALL=C sort > $@ + +include/generated/vdso64-offsets.h: $(obj)/vdso64.so FORCE + $(call if_changed,vdsosym) diff --git a/arch/parisc/kernel/vdso64/gen_vdso_offsets.sh b/arch/parisc/kernel/vdso64/gen_vdso_offsets.sh new file mode 100755 index 000000000000..37f05cb38dad --- /dev/null +++ b/arch/parisc/kernel/vdso64/gen_vdso_offsets.sh @@ -0,0 +1,15 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 + +# +# Match symbols in the DSO that look like VDSO_*; produce a header file +# of constant offsets into the shared object. +# +# Doing this inside the Makefile will break the $(filter-out) function, +# causing Kbuild to rebuild the vdso-offsets header file every time. +# +# Inspired by arm64 version. +# + +LC_ALL=C +sed -n 's/\([0-9a-f]*\) . __kernel_\(.*\)/\#define vdso64_offset_\2\t0x\1/p' diff --git a/arch/parisc/kernel/vdso64/note.S b/arch/parisc/kernel/vdso64/note.S new file mode 100644 index 000000000000..bd1fa23597d6 --- /dev/null +++ b/arch/parisc/kernel/vdso64/note.S @@ -0,0 +1,2 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#include "../vdso32/note.S" diff --git a/arch/parisc/kernel/vdso64/restart_syscall.S b/arch/parisc/kernel/vdso64/restart_syscall.S new file mode 100644 index 000000000000..83004451f6b1 --- /dev/null +++ b/arch/parisc/kernel/vdso64/restart_syscall.S @@ -0,0 +1,3 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#include "../vdso32/restart_syscall.S" diff --git a/arch/parisc/kernel/vdso64/sigtramp.S b/arch/parisc/kernel/vdso64/sigtramp.S new file mode 100644 index 000000000000..66a6d2b241e1 --- /dev/null +++ b/arch/parisc/kernel/vdso64/sigtramp.S @@ -0,0 +1,166 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Signal trampolines for 64 bit processes. + * + * Copyright (C) 2006 Randolph Chung <tausq@debian.org> + * Copyright (C) 2018-2022 Helge Deller <deller@gmx.de> + * Copyright (C) 2022 John David Anglin <dave.anglin@bell.net> + */ +#include <asm/unistd.h> +#include <linux/linkage.h> +#include <generated/asm-offsets.h> + + .text + +/* Gdb expects the trampoline is on the stack and the pc is offset from + a 64-byte boundary by 0, 4 or 5 instructions. Since the vdso trampoline + is not on the stack, we need a new variant with different offsets and + data to tell gdb where to find the signal context on the stack. + + Here we put the offset to the context data at the start of the trampoline + region and offset the first trampoline by 2 instructions. Please do + not change the trampoline as the code in gdb depends on the following + instruction sequence exactly. + */ + .align 64 + .word SIGFRAME_CONTEXT_REGS + +/* The nop here is a hack. The dwarf2 unwind routines subtract 1 from + the return address to get an address in the middle of the presumed + call instruction. Since we don't have a call here, we artifically + extend the range covered by the unwind info by adding a nop before + the real start. + */ + nop + + .globl __kernel_sigtramp_rt + .type __kernel_sigtramp_rt, @function +__kernel_sigtramp_rt: + .proc + .callinfo FRAME=ASM_SIGFRAME_SIZE,CALLS,SAVE_RP + .entry + +.Lsigrt_start = . - 4 +0: ldi 0, %r25 /* (in_syscall=0) */ + ldi __NR_rt_sigreturn, %r20 + ble 0x100(%sr2, %r0) + nop + +1: ldi 1, %r25 /* (in_syscall=1) */ + ldi __NR_rt_sigreturn, %r20 + ble 0x100(%sr2, %r0) + nop +.Lsigrt_end: + .exit + .procend + .size __kernel_sigtramp_rt,.-__kernel_sigtramp_rt + + .section .eh_frame,"a",@progbits + +/* This is where the mcontext_t struct can be found on the stack. */ +#define PTREGS SIGFRAME_CONTEXT_REGS /* 64-bit process offset is -720 */ + +/* Register REGNO can be found at offset OFS of the mcontext_t structure. */ + .macro rsave regno,ofs + .byte 0x05 /* DW_CFA_offset_extended */ + .uleb128 \regno; /* regno */ + .uleb128 \ofs /* factored offset */ + .endm + +.Lcie: + .long .Lcie_end - .Lcie_start +.Lcie_start: + .long 0 /* CIE ID */ + .byte 1 /* Version number */ + .stringz "zRS" /* NUL-terminated augmentation string */ + .uleb128 4 /* Code alignment factor */ + .sleb128 8 /* Data alignment factor */ + .byte 61 /* Return address register column, iaoq[0] */ + .uleb128 1 /* Augmentation value length */ + .byte 0x1b /* DW_EH_PE_pcrel | DW_EH_PE_sdata4. */ + .byte 0x0f /* DW_CFA_def_cfa_expresion */ + .uleb128 9f - 1f /* length */ +1: + .byte 0x8e /* DW_OP_breg30 */ + .sleb128 PTREGS +9: + .balign 8 +.Lcie_end: + + .long .Lfde0_end - .Lfde0_start +.Lfde0_start: + .long .Lfde0_start - .Lcie /* CIE pointer. */ + .long .Lsigrt_start - . /* PC start, length */ + .long .Lsigrt_end - .Lsigrt_start + .uleb128 0 /* Augmentation */ + + /* General registers */ + rsave 1, 2 + rsave 2, 3 + rsave 3, 4 + rsave 4, 5 + rsave 5, 6 + rsave 6, 7 + rsave 7, 8 + rsave 8, 9 + rsave 9, 10 + rsave 10, 11 + rsave 11, 12 + rsave 12, 13 + rsave 13, 14 + rsave 14, 15 + rsave 15, 16 + rsave 16, 17 + rsave 17, 18 + rsave 18, 19 + rsave 19, 20 + rsave 20, 21 + rsave 21, 22 + rsave 22, 23 + rsave 23, 24 + rsave 24, 25 + rsave 25, 26 + rsave 26, 27 + rsave 27, 28 + rsave 28, 29 + rsave 29, 30 + rsave 30, 31 + rsave 31, 32 + + /* Floating-point registers */ + rsave 32, 36 + rsave 33, 37 + rsave 34, 38 + rsave 35, 39 + rsave 36, 40 + rsave 37, 41 + rsave 38, 42 + rsave 39, 43 + rsave 40, 44 + rsave 41, 45 + rsave 42, 46 + rsave 43, 47 + rsave 44, 48 + rsave 45, 49 + rsave 46, 50 + rsave 47, 51 + rsave 48, 52 + rsave 49, 53 + rsave 50, 54 + rsave 51, 55 + rsave 52, 56 + rsave 53, 57 + rsave 54, 58 + rsave 55, 59 + rsave 56, 60 + rsave 57, 61 + rsave 58, 62 + rsave 59, 63 + + /* SAR register */ + rsave 60, 67 + + /* iaoq[0] return address register */ + rsave 61, 65 + .balign 8 +.Lfde0_end: diff --git a/arch/parisc/kernel/vdso64/vdso64.lds.S b/arch/parisc/kernel/vdso64/vdso64.lds.S new file mode 100644 index 000000000000..de1fb4b19286 --- /dev/null +++ b/arch/parisc/kernel/vdso64/vdso64.lds.S @@ -0,0 +1,109 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * This is the infamous ld script for the 64 bits vdso library + */ +#include <asm/vdso.h> + +/* Default link addresses for the vDSOs */ +OUTPUT_FORMAT("elf64-hppa-linux") +OUTPUT_ARCH(hppa:hppa2.0w) +ENTRY(_start) + +SECTIONS +{ + . = VDSO_LBASE + SIZEOF_HEADERS; + .hash : { *(.hash) } :text + .gnu.hash : { *(.gnu.hash) } + .dynsym : { *(.dynsym) } + .dynstr : { *(.dynstr) } + .gnu.version : { *(.gnu.version) } + .gnu.version_d : { *(.gnu.version_d) } + .gnu.version_r : { *(.gnu.version_r) } + + .note : { *(.note.*) } :text :note + + . = ALIGN (16); + .text : + { + *(.text .stub .text.* .gnu.linkonce.t.*) + } + PROVIDE (__etext = .); + PROVIDE (_etext = .); + PROVIDE (etext = .); + + /* Other stuff is appended to the text segment: */ + .rodata : { *(.rodata .rodata.* .gnu.linkonce.r.*) } + .rodata1 : { *(.rodata1) } + + .eh_frame_hdr : { *(.eh_frame_hdr) } :text :eh_frame_hdr + .eh_frame : { KEEP (*(.eh_frame)) } :text + .gcc_except_table : { *(.gcc_except_table) } + .fixup : { *(.fixup) } + + .dynamic : { *(.dynamic) } :text :dynamic + .plt : { *(.plt) } + .got : { *(.got) } + + _end = .; + __end = .; + PROVIDE (end = .); + + + /* Stabs debugging sections are here too + */ + .stab 0 : { *(.stab) } + .stabstr 0 : { *(.stabstr) } + .stab.excl 0 : { *(.stab.excl) } + .stab.exclstr 0 : { *(.stab.exclstr) } + .stab.index 0 : { *(.stab.index) } + .stab.indexstr 0 : { *(.stab.indexstr) } + .comment 0 : { *(.comment) } + .debug 0 : { *(.debug) } + .line 0 : { *(.line) } + + .debug_srcinfo 0 : { *(.debug_srcinfo) } + .debug_sfnames 0 : { *(.debug_sfnames) } + + .debug_aranges 0 : { *(.debug_aranges) } + .debug_pubnames 0 : { *(.debug_pubnames) } + + .debug_info 0 : { *(.debug_info .gnu.linkonce.wi.*) } + .debug_abbrev 0 : { *(.debug_abbrev) } + .debug_line 0 : { *(.debug_line) } + .debug_frame 0 : { *(.debug_frame) } + .debug_str 0 : { *(.debug_str) } + .debug_loc 0 : { *(.debug_loc) } + .debug_macinfo 0 : { *(.debug_macinfo) } + + .debug_weaknames 0 : { *(.debug_weaknames) } + .debug_funcnames 0 : { *(.debug_funcnames) } + .debug_typenames 0 : { *(.debug_typenames) } + .debug_varnames 0 : { *(.debug_varnames) } + + /DISCARD/ : { *(.note.GNU-stack) } + /DISCARD/ : { *(.data .data.* .gnu.linkonce.d.* .sdata*) } + /DISCARD/ : { *(.bss .sbss .dynbss .dynsbss) } +} + + +PHDRS +{ + text PT_LOAD FILEHDR PHDRS FLAGS(5); /* PF_R|PF_X */ + note PT_NOTE FLAGS(4); /* PF_R */ + dynamic PT_DYNAMIC FLAGS(4); /* PF_R */ + eh_frame_hdr PT_GNU_EH_FRAME; +} + + +/* + * This controls what symbols we export from the DSO. + */ +VERSION +{ + VDSO_VERSION_STRING { + global: + __kernel_sigtramp_rt64; + __kernel_restart_syscall64; + local: *; + }; +} diff --git a/arch/parisc/kernel/vdso64/vdso64_wrapper.S b/arch/parisc/kernel/vdso64/vdso64_wrapper.S new file mode 100644 index 000000000000..66f929482d3d --- /dev/null +++ b/arch/parisc/kernel/vdso64/vdso64_wrapper.S @@ -0,0 +1,14 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#include <linux/linkage.h> +#include <asm/page.h> + + __PAGE_ALIGNED_DATA + + .globl vdso64_start, vdso64_end + .balign PAGE_SIZE +vdso64_start: + .incbin "arch/parisc/kernel/vdso64/vdso64.so" + .balign PAGE_SIZE +vdso64_end: + + .previous |