eBPF bug fixes (#14869)

collectors/ebpf.plugin/ebpf.c

@@ -1820,8 +1820,9 @@ void set_global_variables()
         ebpf_configured_log_dir = LOG_DIR;
 
     ebpf_nprocs = (int)sysconf(_SC_NPROCESSORS_ONLN);
-    if (ebpf_nprocs > NETDATA_MAX_PROCESSOR) {
+    if (ebpf_nprocs < 0) {
         ebpf_nprocs = NETDATA_MAX_PROCESSOR;
+        error("Cannot identify number of process, using default value %d", ebpf_nprocs);
     }
 
     isrh = get_redhat_release();

collectors/ebpf.plugin/ebpf_hardirq.c

@@ -129,11 +129,54 @@ static hardirq_static_val_t hardirq_static_vals[] = {
 // thread will write to netdata agent.
 static avl_tree_lock hardirq_pub;
 
-// tmp store for dynamic hard IRQ values we get from a per-CPU eBPF map.
-static hardirq_ebpf_val_t *hardirq_ebpf_vals = NULL;
+/*****************************************************************
+ *
+ *  ARAL SECTION
+ *
+ *****************************************************************/
+
+// ARAL vectors used to speed up processing
+ARAL *ebpf_aral_hardirq = NULL;
 
-// tmp store for static hard IRQ values we get from a per-CPU eBPF map.
-static hardirq_ebpf_static_val_t *hardirq_ebpf_static_vals = NULL;
+/**
+ * eBPF hardirq Aral init
+ *
+ * Initiallize array allocator that will be used when integration with apps is enabled.
+ */
+static inline void ebpf_hardirq_aral_init()
+{
+    ebpf_aral_hardirq = ebpf_allocate_pid_aral(NETDATA_EBPF_HARDIRQ_ARAL_NAME, sizeof(hardirq_val_t));
+}
+
+/**
+ * eBPF hardirq get
+ *
+ * Get a hardirq_val_t entry to be used with a specific IRQ.
+ *
+ * @return it returns the address on success.
+ */
+hardirq_val_t *ebpf_hardirq_get(void)
+{
+    hardirq_val_t *target = aral_mallocz(ebpf_aral_hardirq);
+    memset(target, 0, sizeof(hardirq_val_t));
+    return target;
+}
+
+/**
+ * eBPF hardirq release
+ *
+ * @param stat Release a target after usage.
+ */
+void ebpf_hardirq_release(hardirq_val_t *stat)
+{
+    aral_freez(ebpf_aral_hardirq, stat);
+}
+
+/*****************************************************************
+ *
+ *  EXIT FUNCTIONS
+ *
+ *****************************************************************/
 
 /**
  * Hardirq Free
@@ -151,9 +194,6 @@ static void ebpf_hardirq_free(ebpf_module_t *em)
     for (int i = 0; hardirq_tracepoints[i].class != NULL; i++) {
         ebpf_disable_tracepoint(&hardirq_tracepoints[i]);
     }
-    freez(hardirq_ebpf_vals);
-    freez(hardirq_ebpf_static_vals);
-
     pthread_mutex_lock(&ebpf_exit_cleanup);
     em->thread->enabled = NETDATA_THREAD_EBPF_STOPPED;
     pthread_mutex_unlock(&ebpf_exit_cleanup);
@@ -200,8 +240,82 @@ static int hardirq_val_cmp(void *a, void *b)
     }
 }
 
-static void hardirq_read_latency_map(int mapfd)
+/**
+ * Parse interrupts
+ *
+ * Parse /proc/interrupts to get names  used in metrics
+ *
+ * @param irq_name vector to store data.
+ * @param irq      irq value
+ *
+ * @return It returns 0 on success and -1 otherwise
+ */
+static int hardirq_parse_interrupts(char *irq_name, int irq)
 {
+    static procfile *ff = NULL;
+    static int cpus = -1;
+    if(unlikely(!ff)) {
+        char filename[FILENAME_MAX + 1];
+        snprintfz(filename, FILENAME_MAX, "%s%s", netdata_configured_host_prefix, "/proc/interrupts");
+        ff = procfile_open(filename, " \t:", PROCFILE_FLAG_DEFAULT);
+    }
+    if(unlikely(!ff))
+        return -1;
+
+    ff = procfile_readall(ff);
+    if(unlikely(!ff))
+        return -1; // we return 0, so that we will retry to open it next time
+
+    size_t words = procfile_linewords(ff, 0);
+    if(unlikely(cpus == -1)) {
+        uint32_t w;
+        cpus = 0;
+        for(w = 0; w < words ; w++) {
+            if(likely(strncmp(procfile_lineword(ff, 0, w), "CPU", 3) == 0))
+                cpus++;
+        }
+   }
+
+    size_t lines = procfile_lines(ff), l;
+    if(unlikely(!lines)) {
+        collector_error("Cannot read /proc/interrupts, zero lines reported.");
+        return -1;
+    }
+
+    for(l = 1; l < lines ;l++) {
+        words = procfile_linewords(ff, l);
+        if(unlikely(!words)) continue;
+        const char *id = procfile_lineword(ff, l, 0);
+        if (!isdigit(id[0]))
+            continue;
+
+        int cmp = str2i(id);
+        if (cmp != irq)
+            continue;
+
+        if(unlikely((uint32_t)(cpus + 2) < words)) {
+            const char *name = procfile_lineword(ff, l, words - 1);
+            // On some motherboards IRQ can have the same name, so we append IRQ id to differentiate.
+            snprintfz(irq_name, NETDATA_HARDIRQ_NAME_LEN - 1, "%d_%s", irq, name);
+        }
+    }
+
+    return 0;
+}
+
+/**
+ * Read Latency MAP
+ *
+ * Read data from kernel ring to user ring.
+ *
+ * @param mapfd hash map id.
+ *
+ * @return it returns 0 on success and -1 otherwise
+ */
+static int hardirq_read_latency_map(int mapfd)
+{
+    hardirq_ebpf_static_val_t hardirq_ebpf_vals[ebpf_nprocs + 1];
+
     hardirq_ebpf_key_t key = {};
     hardirq_ebpf_key_t next_key = {};
     hardirq_val_t search_v = {};
@@ -234,7 +348,7 @@ static void hardirq_read_latency_map(int mapfd)
         if (unlikely(v == NULL)) {
             // latency/name can only be added reliably at a later time.
             // when they're added, only then will we AVL insert.
-            v = callocz(1, sizeof(hardirq_val_t));
+            v = ebpf_hardirq_get();
             v->irq = key.irq;
             v->dim_exists = false;
 
@@ -246,22 +360,10 @@ static void hardirq_read_latency_map(int mapfd)
         // 2. the name is unfortunately *not* available on all CPU maps - only
         //    a single map contains the name, so we must find it. we only need
         //    to copy it though if the IRQ is new for us.
-        bool name_saved = false;
         uint64_t total_latency = 0;
         int i;
-        int end = (running_on_kernel < NETDATA_KERNEL_V4_15) ? 1 : ebpf_nprocs;
-        for (i = 0; i < end; i++) {
+        for (i = 0; i < ebpf_nprocs; i++) {
             total_latency += hardirq_ebpf_vals[i].latency/1000;
-
-            // copy name for new IRQs.
-            if (v_is_new && !name_saved && hardirq_ebpf_vals[i].name[0] != '\0') {
-                strncpyz(
-                    v->name,
-                    hardirq_ebpf_vals[i].name,
-                    NETDATA_HARDIRQ_NAME_LEN
-                );
-                name_saved = true;
-            }
         }
 
         // can now safely publish latency for existing IRQs.
@@ -269,6 +371,11 @@ static void hardirq_read_latency_map(int mapfd)
 
         // can now safely publish new IRQ.
         if (v_is_new) {
+            if (hardirq_parse_interrupts(v->name, v->irq)) {
+                ebpf_hardirq_release(v);
+                return -1;
+            }
+
             avl_t *check = avl_insert_lock(&hardirq_pub, (avl_t *)v);
             if (check != (avl_t *)v) {
                 error("Internal error, cannot insert the AVL tree.");
@@ -277,10 +384,14 @@ static void hardirq_read_latency_map(int mapfd)
 
         key = next_key;
     }
+
+    return 0;
 }
 
 static void hardirq_read_latency_static_map(int mapfd)
 {
+    hardirq_ebpf_static_val_t hardirq_ebpf_static_vals[ebpf_nprocs + 1];
+
     uint32_t i;
     for (i = 0; i < HARDIRQ_EBPF_STATIC_END; i++) {
         uint32_t map_i = hardirq_static_vals[i].idx;
@@ -302,11 +413,17 @@ static void hardirq_read_latency_static_map(int mapfd)
 
 /**
  * Read eBPF maps for hard IRQ.
+ *
+ * @return When it is not possible to parse /proc, it returns -1, on success it returns 0;
  */
-static void hardirq_reader()
+static int hardirq_reader()
 {
-    hardirq_read_latency_map(hardirq_maps[HARDIRQ_MAP_LATENCY].map_fd);
+    if (hardirq_read_latency_map(hardirq_maps[HARDIRQ_MAP_LATENCY].map_fd))
+        return -1;
+
     hardirq_read_latency_static_map(hardirq_maps[HARDIRQ_MAP_LATENCY_STATIC].map_fd);
+
+    return 0;
 }
 
 static void hardirq_create_charts(int update_every)
@@ -375,16 +492,8 @@ static inline void hardirq_write_static_dims()
 */
 static void hardirq_collector(ebpf_module_t *em)
 {
-    hardirq_ebpf_vals = callocz(
-        (running_on_kernel < NETDATA_KERNEL_V4_15) ? 1 : ebpf_nprocs,
-        sizeof(hardirq_ebpf_val_t)
-    );
-    hardirq_ebpf_static_vals = callocz(
-        (running_on_kernel < NETDATA_KERNEL_V4_15) ? 1 : ebpf_nprocs,
-        sizeof(hardirq_ebpf_static_val_t)
-    );
-
     avl_init_lock(&hardirq_pub, hardirq_val_cmp);
+    ebpf_hardirq_aral_init();
 
     // create chart and static dims.
     pthread_mutex_lock(&lock);
@@ -407,7 +516,9 @@ static void hardirq_collector(ebpf_module_t *em)
             continue;
 
         counter = 0;
-        hardirq_reader();
+        if (hardirq_reader())
+            break;
+
         pthread_mutex_lock(&lock);
 
         // write dims now for all hitherto discovered IRQs.

collectors/ebpf.plugin/ebpf_hardirq.h

@@ -3,6 +3,9 @@
 #ifndef NETDATA_EBPF_HARDIRQ_H
 #define NETDATA_EBPF_HARDIRQ_H 1
 
+#include <stdint.h>
+#include "libnetdata/avl/avl.h"
+
 /*****************************************************************
  *  copied from kernel-collectors repo, with modifications needed
  *  for inclusion here.
@@ -15,12 +18,6 @@ typedef struct hardirq_ebpf_key {
     int irq;
 } hardirq_ebpf_key_t;
 
-typedef struct hardirq_ebpf_val {
-    uint64_t latency;
-    uint64_t ts;
-    char name[NETDATA_HARDIRQ_NAME_LEN];
-} hardirq_ebpf_val_t;
-
 enum hardirq_ebpf_static {
     HARDIRQ_EBPF_STATIC_APIC_THERMAL,
     HARDIRQ_EBPF_STATIC_APIC_THRESHOLD,
@@ -46,6 +43,9 @@ typedef struct hardirq_ebpf_static_val {
  * below this is eBPF plugin-specific code.
  *****************************************************************/
 
+// ARAL Name
+#define NETDATA_EBPF_HARDIRQ_ARAL_NAME "ebpf_harddirq"
+
 #define NETDATA_EBPF_MODULE_NAME_HARDIRQ "hardirq"
 #define NETDATA_HARDIRQ_CONFIG_FILE "hardirq.conf"
 

collectors/ebpf.plugin/ebpf_oomkill.c

@@ -299,7 +299,7 @@ static void oomkill_collector(ebpf_module_t *em)
     int counter = update_every - 1;
     while (!ebpf_exit_plugin) {
         (void)heartbeat_next(&hb, USEC_PER_SEC);
-        if (!ebpf_exit_plugin || ++counter != update_every)
+        if (ebpf_exit_plugin || ++counter != update_every)
             continue;
 
         counter = 0;

libnetdata/ebpf/ebpf.c

@@ -366,20 +366,22 @@ static uint32_t ebpf_select_index(uint32_t kernels, int is_rhf, uint32_t kver)
  *     V - The kernel version in string format.
  *
  *  @param out       the vector where the name will be stored
- *  @param path
  *  @param len       the size of the out vector.
+ *  @param path      where the binaries are stored
  *  @param kver      the kernel version
  *  @param name      the eBPF program name.
  *  @param is_return is return or entry ?
  */
-static void ebpf_mount_name(char *out, size_t len, char *path, uint32_t kver, const char *name, int is_return)
+static void ebpf_mount_name(char *out, size_t len, char *path, uint32_t kver, const char *name,
+                            int is_return, int is_rhf)
 {
     char *version = ebpf_select_kernel_name(kver);
-    snprintfz(out, len, "%s/ebpf.d/%cnetdata_ebpf_%s.%s.o",
+    snprintfz(out, len, "%s/ebpf.d/%cnetdata_ebpf_%s.%s%s.o",
               path,
               (is_return) ? 'r' : 'p',
               name,
-              version);
+              version,
+              (is_rhf != -1) ? ".rhf" : "");
 }
 
 //----------------------------------------------------------------------------------------------------------------------
@@ -781,7 +783,7 @@ struct bpf_link **ebpf_load_program(char *plugins_dir, ebpf_module_t *em, int kv
 
     uint32_t idx = ebpf_select_index(em->kernels, is_rhf, kver);
 
-    ebpf_mount_name(lpath, 4095, plugins_dir, idx, em->thread_name, em->mode);
+    ebpf_mount_name(lpath, 4095, plugins_dir, idx, em->thread_name, em->mode, is_rhf);
 
     // When this function is called ebpf.plugin is using legacy code, so we should reset the variable
     em->load &= ~ NETDATA_EBPF_LOAD_METHODS;

netdata-installer.sh

@@ -1564,6 +1564,11 @@ remove_old_ebpf() {
     rm -f "${NETDATA_PREFIX}/usr/libexec/netdata/plugins.d/pnetdata_ebpf"*.?.*.o
   fi
 
+  # Remove old eBPF programs that did not have "rhf" suffix
+  if [ ! -f "${NETDATA_PREFIX}/usr/libexec/netdata/plugins.d/ebpf.d/pnetdata_ebpf_process.3.10.rhf.o" ]; then
+    rm -f "${NETDATA_PREFIX}/usr/libexec/netdata/plugins.d/ebpf.d/"*.o
+  fi
+
   # Remove old reject list from previous directory
   if [ -f "${NETDATA_PREFIX}/usr/lib/netdata/conf.d/ebpf_kernel_reject_list.txt" ]; then
     echo >&2 "Removing old ebpf_kernel_reject_list.txt."

packaging/ebpf-co-re.checksums

@@ -1 +1 @@
-d1864cd736d236aa3738152d86096529830822a26405a62fe164779949bb3658  netdata-ebpf-co-re-glibc-v1.1.0.tar.xz
+a50e649635cc2fe86c21a08334ee73451f08591ebbda8b5d0012c3b8fad2cc1e  netdata-ebpf-co-re-glibc-v1.1.2.tar.xz

packaging/ebpf-co-re.version

@@ -1 +1 @@
-v1.1.0
+v1.1.2

packaging/ebpf.checksums

@@ -1,3 +1,3 @@
-7f28bb61b1e9fdac59e5f8f041502c54f319048c1cf4adaa96ace3360f55a80e  ./netdata-kernel-collector-glibc-v1.1.0.tar.xz
-5d927deadac9a4a5bc8a5be386aec2ea4f9b8335e60eadf375b11e7656404270  ./netdata-kernel-collector-musl-v1.1.0.tar.xz
-0d8825b77b8ba20e10b6e24f15c1d65a43f1c47dced93798839adc789f1427d3  ./netdata-kernel-collector-static-v1.1.0.tar.xz
+597a20895bbedcf87528b08fa9057426bd3c7638aa1ffac94f8987a90634513d  ./netdata-kernel-collector-glibc-v1.1.2.tar.xz
+25db2232b75bdb7fc6e10db870c3a3290f52ecfcdcf546d0e51947f2a4c17ccf  ./netdata-kernel-collector-musl-v1.1.2.tar.xz
+1d60425f5e8c6e30b3be86028dfc62c16022d8fe561e4c21c84cf6e8b998cd7d  ./netdata-kernel-collector-static-v1.1.2.tar.xz

packaging/ebpf.version

@@ -1 +1 @@
-v1.1.0
+v1.1.2