1. Add loongarch64_init() implementation, do all necessary machine-specific setup, which will be called multiple times during initialization. 2. Add the implementation of the vtop command, which is used to convert a virtual address to a physical address. When entering the crash command line, the corresponding symbols in the kernel will be read, and at the same time, the conversion of virtual and real addresses will also be used, so the vtop command is a prerequisite for entering the crash command line. 3. Add loongarch64_get_smp_cpus() implementation, get the number of online cpus. 4. Add loongarch64_get_page_size() implementation, get page size. 5. Add to get processor speed. Obtain the processor speed from the kernel symbol "cpu_clock_freq". 6. Add loongarch64_verify_symbol() implementation, accept or reject a symbol from the kernel namelist. With this patch, we can enter crash command line. Tested on Loongson-3C5000 platform. For help, type "help". Type "apropos word" to search for commands related to "word"... KERNEL: /usr/lib/debug/lib/modules/5.10.0-60.103.0.130.oe2203.loongarch64/vmlinux DUMPFILE: /proc/kcore CPUS: 16 DATE: Mon Aug 21 14:33:19 CST 2023 UPTIME: 05:01:34 LOAD AVERAGE: 0.43, 0.11, 0.17 TASKS: 265 NODENAME: localhost.localdomain RELEASE: 5.10.0-60.103.0.130.oe2203.loongarch64 VERSION: #1 SMP Fri Jul 21 12:48:08 UTC 2023 MACHINE: loongarch64 (2200 Mhz) MEMORY: 64 GB PID: 114499 COMMAND: "crash" TASK: 900000009676ff00 [THREAD_INFO: 90000000981a8000] CPU: 12 STATE: TASK_RUNNING (ACTIVE) Co-developed-by: Youling Tang <tangyouling@xxxxxxxxxxx> Signed-off-by: Youling Tang <tangyouling@xxxxxxxxxxx> Signed-off-by: Ming Wang <wangming01@xxxxxxxxxxx> --- loongarch64.c | 313 ++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 313 insertions(+) diff --git a/loongarch64.c b/loongarch64.c index 0a6a502..6018bed 100644 --- a/loongarch64.c +++ b/loongarch64.c @@ -40,14 +40,327 @@ struct loongarch64_unwind_frame { unsigned long ra; }; +static int loongarch64_pgd_vtop(ulong *pgd, ulong vaddr, + physaddr_t *paddr, int verbose); +static int loongarch64_uvtop(struct task_context *tc, ulong vaddr, + physaddr_t *paddr, int verbose); +static int loongarch64_kvtop(struct task_context *tc, ulong kvaddr, + physaddr_t *paddr, int verbose); + +/* + * 3 Levels paging PAGE_SIZE=16KB + * PGD | PMD | PTE | OFFSET | + * 11 | 11 | 11 | 14 | + */ +/* From arch/loongarch/include/asm/pgtable{,-64}.h */ +typedef struct { ulong pgd; } pgd_t; +typedef struct { ulong pmd; } pmd_t; +typedef struct { ulong pte; } pte_t; + +#define TASK_SIZE64 (1UL << 40) + +#define PMD_SHIFT (PAGESHIFT() + (PAGESHIFT() - 3)) +#define PMD_SIZE (1UL << PMD_SHIFT) +#define PMD_MASK (~(PMD_SIZE - 1)) + +#define PGDIR_SHIFT (PMD_SHIFT + (PAGESHIFT() - 3)) +#define PGDIR_SIZE (1UL << PGDIR_SHIFT) +#define PGDIR_MASK (~(PGDIR_SIZE - 1)) + +#define PTRS_PER_PTE (1UL << (PAGESHIFT() - 3)) +#define PTRS_PER_PMD PTRS_PER_PTE +#define PTRS_PER_PGD PTRS_PER_PTE +#define USER_PTRS_PER_PGD ((TASK_SIZE64 / PGDIR_SIZE)?(TASK_SIZE64 / PGDIR_SIZE) : 1) + +#define pte_index(addr) (((addr) >> PAGESHIFT()) & (PTRS_PER_PTE - 1)) +#define pmd_index(addr) (((addr) >> PMD_SHIFT) & (PTRS_PER_PMD - 1)) +#define pgd_index(addr) (((addr) >> PGDIR_SHIFT) & (PTRS_PER_PGD - 1)) + +#define LOONGARCH64_CPU_RIXI (1UL << 23) /* CPU has TLB Read/eXec Inhibit */ + +static struct machine_specific loongarch64_machine_specific = { 0 }; + +/* + * Identify and print the segment name to which the virtual address belongs + */ +static void +get_segment_name(ulong vaddr, int verbose) +{ + const char * segment; + + if (verbose) { + if (vaddr < 0x4000000000000000lu) + segment = "xuvrange"; + else if (vaddr < 0x8000000000000000lu) + segment = "xsprange"; + else if (vaddr < 0xc000000000000000lu) + segment = "xkprange"; + else + segment = "xkvrange"; + + fprintf(fp, "SEGMENT: %s\n", segment); + } +} + +/* + * Virtual to physical memory translation. This function will be called + * by both loongarch64_kvtop and loongarch64_uvtop. + */ +static int +loongarch64_pgd_vtop(ulong *pgd, ulong vaddr, physaddr_t *paddr, int verbose) +{ + ulong *pgd_ptr, pgd_val; + ulong *pmd_ptr, pmd_val; + ulong *pte_ptr, pte_val; + + get_segment_name(vaddr, verbose); + + if (IS_XKPRANGE(vaddr)) { + *paddr = VTOP(vaddr); + return TRUE; + } + + if (verbose) + fprintf(fp, "PAGE DIRECTORY: %016lx\n", (ulong)pgd); + + pgd_ptr = pgd + pgd_index(vaddr); + FILL_PGD(PAGEBASE(pgd), KVADDR, PAGESIZE()); + pgd_val = ULONG(machdep->pgd + PAGEOFFSET(pgd_ptr)); + if (verbose) + fprintf(fp, " PGD: %16lx => %16lx\n", (ulong)pgd_ptr, pgd_val); + if (!pgd_val) + goto no_page; + + pmd_ptr = (ulong *)(VTOP(pgd_val) + sizeof(pmd_t) * pmd_index(vaddr)); + FILL_PMD(PAGEBASE(pmd_ptr), PHYSADDR, PAGESIZE()); + pmd_val = ULONG(machdep->pmd + PAGEOFFSET(pmd_ptr)); + if (verbose) + fprintf(fp, " PMD: %016lx => %016lx\n", (ulong)pmd_ptr, pmd_val); + if (!pmd_val) + goto no_page; + + pte_ptr = (ulong *)(VTOP(pmd_val) + sizeof(pte_t) * pte_index(vaddr)); + FILL_PTBL(PAGEBASE(pte_ptr), PHYSADDR, PAGESIZE()); + pte_val = ULONG(machdep->ptbl + PAGEOFFSET(pte_ptr)); + if (verbose) + fprintf(fp, " PTE: %016lx => %016lx\n", (ulong)pte_ptr, pte_val); + if (!pte_val) + goto no_page; + + return TRUE; +no_page: + fprintf(fp, "invalid\n"); + return FALSE; +} + +/* Translates a user virtual address to its physical address. cmd_vtop() sets + * the verbose flag so that the pte translation gets displayed; all other + * callers quietly accept the translation. + */ +static int +loongarch64_uvtop(struct task_context *tc, ulong vaddr, physaddr_t *paddr, int verbose) +{ + ulong mm, active_mm; + ulong *pgd; + + if (!tc) + error(FATAL, "current context invalid\n"); + + *paddr = 0; + + if (is_kernel_thread(tc->task) && IS_KVADDR(vaddr)) { + readmem(tc->task + OFFSET(task_struct_active_mm), + KVADDR, &active_mm, sizeof(void *), + "task active_mm contents", FAULT_ON_ERROR); + + if (!active_mm) + error(FATAL, + "no active_mm for this kernel thread\n"); + + readmem(active_mm + OFFSET(mm_struct_pgd), + KVADDR, &pgd, sizeof(long), + "mm_struct pgd", FAULT_ON_ERROR); + } else { + if ((mm = task_mm(tc->task, TRUE))) + pgd = ULONG_PTR(tt->mm_struct + OFFSET(mm_struct_pgd)); + else + readmem(tc->mm_struct + OFFSET(mm_struct_pgd), + KVADDR, &pgd, sizeof(long), "mm_struct pgd", + FAULT_ON_ERROR); + } + + return loongarch64_pgd_vtop(pgd, vaddr, paddr, verbose);; +} + +/* Translates a user virtual address to its physical address. cmd_vtop() sets + * the verbose flag so that the pte translation gets displayed; all other + * callers quietly accept the translation. + */ +static int +loongarch64_kvtop(struct task_context *tc, ulong kvaddr, physaddr_t *paddr, int verbose) +{ + if (!IS_KVADDR(kvaddr)) + return FALSE; + + if (!verbose) { + if (IS_XKPRANGE(kvaddr)) { + *paddr = VTOP(kvaddr); + return TRUE; + } + } + + return loongarch64_pgd_vtop((ulong *)vt->kernel_pgd[0], kvaddr, paddr, + verbose); +} + +/* + * Accept or reject a symbol from the kernel namelist. + */ +static int +loongarch64_verify_symbol(const char *name, ulong value, char type) +{ + if (!strncmp(name, ".L", 2) || !strncmp(name, "L0", 2)) + return FALSE; + + if (CRASHDEBUG(8) && name && strlen(name)) + fprintf(fp, "%08lx %s\n", value, name); + + if (STREQ(name, "_text") || STREQ(name, "_stext")) + machdep->flags |= KSYMS_START; + + return (name && strlen(name) && (machdep->flags & KSYMS_START) && + !STRNEQ(name, "__func__.") && !STRNEQ(name, "__crc_")); +} + +/* + * Override smp_num_cpus if possible and necessary. + */ +static int +loongarch64_get_smp_cpus(void) +{ + return (get_cpus_online() > 0) ? get_cpus_online() : kt->cpus; +} + +static ulong +loongarch64_get_page_size(void) +{ + return memory_page_size(); +} + +/* + * Determine where vmalloc'd memory starts. + */ +static ulong +loongarch64_vmalloc_start(void) +{ +} + +/* + * Calculate and return the speed of the processor. + */ +static ulong +loongarch64_processor_speed(void) +{ + unsigned long cpu_hz = 0; + + if (machdep->mhz) + return (machdep->mhz); + + if (symbol_exists("cpu_clock_freq")) { + get_symbol_data("cpu_clock_freq", sizeof(int), &cpu_hz); + if (cpu_hz) + return(machdep->mhz = cpu_hz/1000000); + } + + return 0; +} + +/* + * Checks whether given task is valid task address. + */ +static int +loongarch64_is_task_addr(ulong task) +{ + if (tt->flags & THREAD_INFO) + return IS_KVADDR(task); + + return (IS_KVADDR(task) && ALIGNED_STACK_OFFSET(task) == 0); +} + void loongarch64_dump_machdep_table(ulong arg) { } +static void +pt_level_alloc(char **lvl, char *name) +{ + size_t sz = PAGESIZE(); + void *pointer = malloc(sz); + + if (!pointer) + error(FATAL, name); + *lvl = pointer; +} + void loongarch64_init(int when) { + switch (when) { + case SETUP_ENV: + machdep->process_elf_notes = process_elf64_notes; + break; + + case PRE_SYMTAB: + machdep->verify_symbol = loongarch64_verify_symbol; + machdep->machspec = &loongarch64_machine_specific; + if (pc->flags & KERNEL_DEBUG_QUERY) + return; + machdep->last_pgd_read = 0; + machdep->last_pmd_read = 0; + machdep->last_ptbl_read = 0; + machdep->verify_paddr = generic_verify_paddr; + machdep->ptrs_per_pgd = PTRS_PER_PGD; + break; + + case PRE_GDB: + machdep->pagesize = loongarch64_get_page_size(); + machdep->pageshift = ffs(machdep->pagesize) - 1; + machdep->pageoffset = machdep->pagesize - 1; + machdep->pagemask = ~((ulonglong)machdep->pageoffset); + if (machdep->pagesize >= 16384) + machdep->stacksize = machdep->pagesize; + else + machdep->stacksize = machdep->pagesize * 2; + + pt_level_alloc(&machdep->pgd, "cannot malloc pgd space."); + pt_level_alloc(&machdep->pmd, "cannot malloc pmd space."); + pt_level_alloc(&machdep->ptbl, "cannot malloc ptbl space."); + machdep->kvbase = 0x8000000000000000lu; + machdep->identity_map_base = machdep->kvbase; + machdep->is_kvaddr = generic_is_kvaddr; + machdep->is_uvaddr = generic_is_uvaddr; + machdep->uvtop = loongarch64_uvtop; + machdep->kvtop = loongarch64_kvtop; + machdep->vmalloc_start = loongarch64_vmalloc_start; + machdep->processor_speed = loongarch64_processor_speed; + machdep->get_stackbase = generic_get_stackbase; + machdep->get_stacktop = generic_get_stacktop; + machdep->memory_size = generic_memory_size; + machdep->is_task_addr = loongarch64_is_task_addr; + machdep->get_smp_cpus = loongarch64_get_smp_cpus; + machdep->dis_filter = generic_dis_filter; + machdep->value_to_symbol = generic_machdep_value_to_symbol; + machdep->init_kernel_pgd = NULL; + break; + + case POST_GDB: + machdep->section_size_bits = _SECTION_SIZE_BITS; + machdep->max_physmem_bits = _MAX_PHYSMEM_BITS; + + case POST_VM: + break; + } } void -- 2.39.2 -- Crash-utility mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxxxxxx %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s Contribution Guidelines: https://github.com/crash-utility/crash/wiki