From patchwork Fri May 27 05:48:59 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Rashmica Gupta X-Patchwork-Id: 626989 Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Received: from lists.ozlabs.org (lists.ozlabs.org [103.22.144.68]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 3rGFVR66y5z9t3c for ; Fri, 27 May 2016 15:50:39 +1000 (AEST) Authentication-Results: ozlabs.org; dkim=fail reason="signature verification failed" (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b=kVixFl+q; dkim-atps=neutral Received: from ozlabs.org (lists.ozlabs.org [IPv6:2401:3900:2:1::3]) by lists.ozlabs.org (Postfix) with ESMTP id 3rGFVR52TMzDqbW for ; Fri, 27 May 2016 15:50:39 +1000 (AEST) Authentication-Results: lists.ozlabs.org; dkim=fail reason="signature verification failed" (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b=kVixFl+q; dkim-atps=neutral X-Original-To: linuxppc-dev@lists.ozlabs.org Delivered-To: linuxppc-dev@lists.ozlabs.org Received: from mail-pf0-x242.google.com (mail-pf0-x242.google.com [IPv6:2607:f8b0:400e:c00::242]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 3rGFSs5jWHzDqCT for ; Fri, 27 May 2016 15:49:17 +1000 (AEST) Authentication-Results: lists.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=gmail.com header.i=@gmail.com header.b=kVixFl+q; dkim-atps=neutral Received: by mail-pf0-x242.google.com with SMTP id f144so4800095pfa.2 for ; Thu, 26 May 2016 22:49:17 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=from:to:cc:subject:date:message-id; bh=HVT7d5aHpIk0xzf/Q8yHM7UIhX1qQuzZjLjd6BWLvhk=; b=kVixFl+qG98bCZq2+ylEgRgmy62+VZANu3qslqjA2jbpkoWwJ34B2qSWUzvE8lhQ23 CwjxY+xUWxBGam0n8OQDV+srwuIh4Pj4BuH8iKlE1srlgAajJnPI5LKZ2UGtCXdTUJD9 AL11E5m+rbWA9choTDV7RE/VDfBhDZvH8L98yvYGELi2v4Hof0yuQ4ZfYjBadFd5+4bO HEaoWdNYrn95pZwRdXXnwi4KCn5BDOwegBKLpTt3RhsEIeLO765JvVE2sFJMG1FJaiZB RnACX7jd6RddckC81FjZyl+gNz4ucMVsgoo+DscyrSvMY0c36IdgHL7vLYGbHMDSzldJ Xg6Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:from:to:cc:subject:date:message-id; bh=HVT7d5aHpIk0xzf/Q8yHM7UIhX1qQuzZjLjd6BWLvhk=; b=JJOWAx14kqqM+edsDkCNz/TgYbOrfqnjZOe7GcaiRkyab2B4DXcxyeR/IJhc6VQ5Bv UOVWRuxFl2k+AuIDKRE3aI4o9Zdiux+G1Nui9rNwweEZyuG+c8ir+fhchMIcj7hgdeO5 NcedBGvnzLB8Vplplc8uFFL+K8Ssx1pWbXAL5t97HlMnIVqer6hIRB8dONXnJ/5ZwY8Y JIj0Vy5Vb1zVECNKT2cAtgwY4qinSJcVKLjrRrIDaxIG3bp3osvQwPll4rzdnfs7vhVz 19WzSabLVHalc9Ga+F4cWfUrJXwx8Y5tkqOtcVryYBCm0/hV9sESIm6Noha+HI/L1kgh D+BA== X-Gm-Message-State: ALyK8tLDys3Tqf4ZydUG/lpB2VlFPHka34COm3BSTGfPmk2+QZSu+on7A6W2WR/MIB4JIw== X-Received: by 10.98.100.83 with SMTP id y80mr19789717pfb.84.1464328155540; Thu, 26 May 2016 22:49:15 -0700 (PDT) Received: from circle.ozlabs.ibm.com ([122.99.82.10]) by smtp.gmail.com with ESMTPSA id vy8sm24682185pab.22.2016.05.26.22.49.12 (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Thu, 26 May 2016 22:49:14 -0700 (PDT) From: Rashmica Gupta To: linuxppc-dev@lists.ozlabs.org, mpe@ellerman.id.au Subject: [v3 PATCH 1/2] powerpc/mm: Dump linux pagetables Date: Fri, 27 May 2016 15:48:59 +1000 Message-Id: <1464328140-31178-1-git-send-email-rashmicy@gmail.com> X-Mailer: git-send-email 2.5.0 X-BeenThere: linuxppc-dev@lists.ozlabs.org X-Mailman-Version: 2.1.22 Precedence: list List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Errors-To: linuxppc-dev-bounces+patchwork-incoming=ozlabs.org@lists.ozlabs.org Sender: "Linuxppc-dev" Useful to be able to dump the kernels page tables to check permissions and memory types - derived from arm64's implementation. Add a debugfs file to check the page tables. To use this the PPC_PTDUMP config option must be selected. Signed-off-by: Rashmica Gupta --- v2 -> v3: Changed to be compatible with P9. arch/powerpc/Kconfig.debug | 12 + arch/powerpc/mm/Makefile | 1 + arch/powerpc/mm/dump_linuxpagetables.c | 440 +++++++++++++++++++++++++++++++++ 3 files changed, 453 insertions(+) create mode 100644 arch/powerpc/mm/dump_linuxpagetables.c diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug index 2be221874de8..d19f496a7e12 100644 --- a/arch/powerpc/Kconfig.debug +++ b/arch/powerpc/Kconfig.debug @@ -350,4 +350,16 @@ config FAIL_IOMMU If you are unsure, say N. +config PPC_PTDUMP + bool "Export kernel pagetable layout to userspace via debugfs" + depends on DEBUG_KERNEL + select DEBUG_FS + help + This option exports the state of the kernel pagetables to a + debugfs file. This is only useful for kernel developers who are + working in architecture specific areas of the kernel - probably + not a good idea to enable this feature in a production kernel. + + If you are unsure, say N. + endmenu diff --git a/arch/powerpc/mm/Makefile b/arch/powerpc/mm/Makefile index f2cea6d5e764..d3dfa02082b1 100644 --- a/arch/powerpc/mm/Makefile +++ b/arch/powerpc/mm/Makefile @@ -43,3 +43,4 @@ obj-$(CONFIG_NOT_COHERENT_CACHE) += dma-noncoherent.o obj-$(CONFIG_HIGHMEM) += highmem.o obj-$(CONFIG_PPC_COPRO_BASE) += copro_fault.o obj-$(CONFIG_SPAPR_TCE_IOMMU) += mmu_context_iommu.o +obj-$(CONFIG_PPC_PTDUMP) += dump_linuxpagetables.o diff --git a/arch/powerpc/mm/dump_linuxpagetables.c b/arch/powerpc/mm/dump_linuxpagetables.c new file mode 100644 index 000000000000..d242bc79ae64 --- /dev/null +++ b/arch/powerpc/mm/dump_linuxpagetables.c @@ -0,0 +1,440 @@ +/* + * Copyright 2016, Rashmica Gupta, IBM Corp. + * + * This traverses the kernel pagetables and dumps the + * information about the used sections of memory to + * /sys/kernel/debug/kernel_pagetables. + * + * Derived from the arm64 implementation: + * Copyright (c) 2014, The Linux Foundation, Laura Abbott. + * (C) Copyright 2008 Intel Corporation, Arjan van de Ven. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; version 2 + * of the License. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* + * To visualise what is happening, + * + * - PTRS_PER_P** = how many entries there are in the corresponding P** + * - P**_SHIFT = how many bits of the address we use to index into the + * corresponding P** + * - P**_SIZE is how much memory we can access through the table - not the + * size of the table itself. + * P**={PGD, PUD, PMD, PTE} + * + * + * Each entry of the PGD points to a PUD. Each entry of a PUD points to a + * PMD. Each entry of a PMD points to a PTE. And every PTE entry points to + * a page. + * + * In the case where there are only 3 levels, the PUD is folded into the + * PGD: every PUD has only one entry which points to the PMD. + * + * The page dumper groups page table entries of the same type into a single + * description. It uses pg_state to track the range information while + * iterating over the PTE entries. When the continuity is broken it then + * dumps out a description of the range - ie PTEs that are virtually contiguous + * with the same PTE flags are chunked together. This is to make it clear how + * different areas of the kernel virtual memory are used. + * + */ +struct pg_state { + struct seq_file *seq; + const struct addr_marker *marker; + unsigned long start_address; + unsigned int level; + u64 current_flags; +}; + +struct addr_marker { + unsigned long start_address; + const char *name; +}; + +static struct addr_marker address_markers[] = { + { 0, "Start of kernel VM" }, + { 0, "vmalloc() Area" }, + { 0, "vmalloc() End" }, + { 0, "isa I/O start" }, + { 0, "isa I/O end" }, + { 0, "phb I/O start" }, + { 0, "phb I/O end" }, + { 0, "I/O remap start" }, + { 0, "I/O remap end" }, + { 0, "vmemmap start" }, + { -1, NULL }, +}; + +struct flag_info { + u64 mask; + u64 val; + const char *set; + const char *clear; + bool is_val; + int shift; +}; + +static const struct flag_info flag_array[] = { + { +#ifdef CONFIG_PPC_STD_MMU_64 + .mask = _PAGE_PRIVILEGED, + .val = 0, +#else + .mask = _PAGE_USER, + .val = _PAGE_USER, +#endif + .set = "user", + .clear = " ", + }, { + .mask = _PAGE_RW, + .val = _PAGE_RW, + .set = "rw", + .clear = "ro", + }, { + .mask = _PAGE_EXEC, + .val = _PAGE_EXEC, + .set = " X ", + .clear = " ", + }, { + .mask = _PAGE_PTE, + .val = _PAGE_PTE, + .set = "pte", + .clear = " ", + }, { + .mask = _PAGE_PRESENT, + .val = _PAGE_PRESENT, + .set = "present", + .clear = " ", + }, { +#ifdef CONFIG_PPC_STD_MMU_64 + .mask = H_PAGE_HASHPTE, + .val = H_PAGE_HASHPTE, +#else + .mask = _PAGE_HASHPTE, + .val = _PAGE_HASHPTE, +#endif + .set = "hpte", + .clear = " ", + }, { +#ifndef CONFIG_PPC_STD_MMU_64 + .mask = _PAGE_GUARDED, + .val = _PAGE_GUARDED, + .set = "guarded", + .clear = " ", + }, { +#endif + .mask = _PAGE_DIRTY, + .val = _PAGE_DIRTY, + .set = "dirty", + .clear = " ", + }, { + .mask = _PAGE_ACCESSED, + .val = _PAGE_ACCESSED, + .set = "accessed", + .clear = " ", + }, { +#ifndef CONFIG_PPC_STD_MMU_64 + .mask = _PAGE_WRITETHRU, + .val = _PAGE_WRITETHRU, + .set = "write through", + .clear = " ", + }, { +#endif + .mask = _PAGE_NO_CACHE, + .val = _PAGE_NO_CACHE, + .set = "no cache", + .clear = " ", + }, { + .mask = H_PAGE_BUSY, + .val = H_PAGE_BUSY, + .set = "busy", + }, { +#ifdef CONFIG_PPC_64K_PAGES + .mask = H_PAGE_COMBO, + .val = H_PAGE_COMBO, + .set = "combo", + }, { + .mask = H_PAGE_4K_PFN, + .val = H_PAGE_4K_PFN, + .set = "4K_pfn", + }, { +#endif + .mask = H_PAGE_F_GIX, + .val = H_PAGE_F_GIX, + .set = "f_gix", + .is_val = true, + .shift = H_PAGE_F_GIX_SHIFT, + }, { + .mask = H_PAGE_F_SECOND, + .val = H_PAGE_F_SECOND, + .set = "f_second", + }, { + .mask = _PAGE_SPECIAL, + .val = _PAGE_SPECIAL, + .set = "special", + } +}; + +struct pgtable_level { + const struct flag_info *flag; + size_t num; + u64 mask; +}; + +static struct pgtable_level pg_level[] = { + { + }, { /* pgd */ + .flag = flag_array, + .num = ARRAY_SIZE(flag_array), + }, { /* pud */ + .flag = flag_array, + .num = ARRAY_SIZE(flag_array), + }, { /* pmd */ + .flag = flag_array, + .num = ARRAY_SIZE(flag_array), + }, { /* pte */ + .flag = flag_array, + .num = ARRAY_SIZE(flag_array), + }, +}; + +static void dump_flag_info(struct pg_state *st, const struct flag_info + *flag, u64 pte, int num) +{ + unsigned int i; + + for (i = 0; i < num; i++, flag++) { + const char *s = NULL; + u64 val; + + /* flag not defined so don't check it */ + if (flag->mask == 0) + continue; + /* Some 'flags' are actually values */ + if (flag->is_val) { + val = pte & flag->val; + if (flag->shift) + val = val >> flag->shift; + seq_printf(st->seq, " %s:%llx", flag->set, val); + } else { + if ((pte & flag->mask) == flag->val) + s = flag->set; + else + s = flag->clear; + if (s) + seq_printf(st->seq, " %s", s); + } + st->current_flags &= ~flag->mask; + } + if (st->current_flags != 0) + seq_printf(st->seq, " unknown flags:%llx", st->current_flags); +} + +static void dump_addr(struct pg_state *st, unsigned long addr) +{ + static const char units[] = "KMGTPE"; + const char *unit = units; + unsigned long delta; + + seq_printf(st->seq, "0x%016lx-0x%016lx ", st->start_address, addr-1); + delta = (addr - st->start_address) >> 10; + /* Work out what appropriate unit to use */ + while (!(delta & 1023) && unit[1]) { + delta >>= 10; + unit++; + } + seq_printf(st->seq, "%9lu%c", delta, *unit); + +} + +static void note_page(struct pg_state *st, unsigned long addr, + unsigned int level, u64 val) +{ + u64 flag = val & pg_level[level].mask; + /* At first no level is set */ + if (!st->level) { + st->level = level; + st->current_flags = flag; + st->start_address = addr; + seq_printf(st->seq, "---[ %s ]---\n", st->marker->name); + /* + * Dump the section of virtual memory when: + * - the PTE flags from one entry to the next differs. + * - we change levels in the tree. + * - the address is in a different section of memory and is thus + * used for a different purpose, regardless of the flags. + */ + } else if (flag != st->current_flags || level != st->level || + addr >= st->marker[1].start_address) { + + /* Check the PTE flags */ + if (st->current_flags) { + dump_addr(st, addr); + + /* Dump all the flags */ + if (pg_level[st->level].flag) + dump_flag_info(st, pg_level[st->level].flag, + st->current_flags, + pg_level[st->level].num); + + seq_puts(st->seq, "\n"); + } + + /* + * Address indicates we have passed the end of the + * current section of virtual memory + */ + while (addr >= st->marker[1].start_address) { + st->marker++; + seq_printf(st->seq, "---[ %s ]---\n", st->marker->name); + } + st->start_address = addr; + st->current_flags = flag; + st->level = level; + } +} + +static void walk_pte(struct pg_state *st, pmd_t *pmd, unsigned long start) +{ + pte_t *pte = pte_offset_kernel(pmd, 0); + unsigned long addr; + unsigned int i; + + for (i = 0; i < PTRS_PER_PTE; i++, pte++) { + addr = start + i * PAGE_SIZE; + note_page(st, addr, 4, pte_val(*pte)); + + } +} + +static void walk_pmd(struct pg_state *st, pud_t *pud, unsigned long start) +{ + pmd_t *pmd = pmd_offset(pud, 0); + unsigned long addr; + unsigned int i; + + for (i = 0; i < PTRS_PER_PMD; i++, pmd++) { + addr = start + i * PMD_SIZE; + if (!pmd_none(*pmd)) + /* pmd exists */ + walk_pte(st, pmd, addr); + else + note_page(st, addr, 3, pmd_val(*pmd)); + } +} + +static void walk_pud(struct pg_state *st, pgd_t *pgd, unsigned long start) +{ + pud_t *pud = pud_offset(pgd, 0); + unsigned long addr; + unsigned int i; + + for (i = 0; i < PTRS_PER_PUD; i++, pud++) { + addr = start + i * PUD_SIZE; + if (!pud_none(*pud)) + /* pud exists */ + walk_pmd(st, pud, addr); + else + note_page(st, addr, 2, pud_val(*pud)); + } +} + +static void walk_pagetables(struct pg_state *st) +{ + pgd_t *pgd = pgd_offset_k(0UL); + unsigned int i; + unsigned long addr; + + /* + * Traverse the linux pagetable structure and dump pages that are in + * the hash pagetable. + */ + for (i = 0; i < PTRS_PER_PGD; i++, pgd++) { + addr = KERN_VIRT_START + i * PGDIR_SIZE; + if (!pgd_none(*pgd)) + /* pgd exists */ + walk_pud(st, pgd, addr); + else + note_page(st, addr, 1, pgd_val(*pgd)); + } +} + +static void populate_markers(void) +{ + address_markers[0].start_address = PAGE_OFFSET; + address_markers[1].start_address = VMALLOC_START; + address_markers[2].start_address = VMALLOC_END; + address_markers[3].start_address = ISA_IO_BASE; + address_markers[4].start_address = ISA_IO_END; + address_markers[5].start_address = PHB_IO_BASE; + address_markers[6].start_address = PHB_IO_END; + address_markers[7].start_address = IOREMAP_BASE; + address_markers[8].start_address = IOREMAP_END; +#ifdef CONFIG_PPC_STD_MMU_64 + address_markers[9].start_address = H_VMEMMAP_BASE; +#else + address_markers[9].start_address = VMEMMAP_BASE; +#endif +} + +static int ptdump_show(struct seq_file *m, void *v) +{ + struct pg_state st = { + .seq = m, + .start_address = KERN_VIRT_START, + .marker = address_markers, + }; + /* Traverse kernel page tables */ + walk_pagetables(&st); + note_page(&st, 0, 0, 0); + return 0; +} + + +static int ptdump_open(struct inode *inode, struct file *file) +{ + return single_open(file, ptdump_show, NULL); +} + +static const struct file_operations ptdump_fops = { + .open = ptdump_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static void build_pgtable_complete_mask(void) +{ + unsigned int i, j; + + for (i = 0; i < ARRAY_SIZE(pg_level); i++) + if (pg_level[i].flag) + for (j = 0; j < pg_level[i].num; j++) + pg_level[i].mask |= pg_level[i].flag[j].mask; +} + +static int ptdump_init(void) +{ + struct dentry *debugfs_file; + + populate_markers(); + build_pgtable_complete_mask(); + debugfs_file = debugfs_create_file("kernel_pagetables", 0400, NULL, + NULL, &ptdump_fops); + return debugfs_file ? 0 : -ENOMEM; +} +device_initcall(ptdump_init);