lkml.org 
[lkml]   [2011]   [Jan]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[PATCH] ras: Add RAS daemon
Date
From: Borislav Petkov <borislav.petkov@amd.com>

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
---
tools/Makefile | 4 +
tools/ras/Makefile | 16 +++
tools/ras/rasd.c | 311 ++++++++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 331 insertions(+), 0 deletions(-)
create mode 100644 tools/ras/Makefile
create mode 100644 tools/ras/rasd.c

diff --git a/tools/Makefile b/tools/Makefile
index 71dce04..a012fa3 100644
--- a/tools/Makefile
+++ b/tools/Makefile
@@ -22,11 +22,15 @@ liblk: .FORCE
liblkperf:
$(QUIET_SUBDIR0)lib/perf/ $(QUIET_SUBDIR1)

+ras: libtrace liblk liblkperf .FORCE
+ $(QUIET_SUBDIR0)ras/ $(QUIET_SUBDIR1)
+
clean:
$(QUIET_SUBDIR0)lib/trace/ $(QUIET_SUBDIR1) clean
$(QUIET_SUBDIR0)lib/lk/ $(QUIET_SUBDIR1) clean
$(QUIET_SUBDIR0)lib/perf/ $(QUIET_SUBDIR1) clean
$(QUIET_SUBDIR0)perf/ $(QUIET_SUBDIR1) clean
+ $(QUIET_SUBDIR0)ras/ $(QUIET_SUBDIR1) clean


.PHONY: clean .FORCE
diff --git a/tools/ras/Makefile b/tools/ras/Makefile
new file mode 100644
index 0000000..0b2f458
--- /dev/null
+++ b/tools/ras/Makefile
@@ -0,0 +1,16 @@
+include ../scripts/Makefile.lib
+
+CFLAGS = -ggdb3 -Wall -Wextra -std=gnu99 $(CFLAGS_OPTIMIZE) -D_FORTIFY_SOURCE=2 -DNO_NEWT_SUPPORT $(EXTRA_WARNINGS) $(EXTRA_CFLAGS)
+ALL_CFLAGS = $(CFLAGS) $(BASIC_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64
+ALL_LDFLAGS = $(LDFLAGS)
+
+RASLIBS=$(LIB_OUTPUT)libtrace.a $(LIB_OUTPUT)liblk.a $(LIB_OUTPUT)liblkperf.a
+
+rasd: rasd.o
+ $(QUIET_CC)$(CC) $(ALL_CFLAGS) -o $@ $^ $(RASLIBS)
+
+%.o: %.c
+ $(QUIET_CC)$(CC) $(ALL_CFLAGS) -c $<
+
+clean:
+ rm -rf *.o rasd
diff --git a/tools/ras/rasd.c b/tools/ras/rasd.c
new file mode 100644
index 0000000..a8b14b7
--- /dev/null
+++ b/tools/ras/rasd.c
@@ -0,0 +1,311 @@
+/*
+ * Linux RAS daemon.
+ *
+ * Initial code reused from Linux Daemon Writing HOWTO
+ */
+
+#include <errno.h>
+#include <fcntl.h>
+#include <stdio.h>
+#include <stdint.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <sys/stat.h>
+#include <sys/mman.h>
+#include <sys/types.h>
+
+#include <lk/util.h>
+#include <lk/debugfs.h>
+#include <perf/mmap.h>
+#include <perf/util.h>
+#include <trace/trace-event.h>
+
+#include "../../arch/x86/include/asm/mce.h"
+
+#ifdef DEBUG
+#define dbg(fmt, args...) \
+ fprintf(stderr, "DBG %s: " fmt "\n", __func__, ##args)
+#else
+#define dbg(fmt, args...) do { } while (0)
+#endif
+
+#define MMAP_PAGES 128
+#define MCE_TP "mce/mce_record"
+
+#define PFX "rasd: "
+
+static int fds[MAX_NR_CPUS];
+static struct mmap_data mmaps[MAX_NR_CPUS];
+static struct event *mce_event;
+static struct mce m;
+static const char *dfs_root;
+
+static int nr_cpus;
+static unsigned int page_size;
+
+const char *logf_path = "/var/log/ras.log";
+
+static unsigned long long read_file(const char *file, void *buf)
+{
+ unsigned long long size = 0;
+ int fd, r;
+
+ fd = open(file, O_RDONLY);
+ if (fd < 0)
+ die("Can't read '%s'", file);
+
+ do {
+ r = read(fd, buf, BUFSIZ);
+ if (r > 0)
+ size += r;
+ } while (r > 0);
+
+ close(fd);
+
+ return size;
+}
+
+static void parse_mce_event(void)
+{
+ struct stat st;
+ char *format_path, *format_buf, *path;
+ int fsize, err = 0;
+
+ path = get_tracing_file("events");
+
+ dbg("Got %s", path);
+
+ format_path = malloc_or_die(MAXPATHLEN + sizeof(MCE_TP) + 10);
+ sprintf(format_path, "%s/%s/format", path, MCE_TP);
+
+ err = stat(format_path, &st);
+ if (err < 0)
+ die("accessing %s", format_path);
+
+ dbg("Format access %s ok", format_path);
+
+ fsize = get_filesize(format_path);
+ dbg("Format file size: %d", fsize);
+
+ format_buf = malloc_or_die(fsize);
+ if (!format_buf)
+ die("allocating format buffer");
+
+ if (!read_file(format_path, format_buf))
+ die("reading in format file");
+
+ dbg("Format file contents:\n%s", format_buf);
+
+ init_input_buf(format_buf, fsize);
+
+ mce_event = alloc_event();
+ if (!mce_event)
+ die("Cannot alloc mce_event");
+
+ mce_event->name = event_read_name();
+ if (!mce_event->name)
+ error("no event name");
+
+ mce_event->id = event_read_id();
+ if (mce_event->id < 0)
+ error(PFX "failed to read event id");
+
+ if (event_read_format(mce_event))
+ die("parsing event");
+
+
+ free(format_buf);
+ free(format_path);
+ free(path);
+}
+
+static void fill_mce_data(void *vbuf, size_t buflen)
+{
+ struct format_field *field;
+ char *buf = vbuf;
+ u32 tp_len;
+#ifdef DEBUG
+ unsigned i;
+#endif
+
+ if (!buflen)
+ return;
+
+#ifdef DEBUG
+ dbg("buflen %lu", buflen);
+
+ for (i = 0; i < buflen; i++) {
+
+ if (!(i & 0xf) && i)
+ printf("\n");
+
+ printf("0x%2.2x ", *(unsigned char *)(buf + i));
+ }
+#endif
+
+ /* skip event header for now, u32 size inclusive */
+ buf += sizeof(struct perf_event_header);
+ buflen -= sizeof(struct perf_event_header) + 4;
+
+ tp_len = *(u32 *)buf;
+
+ if (tp_len != buflen)
+ warning("bufsize mismatch: %lu <-> %u (tp)\n", buflen, tp_len);
+
+ /* skip size */
+ buf += 4;
+
+ for (field = mce_event->format.fields; field; field = field->next) {
+ if ((size_t)(field->offset + field->size) > buflen)
+ warning("MCE buf truncated? (off: %d <-> buflen: %lu)",
+ field->offset, buflen);
+
+ dbg("field %s, offset: %d", field->name, field->offset);
+
+ if (!strncmp(field->name, "bank", 4))
+ m.bank = *(u8 *)(buf + field->offset);
+ else if (!strncmp(field->name, "status", 6))
+ m.status = *(u64 *)(buf + field->offset);
+ else if (!strncmp(field->name, "addr", 4))
+ m.addr = *(u64 *)(buf + field->offset);
+ else if (!strncmp(field->name, "misc", 4))
+ m.misc = *(u64 *)(buf + field->offset);
+ else if (!strncmp(field->name, "ip", 2))
+ m.ip = *(u64 *)(buf + field->offset);
+ else if (!strncmp(field->name, "cs", 2))
+ m.cs = *(u8 *)(buf + field->offset);
+ else if (!strncmp(field->name, "tsc", 3))
+ m.tsc = *(u64 *)(buf + field->offset);
+ else if (!strncmp(field->name, "cpu", 3))
+ m.cpu = *(u8 *)(buf + field->offset);
+ else
+ warning("skipping %s", field->name);
+ }
+}
+
+static int ras_init(void)
+{
+ int cpu;
+
+ fprintf(stderr, PFX "Starting daemon.\n");
+
+ page_size = sysconf(_SC_PAGE_SIZE);
+
+ dfs_root = debugfs_mount(NULL);
+ if (!dfs_root) {
+ error("Cannot mount debugfs, exiting... ");
+ return 1;
+ }
+
+ nr_cpus = sysconf(_SC_NPROCESSORS_CONF);
+ if (nr_cpus < 0) {
+ error("Cannot get # CPUs, exiting... ");
+ return 1;
+ }
+
+ parse_mce_event();
+ assert(mce_event);
+
+ for (cpu = 0; cpu < nr_cpus; cpu++) {
+ char dfs_path[MAXPATHLEN];
+
+ snprintf(dfs_path, MAXPATHLEN, "%s/%s%d",
+ dfs_root, MCE_TP, cpu);
+
+ dbg("dfs_path: %s", dfs_path);
+
+ fds[cpu] = open(dfs_path, O_RDWR, O_NONBLOCK);
+ if (fds[cpu] < 0) {
+ error("open perf event on cpu %d\n", cpu);
+ return 1;
+ } else
+ dbg("cpu %d, fd %d", cpu, fds[cpu]);
+ }
+
+ for (cpu = 0; cpu < nr_cpus; cpu++) {
+ mmaps[cpu].prev = 0;
+ mmaps[cpu].mask = MMAP_PAGES*page_size - 1;
+ mmaps[cpu].base = mmap(NULL, (MMAP_PAGES + 1) * page_size,
+ PROT_READ | PROT_WRITE, MAP_SHARED,
+ fds[cpu], 0);
+
+ if (mmaps[cpu].base == MAP_FAILED) {
+ error("cannot mmap: %s (%d).", strerror(errno), errno);
+ return 1;
+ }
+ }
+
+ return 0;
+}
+
+static void ras_exit(void)
+{
+ free(mce_event);
+}
+
+int main(void)
+{
+ pid_t pid, sid;
+ FILE *logfile = NULL;
+ int err = 0;
+
+ pid = fork();
+ if (pid < 0) {
+ error(PFX "Error forking daemon thread.");
+ exit(EXIT_FAILURE);
+ }
+
+ /* parent can disappear now */
+ if (pid > 0)
+ exit(EXIT_SUCCESS);
+
+ umask(0);
+
+ sid = setsid();
+ if (sid < 0) {
+ error(PFX "Error creating session.");
+ exit(EXIT_FAILURE);
+ }
+
+ if (chdir("/") < 0) {
+ error(PFX "Error chdir to /");
+ exit(EXIT_FAILURE);
+ }
+
+ logfile = fopen(logf_path, "a");
+ if (!logfile) {
+ error(PFX "Error opening logs: %s\n", strerror(errno));
+ err = errno;
+ goto exit;
+ }
+
+ close(STDIN_FILENO);
+ close(STDOUT_FILENO);
+ close(STDERR_FILENO);
+
+ if (ras_init()) {
+ err = -EINVAL;
+ goto cleanup;
+ }
+
+ while (1) {
+
+ if (mmap_read_all(mmaps, nr_cpus, fill_mce_data)) {
+ fprintf(logfile,
+ "Got MCE, cpu: %d, status: 0x%016llx, addr: 0x%016llx\n",
+ m.cpu, m.status, m.addr);
+ fflush(logfile);
+ }
+
+ sleep(30);
+ }
+
+ ras_exit();
+
+cleanup:
+ fclose(logfile);
+
+exit:
+ return err;
+
+}
--
1.7.4.rc2


\
 
 \ /
  Last update: 2011-01-21 16:11    [W:0.079 / U:2.568 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site