Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751154AbVKJQoV (ORCPT ); Thu, 10 Nov 2005 11:44:21 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751162AbVKJQoU (ORCPT ); Thu, 10 Nov 2005 11:44:20 -0500 Received: from fed1rmmtao12.cox.net ([68.230.241.27]:41393 "EHLO fed1rmmtao12.cox.net") by vger.kernel.org with ESMTP id S1751154AbVKJQoS (ORCPT ); Thu, 10 Nov 2005 11:44:18 -0500 From: Tom Rini To: Andrew Morton Cc: Tom Rini , lkml Message-Id: <20051110164409.20950.43161.sendpatchset@localhost.localdomain> In-Reply-To: <20051110163906.20950.45704.sendpatchset@localhost.localdomain> References: <20051110163906.20950.45704.sendpatchset@localhost.localdomain> Subject: [PATCH,RFC 2.6.14 11/15] KGDB: ppc64-specific changes Date: Thu, 10 Nov 2005 11:43:15 -0500 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 18133 Lines: 619 This adds basic KGDB support to ppc64, and support for kgdb8250 on the 'Maple' board. All of this was done by Frank Rowand (who is on vacation right now, but I'll try and answer for him). This should work on any ppc64 board via kgdboe, so long as there is an eth driver that supports netpoll. At the moment this is mutually exclusive with XMON. It is probably possible to allow them to be chained, but that sounds dangerous to me. This is similar to ppc32, but ppc32 does not explicitly test. arch/ppc64/Kconfig.debug | 14 - arch/ppc64/kernel/Makefile | 1 arch/ppc64/kernel/kgdb.c | 421 +++++++++++++++++++++++++++++++++++++++++++++ arch/ppc64/kernel/setup.c | 6 arch/ppc64/mm/fault.c | 8 include/asm-ppc64/kgdb.h | 54 +++++ lib/Kconfig.debug | 2 7 files changed, 497 insertions(+), 9 deletions(-) Index: linux-2.6.14/arch/ppc64/Kconfig.debug =================================================================== --- linux-2.6.14.orig/arch/ppc64/Kconfig.debug +++ linux-2.6.14/arch/ppc64/Kconfig.debug @@ -28,16 +28,9 @@ config DEBUG_STACK_USAGE This option will slow down process creation somewhat. -config DEBUGGER - bool "Enable debugger hooks" - depends on DEBUG_KERNEL - help - Include in-kernel hooks for kernel debuggers. Unless you are - intending to debug the kernel, say N here. - config XMON bool "Include xmon kernel debugger" - depends on DEBUGGER && !PPC_ISERIES + depends on DEBUG_KERNEL && !PPC_ISERIES help Include in-kernel hooks for the xmon kernel monitor/debugger. Unless you are intending to debug the kernel, say N here. @@ -55,6 +48,11 @@ config XMON_DEFAULT xmon is normally disabled unless booted with 'xmon=on'. Use 'xmon=off' to disable xmon init during runtime. +config DEBUGGER + bool + depends on KGDB || XMON + default y + config PPCDBG bool "Include PPCDBG realtime debugging" depends on DEBUG_KERNEL Index: linux-2.6.14/arch/ppc64/kernel/kgdb.c =================================================================== --- /dev/null +++ linux-2.6.14/arch/ppc64/kernel/kgdb.c @@ -0,0 +1,421 @@ +/* + * arch/ppc64/kernel/kgdb.c + * + * PowerPC64 backend to the KGDB stub. + * + * Maintainer: Tom Rini + * + * Copied from arch/ppc/kernel/kgdb.c, updated for ppc64 + * + * Copyright (C) 1996 Paul Mackerras (setjmp/longjmp) + * 1998 (c) Michael AK Tesch (tesch@cs.wisc.edu) + * Copyright (C) 2003 Timesys Corporation. + * 2004 (c) MontaVista Software, Inc. + * 2005 (c) MontaVista Software, Inc. + * PPC64 Mods (C) 2005 Frank Rowand (frowand@mvista.com) + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program as licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* + * This table contains the mapping between PowerPC64 hardware trap types, and + * signals, which are primarily what GDB understands. GDB and the kernel + * don't always agree on values, so we use constants taken from gdb-6.2. + */ +static struct hard_trap_info +{ + unsigned int tt; /* Trap type code for powerpc */ + unsigned char signo; /* Signal that we map this trap into */ +} hard_trap_info[] = { + { 0x0100, 0x02 /* SIGINT */ }, /* system reset */ + { 0x0200, 0x0b /* SIGSEGV */ }, /* machine check */ + { 0x0300, 0x0b /* SIGSEGV */ }, /* data access */ + { 0x0380, 0x0b /* SIGSEGV */ }, /* data SLB access */ + { 0x0400, 0x0a /* SIGBUS */ }, /* instruction access */ + { 0x0480, 0x0a /* SIGBUS */ }, /* instruction segment */ + { 0x0500, 0x02 /* SIGINT */ }, /* interrupt */ + { 0x0600, 0x0a /* SIGBUS */ }, /* alignment */ + { 0x0700, 0x04 /* SIGILL */ }, /* program */ + { 0x0800, 0x08 /* SIGFPE */ }, /* fpu unavailable */ + { 0x0900, 0x0e /* SIGALRM */ }, /* decrementer */ + { 0x0a00, 0x04 /* SIGILL */ }, /* reserved */ + { 0x0b00, 0x04 /* SIGILL */ }, /* reserved */ + { 0x0c00, 0x14 /* SIGCHLD */ }, /* syscall */ + { 0x0d00, 0x05 /* SIGTRAP */ }, /* single step */ + { 0x0e00, 0x04 /* SIGILL */ }, /* reserved */ + { 0x0f00, 0x04 /* SIGILL */ }, /* performance monitor */ + { 0x0f20, 0x08 /* SIGFPE */ }, /* altivec unavailable */ + { 0x1300, 0x05 /* SIGTRAP */ }, /* instruction address break */ + { 0x1500, 0x04 /* SIGILL */ }, /* soft patch */ + { 0x1600, 0x04 /* SIGILL */ }, /* maintenance */ + { 0x1700, 0x04 /* SIGILL */ }, /* altivec assist */ + { 0x1800, 0x04 /* SIGILL */ }, /* thermal */ + { 0x0000, 0x000 } /* Must be last */ +}; + +extern atomic_t cpu_doing_single_step; + +static int computeSignal(unsigned int tt) +{ + struct hard_trap_info *ht; + + for (ht = hard_trap_info; ht->tt && ht->signo; ht++) + if (ht->tt == tt) + return ht->signo; + + return SIGHUP; /* default for things we don't know about */ +} + +static int kgdb_call_nmi_hook(struct pt_regs *regs) +{ + kgdb_nmihook(smp_processor_id(), regs); + return 0; +} + +void kgdb_roundup_cpus(unsigned long flags) +{ + smp_send_debugger_break(MSG_ALL_BUT_SELF); +} + +/* KGDB functions to use existing PowerPC64 hooks. */ +static int kgdb_debugger(struct pt_regs *regs) +{ + return kgdb_handle_exception(0, computeSignal(TRAP(regs)), 0, regs); +} + +static int kgdb_breakpoint(struct pt_regs *regs) +{ + if (user_mode(regs)) + return 0; + + kgdb_handle_exception(0, SIGTRAP, 0, regs); + + if (*(u32 *) (regs->nip) == *(u32 *) (&arch_kgdb_ops.gdb_bpt_instr)) + regs->nip += 4; + + return 1; +} + +static int kgdb_singlestep(struct pt_regs *regs) +{ + if (user_mode(regs)) + return 0; + + kgdb_handle_exception(0, SIGTRAP, 0, regs); + return 1; +} + +int kgdb_iabr_match(struct pt_regs *regs) +{ + if (user_mode(regs)) + return 0; + + kgdb_handle_exception(0, computeSignal(TRAP(regs)), 0, regs); + return 1; +} + +int kgdb_dabr_match(struct pt_regs *regs) +{ + if (user_mode(regs)) + return 0; + + kgdb_handle_exception(0, computeSignal(TRAP(regs)), 0, regs); + return 1; +} + +#define PACK64(ptr,src) do { *(ptr++) = (src); } while(0) + +#define PACK32(ptr,src) do { \ + u32 *ptr32; \ + ptr32 = (u32 *)ptr; \ + *(ptr32++) = (src); \ + ptr = (unsigned long *)ptr32; \ + } while(0) + +void regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs) +{ + int reg; + unsigned long *ptr = gdb_regs; + + memset(gdb_regs, 0, NUMREGBYTES); + + for (reg = 0; reg < 32; reg++) + PACK64(ptr, regs->gpr[reg]); + + /* fp registers not used by kernel, leave zero */ + ptr += 32; + + PACK64(ptr, regs->nip); + PACK64(ptr, regs->msr); + PACK32(ptr, regs->ccr); + PACK64(ptr, regs->link); + PACK64(ptr, regs->ctr); + PACK32(ptr, regs->xer); + +#if 0 + Following are in struct thread_struct, not struct pt_regs, + ignoring for now since kernel does not use them. Would it + make sense to get them from the thread that kgdb is set to? + + If this code is enabled, update the definition of NUMREGBYTES to + include the vector registers and vector state registers. + + PACK32(ptr, p->thread->fpscr); + + /* vr registers not used by kernel, leave zero */ + ptr += 64; + + PACK32(ptr, p->thread->vscr); + PACK32(ptr, p->thread->vrsave); +#else + /* fpscr not used by kernel, leave zero */ + PACK32(ptr, 0); +#endif + + BUG_ON((unsigned long)ptr > + (unsigned long)(((void *)gdb_regs) + NUMREGBYTES)); +} + +void sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p) +{ + struct pt_regs *regs = (struct pt_regs *)(p->thread.ksp + + STACK_FRAME_OVERHEAD); + int reg; + unsigned long *ptr = gdb_regs; + + memset(gdb_regs, 0, NUMREGBYTES); + + /* Regs GPR0-2 */ + for (reg = 0; reg < 3; reg++) + PACK64(ptr, regs->gpr[reg]); + + /* Regs GPR3-13 are caller saved, not in regs->gpr[] */ + for (reg = 3; reg < 14; reg++) + PACK64(ptr, 0); + + /* Regs GPR14-31 */ + for (reg = 14; reg < 32; reg++) + PACK64(ptr, regs->gpr[reg]); + + /* fp registers not used by kernel, leave zero */ + ptr += 32; + + PACK64(ptr, regs->nip); + PACK64(ptr, regs->msr); + PACK32(ptr, regs->ccr); + PACK64(ptr, regs->link); + PACK64(ptr, regs->ctr); + PACK32(ptr, regs->xer); + +#if 0 + Following are in struct thread_struct, not struct pt_regs, + ignoring for now since kernel does not use them. Would it + make sense to get them from the thread that kgdb is set to? + + If this code is enabled, update the definition of NUMREGBYTES to + include the vector registers and vector state registers. + + PACK32(ptr, p->thread->fpscr); + + /* vr registers not used by kernel, leave zero */ + ptr += 64; + + PACK32(ptr, p->thread->vscr); + PACK32(ptr, p->thread->vrsave); +#else + /* fpscr not used by kernel, leave zero */ + PACK32(ptr, 0); +#endif + + BUG_ON((unsigned long)ptr > + (unsigned long)(((void *)gdb_regs) + NUMREGBYTES)); +} + +#define UNPACK64(dest,ptr) do { dest = *(ptr++); } while(0) + +#define UNPACK32(dest,ptr) do { \ + u32 *ptr32; \ + ptr32 = (u32 *)ptr; \ + dest = *(ptr32++); \ + ptr = (unsigned long *)ptr32; \ + } while(0) + +void gdb_regs_to_regs(unsigned long *gdb_regs, struct pt_regs *regs) +{ + int reg; + unsigned long *ptr = gdb_regs; + + for (reg = 0; reg < 32; reg++) + UNPACK64(regs->gpr[reg], ptr); + + /* fp registers not used by kernel, leave zero */ + ptr += 32; + + UNPACK64(regs->nip, ptr); + UNPACK64(regs->msr, ptr); + UNPACK32(regs->ccr, ptr); + UNPACK64(regs->link, ptr); + UNPACK64(regs->ctr, ptr); + UNPACK32(regs->xer, ptr); + +#if 0 + Following are in struct thread_struct, not struct pt_regs, + ignoring for now since kernel does not use them. Would it + make sense to get them from the thread that kgdb is set to? + + If this code is enabled, update the definition of NUMREGBYTES to + include the vector registers and vector state registers. + + /* fpscr, vscr, vrsave not used by kernel, leave unchanged */ + + UNPACK32(p->thread->fpscr, ptr); + + /* vr registers not used by kernel, leave zero */ + ptr += 64; + + UNPACK32(p->thread->vscr, ptr); + UNPACK32(p->thread->vrsave, ptr); +#endif + + BUG_ON((unsigned long)ptr > + (unsigned long)(((void *)gdb_regs) + NUMREGBYTES)); +} + +/* + * This function does PowerPC64 specific procesing for interfacing to gdb. + */ +int kgdb_arch_handle_exception(int vector, int signo, int err_code, + char *remcom_in_buffer, char *remcom_out_buffer, + struct pt_regs *linux_regs) +{ + char *ptr = &remcom_in_buffer[1]; + unsigned long addr; + + switch (remcom_in_buffer[0]) { + /* + * sAA..AA Step one instruction from AA..AA + * This will return an error to gdb .. + */ + case 's': + case 'c': + /* handle the optional parameter */ + if (kgdb_hex2long(&ptr, &addr)) + linux_regs->nip = addr; + + atomic_set(&cpu_doing_single_step, -1); + /* set the trace bit if we're stepping */ + if (remcom_in_buffer[0] == 's') { + linux_regs->msr |= MSR_SE; + debugger_step = 1; + if (kgdb_contthread) + atomic_set(&cpu_doing_single_step, + smp_processor_id()); + } + return 0; + } + + return -1; +} + +int kgdb_fault_setjmp(unsigned long *curr_context) +{ + __asm__ __volatile__("mflr 0; std 0,0(%0)\n\ + std 1,8(%0)\n\ + std 2,16(%0)\n\ + mfcr 0; std 0,24(%0)\n\ + std 13,32(%0)\n\ + std 14,40(%0)\n\ + std 15,48(%0)\n\ + std 16,56(%0)\n\ + std 17,64(%0)\n\ + std 18,72(%0)\n\ + std 19,80(%0)\n\ + std 20,88(%0)\n\ + std 21,96(%0)\n\ + std 22,104(%0)\n\ + std 23,112(%0)\n\ + std 24,120(%0)\n\ + std 25,128(%0)\n\ + std 26,136(%0)\n\ + std 27,144(%0)\n\ + std 28,152(%0)\n\ + std 29,160(%0)\n\ + std 30,168(%0)\n\ + std 31,176(%0)\n" : : "r" (curr_context)); + return 0; +} + +void kgdb_fault_longjmp(unsigned long *curr_context) +{ + __asm__ __volatile__("ld 13,32(%0)\n\ + ld 14,40(%0)\n\ + ld 15,48(%0)\n\ + ld 16,56(%0)\n\ + ld 17,64(%0)\n\ + ld 18,72(%0)\n\ + ld 19,80(%0)\n\ + ld 20,88(%0)\n\ + ld 21,96(%0)\n\ + ld 22,104(%0)\n\ + ld 23,112(%0)\n\ + ld 24,120(%0)\n\ + ld 25,128(%0)\n\ + ld 26,136(%0)\n\ + ld 27,144(%0)\n\ + ld 28,152(%0)\n\ + ld 29,160(%0)\n\ + ld 30,168(%0)\n\ + ld 31,176(%0)\n\ + ld 0,24(%0)\n\ + mtcrf 0x38,0\n\ + ld 0,0(%0)\n\ + ld 1,8(%0)\n\ + ld 2,16(%0)\n\ + mtlr 0\n\ + mr 3,1\n" : : "r" (curr_context)); +} + +/* + * Global data + */ +struct kgdb_arch arch_kgdb_ops = { + .gdb_bpt_instr = {0x7d, 0x82, 0x10, 0x08}, +}; + +int kgdb_not_implemented(struct pt_regs *regs) +{ + return 0; +} + +int kgdb_arch_init(void) +{ +#ifdef CONFIG_XMON +#error Both XMON and KGDB selected in .config. Unselect one of them. +#endif + + __debugger_ipi = kgdb_call_nmi_hook; + __debugger = kgdb_debugger; + __debugger_bpt = kgdb_breakpoint; + __debugger_sstep = kgdb_singlestep; + __debugger_iabr_match = kgdb_iabr_match; + __debugger_dabr_match = kgdb_dabr_match; + __debugger_fault_handler = kgdb_not_implemented; + + return 0; +} + +arch_initcall(kgdb_arch_init); Index: linux-2.6.14/arch/ppc64/kernel/Makefile =================================================================== --- linux-2.6.14.orig/arch/ppc64/kernel/Makefile +++ linux-2.6.14/arch/ppc64/kernel/Makefile @@ -56,6 +56,7 @@ vio-obj-$(CONFIG_PPC_ISERIES) += iSeries obj-$(CONFIG_IBMVIO) += vio.o $(vio-obj-y) obj-$(CONFIG_XICS) += xics.o obj-$(CONFIG_MPIC) += mpic.o +obj-$(CONFIG_KGDB) += kgdb.o obj-$(CONFIG_PPC_PMAC) += pmac_setup.o pmac_feature.o pmac_pci.o \ pmac_time.o pmac_nvram.o pmac_low_i2c.o \ Index: linux-2.6.14/arch/ppc64/kernel/setup.c =================================================================== --- linux-2.6.14.orig/arch/ppc64/kernel/setup.c +++ linux-2.6.14/arch/ppc64/kernel/setup.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include #include @@ -1219,6 +1220,11 @@ void __init generic_find_legacy_serial_p serial_ports[index].irq, serial_ports[index].uartclk); +#ifdef CONFIG_KGDB_8250 + /* Tell KGDB about this. */ + kgdb8250_add_platform_port(index, &serial_ports[index]); +#endif + /* Get phys address of IO reg for port 1 */ if (index != 0) goto next_port; Index: linux-2.6.14/arch/ppc64/mm/fault.c =================================================================== --- linux-2.6.14.orig/arch/ppc64/mm/fault.c +++ linux-2.6.14/arch/ppc64/mm/fault.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include @@ -327,6 +328,13 @@ void bad_page_fault(struct pt_regs *regs regs->nip = entry->fixup; return; } +#ifdef CONFIG_KGDB + if (atomic_read(&debugger_active) && kgdb_may_fault) { + /* Restore our previous state. */ + kgdb_fault_longjmp(kgdb_fault_jmp_regs); + /* Not reached. */ + } +#endif /* kernel has accessed a bad area */ die("Kernel access of bad area", regs, sig); Index: linux-2.6.14/include/asm-ppc64/kgdb.h =================================================================== --- /dev/null +++ linux-2.6.14/include/asm-ppc64/kgdb.h @@ -0,0 +1,54 @@ +/* + * kgdb.h: Defines and declarations for serial line source level + * remote debugging of the Linux kernel using gdb. + * + * copied from include/asm-ppc, modified for ppc64 + * + * PPC64 Mods (C) 2005 Frank Rowand (frowand@mvista.com) + * PPC Mods (C) 2004 Tom Rini (trini@mvista.com) + * PPC Mods (C) 2003 John Whitney (john.whitney@timesys.com) + * PPC Mods (C) 1998 Michael Tesch (tesch@cs.wisc.edu) + * + * Copyright (C) 1995 David S. Miller (davem@caip.rutgers.edu) + */ +#ifdef __KERNEL__ +#ifndef _PPC64_KGDB_H +#define _PPC64_KGDB_H + +#ifndef __ASSEMBLY__ + +#define BREAK_INSTR_SIZE 4 +#if 1 +/* + * Does not include vector registers and vector state registers. + * + * 64 bit (8 byte) registers: + * 32 gpr, 32 fpr, nip, msr, link, ctr + * 32 bit (4 byte) registers: + * ccr, xer, fpscr + */ +#define NUMREGBYTES ((68 * 8) + (3 * 4)) +#else +/* + * Includes vector registers and vector state registers. + * + * 128 bit (16 byte) registers: + * 32 vr + * 64 bit (8 byte) registers: + * 32 gpr, 32 fpr, nip, msr, link, ctr + * 32 bit (4 byte) registers: + * ccr, xer, fpscr, vscr, vrsave + */ +#define NUMREGBYTES ((128 * 16) + (68 * 8) + (5 * 4)) +#endif +#define NUMCRITREGBYTES 184 +#define BUFMAX ((NUMREGBYTES * 2) + 512) +#define OUTBUFMAX ((NUMREGBYTES * 2) + 512) +#define BREAKPOINT() asm(".long 0x7d821008"); /* twge r2, r2 */ +#define CACHE_FLUSH_IS_SAFE 1 + +#endif /* !(__ASSEMBLY__) */ + +#endif /* !(_PPC64_KGDB_H) */ + +#endif /* __KERNEL__ */ Index: linux-2.6.14/lib/Kconfig.debug =================================================================== --- linux-2.6.14.orig/lib/Kconfig.debug +++ linux-2.6.14/lib/Kconfig.debug @@ -187,7 +187,7 @@ config WANT_EXTRA_DEBUG_INFORMATION config KGDB bool "KGDB: kernel debugging with remote gdb" select WANT_EXTRA_DEBUG_INFORMATION - depends on DEBUG_KERNEL && (ARM || X86 || MIPS || (SUPERH && !SUPERH64) || IA64 || X86_64 || ((!SMP || BROKEN) && PPC32)) + depends on DEBUG_KERNEL && (ARM || X86 || MIPS || PPC64 || (SUPERH && !SUPERH64) || IA64 || X86_64 || ((!SMP || BROKEN) && PPC32)) help If you say Y here, it will be possible to remotely debug the kernel using gdb. It is strongly suggested that you enable -- Tom - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/