From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754631AbXLDSEn (ORCPT ); Tue, 4 Dec 2007 13:04:43 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752678AbXLDSEa (ORCPT ); Tue, 4 Dec 2007 13:04:30 -0500 Received: from an-out-0708.google.com ([209.85.132.242]:28501 "EHLO an-out-0708.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752674AbXLDSE1 (ORCPT ); Tue, 4 Dec 2007 13:04:27 -0500 DomainKey-Signature: a=rsa-sha1; c=nofws; d=googlemail.com; s=gamma; h=received:message-id:date:from:to:subject:cc:mime-version:content-type:content-transfer-encoding:content-disposition; b=Ov8IBO5GGkmNgVsjcZLoPAlRpAbJJPQhpwgTJR5V7vSu5xoxQAG/rW37ZSOf+IfZ40geX7So4hhWs/qYZvwVLzQNlXBR1SkYMjvr4PEX+pQz6lM3skN87nHCgfMIj76og2pu+yTU/QlnkSXpyf643DO4oxfjsF6WKvkerdy7Lzc= Message-ID: Date: Tue, 4 Dec 2007 19:04:26 +0100 From: "Markus Metzger" To: ak@suse.de, hpa@zytor.com, linux-kernel@vger.kernel.org, mingo@elte.hu, tglx@linutronix.de Subject: [patch 1/2] x86, ptrace: support for branch trace store(BTS) Cc: akpm@linux-foundation.org, markus.t.metzger@googlemail.com, markus.t.metzger@intel.com, mtk.manpages@gmail.com, roland@redhat.com, suresh.b.siddha@intel.com MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Changes to the last version: - split implementation into two layers: ds/bts and ptrace - renamed TIF's - save/restore ds save area msr in __switch_to_xtra() - make block-stepping only look at BTF bit Signed-off-by: Markus Metzger Signed-off-by: Suresh Siddha --- Index: linux-2.6-x86/arch/x86/kernel/process_32.c =================================================================== --- linux-2.6-x86.orig/arch/x86/kernel/process_32.c 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/arch/x86/kernel/process_32.c 2007-12-04 18:12:42.%N +0100 @@ -594,11 +594,21 @@ struct tss_struct *tss) { struct thread_struct *prev, *next; + unsigned long debugctl; prev = &prev_p->thread; next = &next_p->thread; - if (next->debugctlmsr != prev->debugctlmsr) + debugctl = prev->debugctlmsr; + if (next->ds_area_msr != prev->ds_area_msr) { + /* we clear debugctl to make sure DS + * is not in use when we change it */ + debugctl = 0; + wrmsrl(MSR_IA32_DEBUGCTLMSR, 0); + wrmsr(MSR_IA32_DS_AREA, next->ds_area_msr, 0); + } + + if (next->debugctlmsr != debugctl) wrmsr(MSR_IA32_DEBUGCTLMSR, next->debugctlmsr, 0); if (test_tsk_thread_flag(next_p, TIF_DEBUG)) { @@ -622,6 +632,13 @@ } #endif + if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS)) + ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS); + + if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS)) + ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES); + + if (!test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) { /* * Disable the bitmap via an invalid offset. We still cache Index: linux-2.6-x86/include/asm-x86/ptrace-abi.h =================================================================== --- linux-2.6-x86.orig/include/asm-x86/ptrace-abi.h 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/include/asm-x86/ptrace-abi.h 2007-12-04 17:21:28.%N +0100 @@ -80,4 +80,56 @@ #define PTRACE_SINGLEBLOCK 33 /* resume execution until next branch */ +/* Return maximal BTS buffer size in number of records, + if successuf; -1, otherwise. + EOPNOTSUPP...processor does not support bts tracing */ +#define PTRACE_BTS_MAX_BUFFER_SIZE 40 + +/* Allocate new bts buffer (free old one, if exists) of size DATA bts records; + parameter ADDR is ignored. + Return 0, if successful; -1, otherwise. + EOPNOTSUPP...processor does not support bts tracing + EINVAL.......invalid size in records + ENOMEM.......out of memory */ +#define PTRACE_BTS_ALLOCATE_BUFFER 41 + +/* Return the size of the bts buffer in number of bts records, + if successful; -1, otherwise. + EOPNOTSUPP...processor does not support bts tracing + ENXIO........no buffer allocated */ +#define PTRACE_BTS_GET_BUFFER_SIZE 42 + +/* Return the index of the next bts record to be written, + if successful; -1, otherwise. + EOPNOTSUPP...processor does not support bts tracing + ENXIO........no buffer allocated + After the first warp-around, this is the start of the circular bts buffer. */ +#define PTRACE_BTS_GET_INDEX 43 + +/* Read the DATA'th bts record into a ptrace_bts_record buffer provided in ADDR. + Return 0, if successful; -1, otherwise + EOPNOTSUPP...processor does not support bts tracing + ENXIO........no buffer allocated + EINVAL.......invalid index */ +#define PTRACE_BTS_READ_RECORD 44 + +/* Configure last branch trace; the configuration is given as a bit-mask of + PTRACE_BTS_O_* options in DATA; parameter ADDR is ignored. + Return 0, if successful; -1, otherwise + EOPNOTSUPP...processor does not support bts tracing + ENXIO........no buffer allocated */ +#define PTRACE_BTS_CONFIG 45 + +/* Return the configuration as bit-mask of PTRACE_BTS_O_* options + if successful; -1, otherwise. + EOPNOTSUPP...processor does not support bts tracing + ENXIO........no buffer allocated */ +#define PTRACE_BTS_STATUS 46 + +/* Trace configuration options */ +/* Collect last branch trace */ +#define PTRACE_BTS_O_TRACE_TASK 0x1 +/* Take timestamps when the task arrives and departs */ +#define PTRACE_BTS_O_TIMESTAMPS 0x2 + #endif Index: linux-2.6-x86/include/asm-x86/ptrace.h =================================================================== --- linux-2.6-x86.orig/include/asm-x86/ptrace.h 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/include/asm-x86/ptrace.h 2007-12-04 18:17:52.%N +0100 @@ -4,8 +4,19 @@ #include /* For __user */ #include + #ifndef __ASSEMBLY__ +#ifdef __KERNEL__ + +#include + +struct task_struct; +extern void ptrace_bts_take_timestamp(struct task_struct *, enum bts_qualifier); + +#endif /* __KERNEL__ */ + + #ifdef __i386__ /* this struct defines the way the registers are stored on the stack during a system call. */ Index: linux-2.6-x86/arch/x86/kernel/Makefile_32 =================================================================== --- linux-2.6-x86.orig/arch/x86/kernel/Makefile_32 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/arch/x86/kernel/Makefile_32 2007-12-04 17:21:28.%N +0100 @@ -11,6 +11,7 @@ quirks.o i8237.o topology.o alternative.o i8253.o tsc_32.o rtc.o obj-y += ptrace.o +obj-y += ds.o obj-y += tls.o obj-y += step.o obj-$(CONFIG_STACKTRACE) += stacktrace.o Index: linux-2.6-x86/include/asm-x86/thread_info_32.h =================================================================== --- linux-2.6-x86.orig/include/asm-x86/thread_info_32.h 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/include/asm-x86/thread_info_32.h 2007-12-04 17:21:28.%N +0100 @@ -139,6 +139,8 @@ #define TIF_NOTSC 20 /* TSC is not accessible in userland */ #define TIF_FORCED_TF 21 /* true if TF in eflags artificially */ #define TIF_DEBUGCTLMSR 22 /* uses thread_struct.debugctlmsr */ +#define TIF_DS_AREA_MSR 23 /* uses thread_struct.ds_area_msr */ +#define TIF_BTS_TRACE_TS 24 /* record scheduling event timestamps */ #define _TIF_SYSCALL_TRACE (1<thread, next = &next_p->thread; - if (next->debugctlmsr != prev->debugctlmsr) + debugctl = prev->debugctlmsr; + if (next->ds_area_msr != prev->ds_area_msr) { + /* we clear debugctl to make sure DS + * is not in use when we change it */ + debugctl = 0; + wrmsrl(MSR_IA32_DEBUGCTLMSR, 0); + wrmsrl(MSR_IA32_DS_AREA, next->ds_area_msr); + } + + if (next->debugctlmsr != debugctl) wrmsrl(MSR_IA32_DEBUGCTLMSR, next->debugctlmsr); if (test_tsk_thread_flag(next_p, TIF_DEBUG)) { @@ -580,6 +590,16 @@ */ memset(tss->io_bitmap, 0xff, prev->io_bitmap_max); } + + /* + * Last branch recording recofiguration of trace hardware and + * disentangling of trace data per task. + */ + if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS)) + ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS); + + if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS)) + ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES); } /* @@ -683,8 +703,8 @@ /* * Now maybe reload the debug registers and handle I/O bitmaps */ - if (unlikely((task_thread_info(next_p)->flags & _TIF_WORK_CTXSW)) - || test_tsk_thread_flag(prev_p, TIF_IO_BITMAP)) + if (unlikely(task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT || + task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV)) __switch_to_xtra(prev_p, next_p, tss); /* If the task has used fpu the last 5 timeslices, just do a full Index: linux-2.6-x86/include/asm-x86/processor_64.h =================================================================== --- linux-2.6-x86.orig/include/asm-x86/processor_64.h 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/include/asm-x86/processor_64.h 2007-12-04 17:21:28.%N +0100 @@ -240,6 +240,9 @@ unsigned io_bitmap_max; /* MSR_IA32_DEBUGCTLMSR value to switch in if TIF_DEBUGCTLMSR is set. */ unsigned long debugctlmsr; +/* Debug Store - if not 0 points to a DS Save Area configuration; + * goes into MSR_IA32_DS_AREA */ + unsigned long ds_area_msr; /* cached TLS descriptors. */ u64 tls_array[GDT_ENTRY_TLS_ENTRIES]; } __attribute__((aligned(16))); Index: linux-2.6-x86/include/asm-x86/thread_info_64.h =================================================================== --- linux-2.6-x86.orig/include/asm-x86/thread_info_64.h 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/include/asm-x86/thread_info_64.h 2007-12-04 17:21:28.%N +0100 @@ -121,6 +121,8 @@ #define TIF_IO_BITMAP 22 /* uses I/O bitmap */ #define TIF_FREEZE 23 /* is freezing for suspend */ #define TIF_DEBUGCTLMSR 24 /* uses thread_struct.debugctlmsr */ +#define TIF_DS_AREA_MSR 25 /* uses thread_struct.ds_area_msr */ +#define TIF_BTS_TRACE_TS 26 /* record scheduling event timestamps */ #define _TIF_SYSCALL_TRACE (1< #include #include +#include +#include #include "cpu.h" @@ -219,6 +221,9 @@ if (!(l1 & (1<<12))) set_bit(X86_FEATURE_PEBS, c->x86_capability); } + + if (cpu_has_bts) + ds_init_intel(c); } static unsigned int __cpuinit intel_size_cache(struct cpuinfo_x86 * c, unsigned int size) Index: linux-2.6-x86/arch/x86/kernel/setup_64.c =================================================================== --- linux-2.6-x86.orig/arch/x86/kernel/setup_64.c 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/arch/x86/kernel/setup_64.c 2007-12-04 17:21:28.%N +0100 @@ -60,6 +60,7 @@ #include #include #include +#include /* * Machine setup.. @@ -850,6 +851,10 @@ set_cpu_cap(c, X86_FEATURE_PEBS); } + + if (cpu_has_bts) + ds_init_intel(c); + n = c->extended_cpuid_level; if (n >= 0x80000008) { unsigned eax = cpuid_eax(0x80000008); Index: linux-2.6-x86/arch/x86/kernel/ptrace.c =================================================================== --- linux-2.6-x86.orig/arch/x86/kernel/ptrace.c 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/arch/x86/kernel/ptrace.c 2007-12-04 18:13:37.%N +0100 @@ -2,6 +2,9 @@ /* * Pentium III FXSR, SSE support * Gareth Hughes , May 2000 + * + * BTS tracing + * Markus Metzger , Dec 2007 */ #include @@ -26,6 +29,14 @@ #include #include #include +#include + + +/* + * The maximal size of a BTS buffer per traced task in number of BTS + * records. + */ +#define PTRACE_BTS_BUFFER_MAX 4000 /* * does not yet catch signals sent when the child dies. @@ -455,6 +466,165 @@ return 0; } +static int ptrace_bts_max_buffer_size(void) +{ + return PTRACE_BTS_BUFFER_MAX; +} + +static int ptrace_bts_get_buffer_size(struct task_struct *child) +{ + if (!child->thread.ds_area_msr) + return -ENXIO; + + return ds_get_bts_size((void *)child->thread.ds_area_msr); +} + +static int ptrace_bts_get_index(struct task_struct *child) +{ + if (!child->thread.ds_area_msr) + return -ENXIO; + + return ds_get_bts_index((void *)child->thread.ds_area_msr); +} + +static int ptrace_bts_read_record(struct task_struct *child, + long index, + struct bts_struct __user *out) +{ + struct bts_struct ret; + int retval; + + if (!child->thread.ds_area_msr) + return -ENXIO; + + retval = ds_read_bts((void *)child->thread.ds_area_msr, + index, &ret); + if (retval) + return retval; + + if (copy_to_user(out, &ret, sizeof(ret))) + return -EFAULT; + + return sizeof(ret); +} + +static int ptrace_bts_write_record(struct task_struct *child, + const struct bts_struct *in) +{ + int retval; + + if (!child->thread.ds_area_msr) + return -ENXIO; + + retval = ds_write_bts((void *)child->thread.ds_area_msr, in); + if (retval) + return retval; + + return sizeof(*in); +} + +static int ptrace_bts_config(struct task_struct *child, + unsigned long options) +{ + unsigned long debugctl_mask = ds_debugctl_mask(); + int retval; + + retval = ptrace_bts_get_buffer_size(child); + if (retval < 0) + return retval; + if (retval == 0) + return -ENXIO; + + if (options & PTRACE_BTS_O_TRACE_TASK) { + child->thread.debugctlmsr |= debugctl_mask; + set_tsk_thread_flag(child, TIF_DEBUGCTLMSR); + } else { + /* there is no way for us to check whether we 'own' + * the respective bits in the DEBUGCTL MSR, we're + * about to clear */ + child->thread.debugctlmsr &= ~debugctl_mask; + + if (!child->thread.debugctlmsr) + clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR); + } + + if (options & PTRACE_BTS_O_TIMESTAMPS) + set_tsk_thread_flag(child, TIF_BTS_TRACE_TS); + else + clear_tsk_thread_flag(child, TIF_BTS_TRACE_TS); + + return 0; +} + +static int ptrace_bts_status(struct task_struct *child) +{ + unsigned long debugctl_mask = ds_debugctl_mask(); + int retval, status = 0; + + retval = ptrace_bts_get_buffer_size(child); + if (retval < 0) + return retval; + if (retval == 0) + return -ENXIO; + + if (ptrace_bts_get_buffer_size(child) <= 0) + return -ENXIO; + + if (test_tsk_thread_flag(child, TIF_DEBUGCTLMSR) && + child->thread.debugctlmsr & debugctl_mask) + status |= PTRACE_BTS_O_TRACE_TASK; + if (test_tsk_thread_flag(child, TIF_BTS_TRACE_TS)) + status |= PTRACE_BTS_O_TIMESTAMPS; + + return status; +} + +static int ptrace_bts_allocate_bts(struct task_struct *child, + int size_in_records) +{ + int retval = 0; + void *ds; + + if (size_in_records < 0) + return -EINVAL; + + if (size_in_records > ptrace_bts_max_buffer_size()) + return -EINVAL; + + if (size_in_records == 0) { + ptrace_bts_config(child, /* options = */ 0); + } else { + retval = ds_allocate(&ds, size_in_records); + if (retval) + return retval; + } + + if (child->thread.ds_area_msr) + ds_free((void **)&child->thread.ds_area_msr); + + child->thread.ds_area_msr = (unsigned long)ds; + if (child->thread.ds_area_msr) + set_tsk_thread_flag(child, TIF_DS_AREA_MSR); + else + clear_tsk_thread_flag(child, TIF_DS_AREA_MSR); + + return retval; +} + +void ptrace_bts_take_timestamp(struct task_struct *tsk, + enum bts_qualifier qualifier) +{ + struct bts_struct rec = { + .qualifier = qualifier, + .variant.timestamp = sched_clock() + }; + + if (ptrace_bts_get_buffer_size(tsk) <= 0) + return; + + ptrace_bts_write_record(tsk, &rec); +} + /* * Called by kernel/ptrace.c when detaching.. * @@ -466,6 +636,11 @@ #ifdef TIF_SYSCALL_EMU clear_tsk_thread_flag(child, TIF_SYSCALL_EMU); #endif + ptrace_bts_config(child, /* options = */ 0); + if (child->thread.ds_area_msr) { + ds_free((void **)&child->thread.ds_area_msr); + clear_tsk_thread_flag(child, TIF_DS_AREA_MSR); + } } long arch_ptrace(struct task_struct *child, long request, long addr, long data) @@ -626,6 +801,36 @@ break; #endif + case PTRACE_BTS_MAX_BUFFER_SIZE: + ret = ptrace_bts_max_buffer_size(); + break; + + case PTRACE_BTS_ALLOCATE_BUFFER: + ret = ptrace_bts_allocate_bts(child, data); + break; + + case PTRACE_BTS_GET_BUFFER_SIZE: + ret = ptrace_bts_get_buffer_size(child); + break; + + case PTRACE_BTS_GET_INDEX: + ret = ptrace_bts_get_index(child); + break; + + case PTRACE_BTS_READ_RECORD: + ret = ptrace_bts_read_record + (child, data, + (struct bts_struct __user *) addr); + break; + + case PTRACE_BTS_CONFIG: + ret = ptrace_bts_config(child, data); + break; + + case PTRACE_BTS_STATUS: + ret = ptrace_bts_status(child); + break; + default: ret = ptrace_request(child, request, addr, data); break; @@ -809,6 +1014,13 @@ case PTRACE_SETOPTIONS: case PTRACE_SET_THREAD_AREA: case PTRACE_GET_THREAD_AREA: + case PTRACE_BTS_MAX_BUFFER_SIZE: + case PTRACE_BTS_ALLOCATE_BUFFER: + case PTRACE_BTS_GET_BUFFER_SIZE: + case PTRACE_BTS_GET_INDEX: + case PTRACE_BTS_READ_RECORD: + case PTRACE_BTS_CONFIG: + case PTRACE_BTS_STATUS: return sys_ptrace(request, pid, addr, data); default: Index: linux-2.6-x86/arch/x86/kernel/Makefile_64 =================================================================== --- linux-2.6-x86.orig/arch/x86/kernel/Makefile_64 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/arch/x86/kernel/Makefile_64 2007-12-04 17:21:28.%N +0100 @@ -13,6 +13,7 @@ i8253.o rtc.o obj-y += ptrace.o +obj-y += ds.o obj-y += step.o obj-$(CONFIG_IA32_EMULATION) += tls.o Index: linux-2.6-x86/arch/x86/kernel/step.c =================================================================== --- linux-2.6-x86.orig/arch/x86/kernel/step.c 2007-12-04 16:44:47.%N +0100 +++ linux-2.6-x86/arch/x86/kernel/step.c 2007-12-04 17:21:28.%N +0100 @@ -169,9 +169,14 @@ */ if (enable_single_step(child) && block) { set_tsk_thread_flag(child, TIF_DEBUGCTLMSR); - write_debugctlmsr(child, DEBUGCTLMSR_BTF); - } else if (test_and_clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR)) { - write_debugctlmsr(child, 0); + write_debugctlmsr(child, + child->thread.debugctlmsr | DEBUGCTLMSR_BTF); + } else { + write_debugctlmsr(child, + child->thread.debugctlmsr & ~TIF_DEBUGCTLMSR); + + if (!child->thread.debugctlmsr) + clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR); } } @@ -190,8 +195,11 @@ /* * Make sure block stepping (BTF) is disabled. */ - if (test_and_clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR)) - write_debugctlmsr(child, 0); + write_debugctlmsr(child, + child->thread.debugctlmsr & ~TIF_DEBUGCTLMSR); + + if (!child->thread.debugctlmsr) + clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR); /* Always clear TIF_SINGLESTEP... */ clear_tsk_thread_flag(child, TIF_SINGLESTEP); Index: linux-2.6-x86/arch/x86/kernel/ds.c =================================================================== --- /dev/null 1970-01-01 00:00:00.000000000 +0000 +++ linux-2.6-x86/arch/x86/kernel/ds.c 2007-12-04 17:21:28.%N +0100 @@ -0,0 +1,429 @@ +/* + * Debug Store support + * + * This provides a low-level interface to the hardware's Debug Store + * feature that is used for last branch recording (LBR) and + * precise-event based sampling (PEBS). + * + * Different architectures use a different DS layout/pointer size. + * The below functions therefore work on a void*. + * + * + * Since there is no user for PEBS, yet, only LBR (or branch + * trace store, BTS) is supported. + * + * + * Copyright (C) 2007 Intel Corporation. + * Markus Metzger , Dec 2007 + */ + +#include + +#include +#include +#include + + +/* + * Debug Store (DS) save area configuration (see Intel64 and IA32 + * Architectures Software Developer's Manual, section 18.5) + * + * The DS configuration consists of the following fields; different + * architetures vary in the size of those fields. + * - double-word aligned base linear address of the BTS buffer + * - write pointer into the BTS buffer + * - end linear address of the BTS buffer (one byte beyond the end of + * the buffer) + * - interrupt pointer into BTS buffer + * (interrupt occurs when write pointer passes interrupt pointer) + * - double-word aligned base linear address of the PEBS buffer + * - write pointer into the PEBS buffer + * - end linear address of the PEBS buffer (one byte beyond the end of + * the buffer) + * - interrupt pointer into PEBS buffer + * (interrupt occurs when write pointer passes interrupt pointer) + * - value to which counter is reset following counter overflow + * + * On later architectures, the last branch recording hardware uses + * 64bit pointers even in 32bit mode. + * + * + * Branch Trace Store (BTS) records store information about control + * flow changes. They at least provide the following information: + * - source linear address + * - destination linear address + * + * Netburst supported a predicated bit that had been dropped in later + * architectures. We do not suppor it. + * + * + * In order to abstract from the actual DS and BTS layout, we describe + * the access to the relevant fields. + * Thanks to Andi Kleen for proposing this design. + * + * The implementation, however, is not as general as it might seem. In + * order to stay somewhat simple and efficient, we assume an + * underlying unsigned type (mostly a pointer type) and we expect the + * field to be at least as big as that type. + */ + +/* + * A special from_ip address to indicate that the BTS record is an + * info record that needs to be interpreted or skipped. + */ +#define BTS_ESCAPE_ADDRESS (-1) + +/* + * A field access descriptor + */ +struct access_desc { + unsigned char offset; + unsigned char size; +}; + +/* + * The configuration for a particular DS/BTS hardware implementation. + */ +struct ds_configuration { + /* the DS configuration */ + unsigned char sizeof_ds; + struct access_desc bts_buffer_base; + struct access_desc bts_index; + struct access_desc bts_absolute_maximum; + struct access_desc bts_interrupt_threshold; + /* the BTS configuration */ + unsigned char sizeof_bts; + struct access_desc from_ip; + struct access_desc to_ip; + /* BTS variants used to store additional information like + timestamps */ + struct access_desc info_type; + struct access_desc info_data; + unsigned long debugctl_mask; +}; + +/* + * The global configuration used by the below accessor functions + */ +static struct ds_configuration ds_cfg; + +/* + * Accessor functions for some DS and BTS fields using the above + * global ptrace_bts_cfg. + */ +static inline void *get_bts_buffer_base(char *base) +{ + return *(void **)(base + ds_cfg.bts_buffer_base.offset); +} +static inline void set_bts_buffer_base(char *base, void *value) +{ + (*(void **)(base + ds_cfg.bts_buffer_base.offset)) = value; +} +static inline void *get_bts_index(char *base) +{ + return *(void **)(base + ds_cfg.bts_index.offset); +} +static inline void set_bts_index(char *base, void *value) +{ + (*(void **)(base + ds_cfg.bts_index.offset)) = value; +} +static inline void *get_bts_absolute_maximum(char *base) +{ + return *(void **)(base + ds_cfg.bts_absolute_maximum.offset); +} +static inline void set_bts_absolute_maximum(char *base, void *value) +{ + (*(void **)(base + ds_cfg.bts_absolute_maximum.offset)) = value; +} +static inline void *get_bts_interrupt_threshold(char *base) +{ + return *(void **)(base + ds_cfg.bts_interrupt_threshold.offset); +} +static inline void set_bts_interrupt_threshold(char *base, void *value) +{ + (*(void **)(base + ds_cfg.bts_interrupt_threshold.offset)) = value; +} +static inline long get_from_ip(char *base) +{ + return *(long *)(base + ds_cfg.from_ip.offset); +} +static inline void set_from_ip(char *base, long value) +{ + (*(long *)(base + ds_cfg.from_ip.offset)) = value; +} +static inline long get_to_ip(char *base) +{ + return *(long *)(base + ds_cfg.to_ip.offset); +} +static inline void set_to_ip(char *base, long value) +{ + (*(long *)(base + ds_cfg.to_ip.offset)) = value; +} +static inline unsigned char get_info_type(char *base) +{ + return *(unsigned char *)(base + ds_cfg.info_type.offset); +} +static inline void set_info_type(char *base, unsigned char value) +{ + (*(unsigned char *)(base + ds_cfg.info_type.offset)) = value; +} +/* + * The info data might overlap with the info type on some architectures. + * We therefore read and write the exact number of bytes. + */ +static inline unsigned long long get_info_data(char *base) +{ + unsigned long long value = 0; + memcpy(&value, + base + ds_cfg.info_data.offset, + ds_cfg.info_data.size); + return value; +} +static inline void set_info_data(char *base, unsigned long long value) +{ + memcpy(base + ds_cfg.info_data.offset, + &value, + ds_cfg.info_data.size); +} + + +int ds_allocate(void **dsp, size_t bts_size_in_records) +{ + size_t bts_size_in_bytes = 0; + void *bts = 0; + void *ds = 0; + + if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts) + return -EOPNOTSUPP; + + if (bts_size_in_records < 0) + return -EINVAL; + + bts_size_in_bytes = + bts_size_in_records * ds_cfg.sizeof_bts; + + if (bts_size_in_bytes <= 0) + return -EINVAL; + + bts = kzalloc(bts_size_in_bytes, GFP_KERNEL); + + if (!bts) + return -ENOMEM; + + ds = kzalloc(ds_cfg.sizeof_ds, GFP_KERNEL); + + if (!ds) { + kfree(bts); + return -ENOMEM; + } + + set_bts_buffer_base(ds, bts); + set_bts_index(ds, bts); + set_bts_absolute_maximum(ds, bts + bts_size_in_bytes); + set_bts_interrupt_threshold(ds, bts + bts_size_in_bytes + 1); + + *dsp = ds; + return 0; +} + +int ds_free(void **dsp) +{ + if (*dsp) + kfree(get_bts_buffer_base(*dsp)); + kfree(*dsp); + *dsp = 0; + + return 0; +} + +int ds_get_bts_size(void *ds) +{ + size_t size_in_bytes; + + if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts) + return -EOPNOTSUPP; + + size_in_bytes = + get_bts_absolute_maximum(ds) - + get_bts_buffer_base(ds); + + return size_in_bytes / ds_cfg.sizeof_bts; +} + +int ds_get_bts_index(void *ds) +{ + size_t index_offset_in_bytes; + + if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts) + return -EOPNOTSUPP; + + index_offset_in_bytes = + get_bts_index(ds) - + get_bts_buffer_base(ds); + + return index_offset_in_bytes / ds_cfg.sizeof_bts; +} + +int ds_read_bts(void *ds, size_t index, struct bts_struct *out) +{ + void *bts; + + if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts) + return -EOPNOTSUPP; + + if (index < 0) + return -EINVAL; + + if (index >= ds_get_bts_size(ds)) + return -EINVAL; + + bts = get_bts_buffer_base(ds); + bts = (char *)bts + (index * ds_cfg.sizeof_bts); + + memset(out, 0, sizeof(*out)); + if (get_from_ip(bts) == BTS_ESCAPE_ADDRESS) { + out->qualifier = get_info_type(bts); + out->variant.timestamp = get_info_data(bts); + } else { + out->qualifier = BTS_BRANCH; + out->variant.lbr.from_ip = get_from_ip(bts); + out->variant.lbr.to_ip = get_to_ip(bts); + } + + return 0; +} + +int ds_write_bts(void *ds, const struct bts_struct *in) +{ + void *bts; + + if (!ds_cfg.sizeof_ds || !ds_cfg.sizeof_bts) + return -EOPNOTSUPP; + + if (ds_get_bts_size(ds) <= 0) + return -ENXIO; + + bts = get_bts_index(ds); + + memset(bts, 0, ds_cfg.sizeof_bts); + switch (in->qualifier) { + case BTS_INVALID: + break; + + case BTS_BRANCH: + set_from_ip(bts, in->variant.lbr.from_ip); + set_to_ip(bts, in->variant.lbr.to_ip); + break; + + case BTS_TASK_ARRIVES: + case BTS_TASK_DEPARTS: + set_from_ip(bts, BTS_ESCAPE_ADDRESS); + set_info_type(bts, in->qualifier); + set_info_data(bts, in->variant.timestamp); + break; + + default: + return -EINVAL; + } + + bts = (char *)bts + ds_cfg.sizeof_bts; + if (bts >= get_bts_absolute_maximum(ds)) + bts = get_bts_buffer_base(ds); + set_bts_index(ds, bts); + + return 0; +} + +unsigned long ds_debugctl_mask(void) +{ + return ds_cfg.debugctl_mask; +} + +#ifdef __i386__ +static const struct ds_configuration ds_cfg_netburst = { + .sizeof_ds = 9 * 4, + .bts_buffer_base = { 0, 4 }, + .bts_index = { 4, 4 }, + .bts_absolute_maximum = { 8, 4 }, + .bts_interrupt_threshold = { 12, 4 }, + .sizeof_bts = 3 * 4, + .from_ip = { 0, 4 }, + .to_ip = { 4, 4 }, + .info_type = { 4, 1 }, + .info_data = { 5, 7 }, + .debugctl_mask = (1<<2)|(1<<3) +}; + +static const struct ds_configuration ds_cfg_pentium_m = { + .sizeof_ds = 9 * 4, + .bts_buffer_base = { 0, 4 }, + .bts_index = { 4, 4 }, + .bts_absolute_maximum = { 8, 4 }, + .bts_interrupt_threshold = { 12, 4 }, + .sizeof_bts = 3 * 4, + .from_ip = { 0, 4 }, + .to_ip = { 4, 4 }, + .info_type = { 4, 1 }, + .info_data = { 5, 7 }, + .debugctl_mask = (1<<6)|(1<<7) +}; +#endif /* _i386_ */ + +static const struct ds_configuration ds_cfg_core2 = { + .sizeof_ds = 9 * 8, + .bts_buffer_base = { 0, 8 }, + .bts_index = { 8, 8 }, + .bts_absolute_maximum = { 16, 8 }, + .bts_interrupt_threshold = { 24, 8 }, + .sizeof_bts = 3 * 8, + .from_ip = { 0, 8 }, + .to_ip = { 8, 8 }, + .info_type = { 8, 1 }, + .info_data = { 9, 7 }, + .debugctl_mask = (1<<6)|(1<<7)|(1<<9) +}; + +static inline void +ds_configure(const struct ds_configuration *cfg) +{ + ds_cfg = *cfg; +} + +void __cpuinit ds_init_intel(struct cpuinfo_x86 *c) +{ + switch (c->x86) { + case 0x6: + switch (c->x86_model) { +#ifdef __i386__ + case 0xD: + case 0xE: /* Pentium M */ + ds_configure(&ds_cfg_pentium_m); + break; +#endif /* _i386_ */ + case 0xF: /* Core2 */ + ds_configure(&ds_cfg_core2); + break; + default: + /* sorry, don't know about them */ + break; + } + break; + case 0xF: + switch (c->x86_model) { +#ifdef __i386__ + case 0x0: + case 0x1: + case 0x2: /* Netburst */ + ds_configure(&ds_cfg_netburst); + break; +#endif /* _i386_ */ + default: + /* sorry, don't know about them */ + break; + } + break; + default: + /* sorry, don't know about them */ + break; + } +} Index: linux-2.6-x86/include/asm-x86/ds.h =================================================================== --- /dev/null 1970-01-01 00:00:00.000000000 +0000 +++ linux-2.6-x86/include/asm-x86/ds.h 2007-12-04 18:18:55.%N +0100 @@ -0,0 +1,65 @@ +/* + * Debug Store (DS) support + * + * This provides a low-level interface to the hardware's Debug Store + * feature that is used for last branch recording (LBR) and + * precise-event based sampling (PEBS). + * + * Different architectures use a different DS layout/pointer size. + * The below functions therefore work on a void*. + * + * + * Since there is no user for PEBS, yet, only LBR (or branch + * trace store, BTS) is supported. + * + * + * Copyright (C) 2007 Intel Corporation. + * Markus Metzger , Dec 2007 + */ + +#ifndef _ASM_X86_DS_H +#define _ASM_X86_DS_H + +#include +#include + +struct cpuinfo_x86; + + +/* a branch trace record entry + * + * In order to unify the interface between various processor versions, + * we use the below data structure for all processors. + */ +enum bts_qualifier { + BTS_INVALID = 0, + BTS_BRANCH, + BTS_TASK_ARRIVES, + BTS_TASK_DEPARTS +}; + +struct bts_struct { + enum bts_qualifier qualifier; + union { + /* BTS_BRANCH */ + struct { + long from_ip; + long to_ip; + } lbr; + /* BTS_TASK_ARRIVES or + BTS_TASK_DEPARTS */ + unsigned long long timestamp; + } variant; +}; + + +extern int ds_allocate(void **, size_t); +extern int ds_free(void **); +extern int ds_get_bts_size(void *); +extern int ds_get_bts_index(void *); +extern int ds_read_bts(void *, size_t, struct bts_struct *); +extern int ds_write_bts(void *, const struct bts_struct *); +extern unsigned long ds_debugctl_mask(void); +extern void __cpuinit ds_init_intel(struct cpuinfo_x86 *c); + +#endif /* _ASM_X86_DS_H */