Adding upstream version 2.4.56.upstream/2.4.56

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 1087 insertions, 0 deletions

diff --git a/modules/cache/mod_socache_shmcb.c b/modules/cache/mod_socache_shmcb.c
new file mode 100644
index 0000000..4727961
--- /dev/null
+++ b/modules/cache/mod_socache_shmcb.c
@@ -0,0 +1,1087 @@
+/* Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#include "httpd.h"
+#include "http_log.h"
+#include "http_request.h"
+#include "http_protocol.h"
+#include "http_config.h"
+#include "mod_status.h"
+
+#include "apr.h"
+#include "apr_strings.h"
+#include "apr_time.h"
+#include "apr_shm.h"
+#define APR_WANT_STRFUNC
+#include "apr_want.h"
+#include "apr_general.h"
+
+#if APR_HAVE_LIMITS_H
+#include <limits.h>
+#endif
+
+#include "ap_socache.h"
+
+/* XXX Unfortunately, there are still many unsigned ints in use here, so we
+ * XXX cannot allow more than UINT_MAX. Since some of the ints are exposed in
+ * XXX public interfaces, a simple search and replace is not enough.
+ * XXX It should be possible to extend that so that the total cache size can
+ * XXX be APR_SIZE_MAX and only the object size needs to be smaller than
+ * XXX UINT_MAX.
+ */
+#define SHMCB_MAX_SIZE (UINT_MAX<APR_SIZE_MAX ? UINT_MAX : APR_SIZE_MAX)
+
+#define DEFAULT_SHMCB_PREFIX "socache-shmcb-"
+
+#define DEFAULT_SHMCB_SUFFIX ".cache"
+
+#define ALIGNED_HEADER_SIZE APR_ALIGN_DEFAULT(sizeof(SHMCBHeader))
+#define ALIGNED_SUBCACHE_SIZE APR_ALIGN_DEFAULT(sizeof(SHMCBSubcache))
+#define ALIGNED_INDEX_SIZE APR_ALIGN_DEFAULT(sizeof(SHMCBIndex))
+
+/*
+ * Header structure - the start of the shared-mem segment
+ */
+typedef struct {
+    /* Stats for cache operations */
+    unsigned long stat_stores;
+    unsigned long stat_replaced;
+    unsigned long stat_expiries;
+    unsigned long stat_scrolled;
+    unsigned long stat_retrieves_hit;
+    unsigned long stat_retrieves_miss;
+    unsigned long stat_removes_hit;
+    unsigned long stat_removes_miss;
+    /* Number of subcaches */
+    unsigned int subcache_num;
+    /* How many indexes each subcache's queue has */
+    unsigned int index_num;
+    /* How large each subcache is, including the queue and data */
+    unsigned int subcache_size;
+    /* How far into each subcache the data area is (optimisation) */
+    unsigned int subcache_data_offset;
+    /* How large the data area in each subcache is (optimisation) */
+    unsigned int subcache_data_size;
+} SHMCBHeader;
+
+/*
+ * Subcache structure - the start of each subcache, followed by
+ * indexes then data
+ */
+typedef struct {
+    /* The start position and length of the cyclic buffer of indexes */
+    unsigned int idx_pos, idx_used;
+    /* Same for the data area */
+    unsigned int data_pos, data_used;
+} SHMCBSubcache;
+
+/*
+ * Index structure - each subcache has an array of these
+ */
+typedef struct {
+    /* absolute time this entry expires */
+    apr_time_t expires;
+    /* location within the subcache's data area */
+    unsigned int data_pos;
+    /* size (most logic ignores this, we keep it only to minimise memcpy) */
+    unsigned int data_used;
+    /* length of the used data which contains the id */
+    unsigned int id_len;
+    /* Used to mark explicitly-removed socache entries */
+    unsigned char removed;
+} SHMCBIndex;
+
+struct ap_socache_instance_t {
+    apr_pool_t *pool;
+    const char *data_file;
+    apr_size_t shm_size;
+    apr_shm_t *shm;
+    SHMCBHeader *header;
+};
+
+/* The SHM data segment is of fixed size and stores data as follows.
+ *
+ *   [ SHMCBHeader | Subcaches ]
+ *
+ * The SHMCBHeader header structure stores metadata concerning the
+ * cache and the contained subcaches.
+ *
+ * Subcaches is a hash table of header->subcache_num SHMCBSubcache
+ * structures.  The hash table is indexed by SHMCB_MASK(id). Each
+ * SHMCBSubcache structure has a fixed size (header->subcache_size),
+ * which is determined at creation time, and looks like the following:
+ *
+ *   [ SHMCBSubcache | Indexes | Data ]
+ *
+ * Each subcache is prefixed by the SHMCBSubcache structure.
+ *
+ * The subcache's "Data" segment is a single cyclic data buffer, of
+ * total size header->subcache_data_size; data inside is referenced
+ * using byte offsets. The offset marking the beginning of the cyclic
+ * buffer is subcache->data_pos; the buffer's length is
+ * subcache->data_used.
+ *
+ * "Indexes" is an array of header->index_num SHMCBIndex structures,
+ * which is used as a cyclic queue; subcache->idx_pos gives the array
+ * index of the first in use, subcache->idx_used gives the number in
+ * use.  Both ->idx_* values have a range of [0, header->index_num)
+ *
+ * Each in-use SHMCBIndex structure represents a single cached object.
+ * The ID and data segment are stored consecutively in the subcache's
+ * cyclic data buffer.  The "Data" segment can thus be seen to
+ * look like this, for example
+ *
+ * offset:  [ 0     1     2     3     4     5     6    ...
+ * contents:[ ID1   Data1       ID2   Data2       ID3  ...
+ *
+ * where the corresponding indices would look like:
+ *
+ * idx1 = { data_pos = 0, data_used = 3, id_len = 1, ...}
+ * idx2 = { data_pos = 3, data_used = 3, id_len = 1, ...}
+ * ...
+ */
+
+/* This macro takes a pointer to the header and a zero-based index and returns
+ * a pointer to the corresponding subcache. */
+#define SHMCB_SUBCACHE(pHeader, num) \
+                (SHMCBSubcache *)(((unsigned char *)(pHeader)) + \
+                        ALIGNED_HEADER_SIZE + \
+                        (num) * ((pHeader)->subcache_size))
+
+/* This macro takes a pointer to the header and an id and returns a
+ * pointer to the corresponding subcache. */
+#define SHMCB_MASK(pHeader, id) \
+                SHMCB_SUBCACHE((pHeader), *(id) & ((pHeader)->subcache_num - 1))
+
+/* This macro takes the same params as the last, generating two outputs for use
+ * in ap_log_error(...). */
+#define SHMCB_MASK_DBG(pHeader, id) \
+                *(id), (*(id) & ((pHeader)->subcache_num - 1))
+
+/* This macro takes a pointer to a subcache and a zero-based index and returns
+ * a pointer to the corresponding SHMCBIndex. */
+#define SHMCB_INDEX(pSubcache, num) \
+                (SHMCBIndex *)(((unsigned char *)pSubcache) + \
+                        ALIGNED_SUBCACHE_SIZE + \
+                        (num) * ALIGNED_INDEX_SIZE)
+
+/* This macro takes a pointer to the header and a subcache and returns a
+ * pointer to the corresponding data area. */
+#define SHMCB_DATA(pHeader, pSubcache) \
+                ((unsigned char *)(pSubcache) + (pHeader)->subcache_data_offset)
+
+/*
+ * Cyclic functions - assists in "wrap-around"/modulo logic
+ */
+
+/* Addition modulo 'mod' */
+#define SHMCB_CYCLIC_INCREMENT(val,inc,mod) \
+                (((val) + (inc)) % (mod))
+
+/* Subtraction (or "distance between") modulo 'mod' */
+#define SHMCB_CYCLIC_SPACE(val1,val2,mod) \
+                ((val2) >= (val1) ? ((val2) - (val1)) : \
+                        ((val2) + (mod) - (val1)))
+
+/* A "normal-to-cyclic" memcpy. */
+static void shmcb_cyclic_ntoc_memcpy(unsigned int buf_size, unsigned char *data,
+                                     unsigned int dest_offset, const unsigned char *src,
+                                     unsigned int src_len)
+{
+    if (dest_offset + src_len < buf_size)
+        /* It be copied all in one go */
+        memcpy(data + dest_offset, src, src_len);
+    else {
+        /* Copy the two splits */
+        memcpy(data + dest_offset, src, buf_size - dest_offset);
+        memcpy(data, src + buf_size - dest_offset,
+               src_len + dest_offset - buf_size);
+    }
+}
+
+/* A "cyclic-to-normal" memcpy. */
+static void shmcb_cyclic_cton_memcpy(unsigned int buf_size, unsigned char *dest,
+                                     const unsigned char *data, unsigned int src_offset,
+                                     unsigned int src_len)
+{
+    if (src_offset + src_len < buf_size)
+        /* It be copied all in one go */
+        memcpy(dest, data + src_offset, src_len);
+    else {
+        /* Copy the two splits */
+        memcpy(dest, data + src_offset, buf_size - src_offset);
+        memcpy(dest + buf_size - src_offset, data,
+               src_len + src_offset - buf_size);
+    }
+}
+
+/* A memcmp against a cyclic data buffer.  Compares SRC of length
+ * SRC_LEN against the contents of cyclic buffer DATA (which is of
+ * size BUF_SIZE), starting at offset DEST_OFFSET. Got that?  Good. */
+static int shmcb_cyclic_memcmp(unsigned int buf_size, unsigned char *data,
+                               unsigned int dest_offset,
+                               const unsigned char *src,
+                               unsigned int src_len)
+{
+    if (dest_offset + src_len < buf_size)
+        /* It be compared all in one go */
+        return memcmp(data + dest_offset, src, src_len);
+    else {
+        /* Compare the two splits */
+        int diff;
+
+        diff = memcmp(data + dest_offset, src, buf_size - dest_offset);
+        if (diff) {
+            return diff;
+        }
+        return memcmp(data, src + buf_size - dest_offset,
+                      src_len + dest_offset - buf_size);
+    }
+}
+
+
+/* Prototypes for low-level subcache operations */
+static void shmcb_subcache_expire(server_rec *, SHMCBHeader *, SHMCBSubcache *,
+                                  apr_time_t);
+/* Returns zero on success, non-zero on failure. */
+static int shmcb_subcache_store(server_rec *s, SHMCBHeader *header,
+                                SHMCBSubcache *subcache,
+                                unsigned char *data, unsigned int data_len,
+                                const unsigned char *id, unsigned int id_len,
+                                apr_time_t expiry);
+/* Returns zero on success, non-zero on failure. */
+static int shmcb_subcache_retrieve(server_rec *, SHMCBHeader *, SHMCBSubcache *,
+                                   const unsigned char *id, unsigned int idlen,
+                                   unsigned char *data, unsigned int *datalen);
+/* Returns zero on success, non-zero on failure. */
+static int shmcb_subcache_remove(server_rec *, SHMCBHeader *, SHMCBSubcache *,
+                                 const unsigned char *, unsigned int);
+
+/* Returns result of the (iterator)() call, zero is success (continue) */
+static apr_status_t shmcb_subcache_iterate(ap_socache_instance_t *instance,
+                                           server_rec *s,
+                                           void *userctx,
+                                           SHMCBHeader *header,
+                                           SHMCBSubcache *subcache,
+                                           ap_socache_iterator_t *iterator,
+                                           unsigned char **buf,
+                                           apr_size_t *buf_len,
+                                           apr_pool_t *pool,
+                                           apr_time_t now);
+
+/*
+ * High-Level "handlers" as per ssl_scache.c
+ * subcache internals are deferred to shmcb_subcache_*** functions lower down
+ */
+
+static const char *socache_shmcb_create(ap_socache_instance_t **context,
+                                        const char *arg,
+                                        apr_pool_t *tmp, apr_pool_t *p)
+{
+    ap_socache_instance_t *ctx;
+    char *path, *cp, *cp2;
+
+    /* Allocate the context. */
+    *context = ctx = apr_pcalloc(p, sizeof *ctx);
+    ctx->pool = p;
+
+    ctx->shm_size  = 1024*512; /* 512KB */
+
+    if (!arg || *arg == '\0') {
+        /* Use defaults. */
+        return NULL;
+    }
+
+    ctx->data_file = path = ap_server_root_relative(p, arg);
+
+    cp = strrchr(path, '(');
+    cp2 = path + strlen(path) - 1;
+    if (cp) {
+        char *endptr;
+        if (*cp2 != ')') {
+            return "Invalid argument: no closing parenthesis or cache size "
+                   "missing after pathname with parenthesis";
+        }
+        *cp++ = '\0';
+        *cp2  = '\0';
+
+
+        ctx->shm_size = strtol(cp, &endptr, 10);
+        if (endptr != cp2) {
+            return "Invalid argument: cache size not numerical";
+        }
+
+        if (ctx->shm_size < 8192) {
+            return "Invalid argument: size has to be >= 8192 bytes";
+
+        }
+
+        if (ctx->shm_size >= SHMCB_MAX_SIZE) {
+            return apr_psprintf(tmp, "Invalid argument: size has "
+                    "to be < %" APR_SIZE_T_FMT " bytes on this platform",
+                    SHMCB_MAX_SIZE);
+        }
+    }
+    else if (cp2 >= path && *cp2 == ')') {
+        return "Invalid argument: no opening parenthesis";
+    }
+
+    return NULL;
+}
+
+static apr_status_t socache_shmcb_cleanup(void *arg)
+{
+    ap_socache_instance_t *ctx = arg;
+    if (ctx->shm) {
+        apr_shm_destroy(ctx->shm);
+        ctx->shm = NULL;
+    }
+    return APR_SUCCESS;
+}
+
+static apr_status_t socache_shmcb_init(ap_socache_instance_t *ctx,
+                                       const char *namespace,
+                                       const struct ap_socache_hints *hints,
+                                       server_rec *s, apr_pool_t *p)
+{
+    void *shm_segment;
+    apr_size_t shm_segsize;
+    apr_status_t rv;
+    SHMCBHeader *header;
+    unsigned int num_subcache, num_idx, loop;
+    apr_size_t avg_obj_size, avg_id_len;
+
+    /* Create shared memory segment */
+    if (ctx->data_file == NULL) {
+        const char *path = apr_pstrcat(p, DEFAULT_SHMCB_PREFIX, namespace,
+                                       DEFAULT_SHMCB_SUFFIX, NULL);
+
+        ctx->data_file = ap_runtime_dir_relative(p, path);
+    }
+
+    /* Use anonymous shm by default, fall back on name-based. */
+    rv = apr_shm_create(&ctx->shm, ctx->shm_size, NULL, p);
+    if (APR_STATUS_IS_ENOTIMPL(rv)) {
+        /* If anon shm isn't supported, fail if no named file was
+         * configured successfully; the ap_server_root_relative call
+         * above will return NULL for invalid paths. */
+        if (ctx->data_file == NULL) {
+            ap_log_error(APLOG_MARK, APLOG_ERR, 0, s, APLOGNO(00818)
+                         "Could not use anonymous shm for '%s' cache",
+                         namespace);
+            ctx->shm = NULL;
+            return APR_EINVAL;
+        }
+
+        /* For a name-based segment, remove it first in case of a
+         * previous unclean shutdown. */
+        apr_shm_remove(ctx->data_file, p);
+
+        rv = apr_shm_create(&ctx->shm, ctx->shm_size, ctx->data_file, p);
+    }
+
+    if (rv != APR_SUCCESS) {
+        ap_log_error(APLOG_MARK, APLOG_ERR, rv, s, APLOGNO(00819)
+                     "Could not allocate shared memory segment for shmcb "
+                     "socache");
+        ctx->shm = NULL;
+        return rv;
+    }
+    apr_pool_cleanup_register(ctx->pool, ctx, socache_shmcb_cleanup,
+                              apr_pool_cleanup_null); 
+
+    shm_segment = apr_shm_baseaddr_get(ctx->shm);
+    shm_segsize = apr_shm_size_get(ctx->shm);
+    if (shm_segsize < (5 * ALIGNED_HEADER_SIZE)) {
+        /* the segment is ridiculously small, bail out */
+        ap_log_error(APLOG_MARK, APLOG_ERR, 0, s, APLOGNO(00820)
+                     "shared memory segment too small");
+        return APR_ENOSPC;
+    }
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00821)
+                 "shmcb_init allocated %" APR_SIZE_T_FMT
+                 " bytes of shared memory",
+                 shm_segsize);
+    /* Discount the header */
+    shm_segsize -= ALIGNED_HEADER_SIZE;
+    /* Select index size based on average object size hints, if given. */
+    avg_obj_size = hints && hints->avg_obj_size ? hints->avg_obj_size : 150;
+    avg_id_len = hints && hints->avg_id_len ? hints->avg_id_len : 30;
+    num_idx = (shm_segsize) / (avg_obj_size + avg_id_len);
+    num_subcache = 256;
+    while ((num_idx / num_subcache) < (2 * num_subcache))
+        num_subcache /= 2;
+    num_idx /= num_subcache;
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00822)
+                 "for %" APR_SIZE_T_FMT " bytes (%" APR_SIZE_T_FMT
+                 " including header), recommending %u subcaches, "
+                 "%u indexes each", shm_segsize,
+                 shm_segsize + ALIGNED_HEADER_SIZE,
+                 num_subcache, num_idx);
+    if (num_idx < 5) {
+        /* we're still too small, bail out */
+        ap_log_error(APLOG_MARK, APLOG_ERR, 0, s, APLOGNO(00823)
+                     "shared memory segment too small");
+        return APR_ENOSPC;
+    }
+    /* OK, we're sorted */
+    ctx->header = header = shm_segment;
+    header->stat_stores = 0;
+    header->stat_replaced = 0;
+    header->stat_expiries = 0;
+    header->stat_scrolled = 0;
+    header->stat_retrieves_hit = 0;
+    header->stat_retrieves_miss = 0;
+    header->stat_removes_hit = 0;
+    header->stat_removes_miss = 0;
+    header->subcache_num = num_subcache;
+    /* Convert the subcache size (in bytes) to a value that is suitable for
+     * structure alignment on the host platform, by rounding down if necessary. */
+    header->subcache_size = (size_t)(shm_segsize / num_subcache);
+    if (header->subcache_size != APR_ALIGN_DEFAULT(header->subcache_size)) {
+        header->subcache_size = APR_ALIGN_DEFAULT(header->subcache_size) -
+                                APR_ALIGN_DEFAULT(1);
+    }
+    header->subcache_data_offset = ALIGNED_SUBCACHE_SIZE +
+                                   num_idx * ALIGNED_INDEX_SIZE;
+    header->subcache_data_size = header->subcache_size -
+                                 header->subcache_data_offset;
+    header->index_num = num_idx;
+
+    /* Output trace info */
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00824)
+                 "shmcb_init_memory choices follow");
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00825)
+                 "subcache_num = %u", header->subcache_num);
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00826)
+                 "subcache_size = %u", header->subcache_size);
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00827)
+                 "subcache_data_offset = %u", header->subcache_data_offset);
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00828)
+                 "subcache_data_size = %u", header->subcache_data_size);
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00829)
+                 "index_num = %u", header->index_num);
+    /* The header is done, make the caches empty */
+    for (loop = 0; loop < header->subcache_num; loop++) {
+        SHMCBSubcache *subcache = SHMCB_SUBCACHE(header, loop);
+        subcache->idx_pos = subcache->idx_used = 0;
+        subcache->data_pos = subcache->data_used = 0;
+    }
+    ap_log_error(APLOG_MARK, APLOG_INFO, 0, s, APLOGNO(00830)
+                 "Shared memory socache initialised");
+    /* Success ... */
+
+    return APR_SUCCESS;
+}
+
+static void socache_shmcb_destroy(ap_socache_instance_t *ctx, server_rec *s)
+{
+    if (ctx) {
+        apr_pool_cleanup_run(ctx->pool, ctx, socache_shmcb_cleanup); 
+    }
+}
+
+static apr_status_t socache_shmcb_store(ap_socache_instance_t *ctx,
+                                        server_rec *s, const unsigned char *id,
+                                        unsigned int idlen, apr_time_t expiry,
+                                        unsigned char *encoded,
+                                        unsigned int len_encoded,
+                                        apr_pool_t *p)
+{
+    SHMCBHeader *header = ctx->header;
+    SHMCBSubcache *subcache = SHMCB_MASK(header, id);
+    int tryreplace;
+
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00831)
+                 "socache_shmcb_store (0x%02x -> subcache %d)",
+                 SHMCB_MASK_DBG(header, id));
+    /* XXX: Says who?  Why shouldn't this be acceptable, or padded if not? */
+    if (idlen < 4) {
+        ap_log_error(APLOG_MARK, APLOG_ERR, 0, s, APLOGNO(00832) "unusably short id provided "
+                "(%u bytes)", idlen);
+        return APR_EINVAL;
+    }
+    tryreplace = shmcb_subcache_remove(s, header, subcache, id, idlen);
+    if (shmcb_subcache_store(s, header, subcache, encoded,
+                             len_encoded, id, idlen, expiry)) {
+        ap_log_error(APLOG_MARK, APLOG_ERR, 0, s, APLOGNO(00833)
+                     "can't store an socache entry!");
+        return APR_ENOSPC;
+    }
+    if (tryreplace == 0) {
+        header->stat_replaced++;
+    }
+    else {
+        header->stat_stores++;
+    }
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00834)
+                 "leaving socache_shmcb_store successfully");
+    return APR_SUCCESS;
+}
+
+static apr_status_t socache_shmcb_retrieve(ap_socache_instance_t *ctx,
+                                           server_rec *s,
+                                           const unsigned char *id, unsigned int idlen,
+                                           unsigned char *dest, unsigned int *destlen,
+                                           apr_pool_t *p)
+{
+    SHMCBHeader *header = ctx->header;
+    SHMCBSubcache *subcache = SHMCB_MASK(header, id);
+    int rv;
+
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00835)
+                 "socache_shmcb_retrieve (0x%02x -> subcache %d)",
+                 SHMCB_MASK_DBG(header, id));
+
+    /* Get the entry corresponding to the id, if it exists. */
+    rv = shmcb_subcache_retrieve(s, header, subcache, id, idlen,
+                                 dest, destlen);
+    if (rv == 0)
+        header->stat_retrieves_hit++;
+    else
+        header->stat_retrieves_miss++;
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00836)
+                 "leaving socache_shmcb_retrieve successfully");
+
+    return rv == 0 ? APR_SUCCESS : APR_NOTFOUND;
+}
+
+static apr_status_t socache_shmcb_remove(ap_socache_instance_t *ctx,
+                                         server_rec *s, const unsigned char *id,
+                                         unsigned int idlen, apr_pool_t *p)
+{
+    SHMCBHeader *header = ctx->header;
+    SHMCBSubcache *subcache = SHMCB_MASK(header, id);
+    apr_status_t rv;
+
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00837)
+                 "socache_shmcb_remove (0x%02x -> subcache %d)",
+                 SHMCB_MASK_DBG(header, id));
+    if (idlen < 4) {
+        ap_log_error(APLOG_MARK, APLOG_ERR, 0, s, APLOGNO(00838) "unusably short id provided "
+                "(%u bytes)", idlen);
+        return APR_EINVAL;
+    }
+    if (shmcb_subcache_remove(s, header, subcache, id, idlen) == 0) {
+        header->stat_removes_hit++;
+        rv = APR_SUCCESS;
+    } else {
+        header->stat_removes_miss++;
+        rv = APR_NOTFOUND;
+    }
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00839)
+                 "leaving socache_shmcb_remove successfully");
+
+    return rv;
+}
+
+static void socache_shmcb_status(ap_socache_instance_t *ctx,
+                                 request_rec *r, int flags)
+{
+    server_rec *s = r->server;
+    SHMCBHeader *header = ctx->header;
+    unsigned int loop, total = 0, cache_total = 0, non_empty_subcaches = 0;
+    apr_time_t idx_expiry, min_expiry = 0, max_expiry = 0;
+    apr_time_t now = apr_time_now();
+    double expiry_total = 0;
+    int index_pct, cache_pct;
+
+    AP_DEBUG_ASSERT(header->subcache_num > 0);
+    ap_log_rerror(APLOG_MARK, APLOG_DEBUG, 0, r, APLOGNO(00840) "inside shmcb_status");
+    /* Perform the iteration inside the mutex to avoid corruption or invalid
+     * pointer arithmetic. The rest of our logic uses read-only header data so
+     * doesn't need the lock. */
+    /* Iterate over the subcaches */
+    for (loop = 0; loop < header->subcache_num; loop++) {
+        SHMCBSubcache *subcache = SHMCB_SUBCACHE(header, loop);
+        shmcb_subcache_expire(s, header, subcache, now);
+        total += subcache->idx_used;
+        cache_total += subcache->data_used;
+        if (subcache->idx_used) {
+            SHMCBIndex *idx = SHMCB_INDEX(subcache, subcache->idx_pos);
+            non_empty_subcaches++;
+            idx_expiry = idx->expires;
+            expiry_total += (double)idx_expiry;
+            max_expiry = ((idx_expiry > max_expiry) ? idx_expiry : max_expiry);
+            if (!min_expiry)
+                min_expiry = idx_expiry;
+            else
+                min_expiry = ((idx_expiry < min_expiry) ? idx_expiry : min_expiry);
+        }
+    }
+    index_pct = (100 * total) / (header->index_num *
+                                 header->subcache_num);
+    cache_pct = (100 * cache_total) / (header->subcache_data_size *
+                                       header->subcache_num);
+    /* Generate Output */
+    if (!(flags & AP_STATUS_SHORT)) {
+        ap_rprintf(r, "cache type: <b>SHMCB</b>, shared memory: <b>%" APR_SIZE_T_FMT "</b> "
+                   "bytes, current entries: <b>%d</b><br>",
+                   ctx->shm_size, total);
+        ap_rprintf(r, "subcaches: <b>%d</b>, indexes per subcache: <b>%d</b><br>",
+                   header->subcache_num, header->index_num);
+        if (non_empty_subcaches) {
+            apr_time_t average_expiry = (apr_time_t)(expiry_total / (double)non_empty_subcaches);
+            ap_rprintf(r, "time left on oldest entries' objects: ");
+            if (now < average_expiry)
+                ap_rprintf(r, "avg: <b>%d</b> seconds, (range: %d...%d)<br>",
+                           (int)apr_time_sec(average_expiry - now),
+                           (int)apr_time_sec(min_expiry - now),
+                           (int)apr_time_sec(max_expiry - now));
+            else
+                ap_rprintf(r, "expiry_threshold: <b>Calculation error!</b><br>");
+        }
+
+        ap_rprintf(r, "index usage: <b>%d%%</b>, cache usage: <b>%d%%</b><br>",
+                   index_pct, cache_pct);
+        ap_rprintf(r, "total entries stored since starting: <b>%lu</b><br>",
+                   header->stat_stores);
+        ap_rprintf(r, "total entries replaced since starting: <b>%lu</b><br>",
+                   header->stat_replaced);
+        ap_rprintf(r, "total entries expired since starting: <b>%lu</b><br>",
+                   header->stat_expiries);
+        ap_rprintf(r, "total (pre-expiry) entries scrolled out of the cache: "
+                   "<b>%lu</b><br>", header->stat_scrolled);
+        ap_rprintf(r, "total retrieves since starting: <b>%lu</b> hit, "
+                   "<b>%lu</b> miss<br>", header->stat_retrieves_hit,
+                   header->stat_retrieves_miss);
+        ap_rprintf(r, "total removes since starting: <b>%lu</b> hit, "
+                   "<b>%lu</b> miss<br>", header->stat_removes_hit,
+                   header->stat_removes_miss);
+    }
+    else {
+        ap_rputs("CacheType: SHMCB\n", r);
+        ap_rprintf(r, "CacheSharedMemory: %" APR_SIZE_T_FMT "\n",
+                   ctx->shm_size);
+        ap_rprintf(r, "CacheCurrentEntries: %d\n", total);
+        ap_rprintf(r, "CacheSubcaches: %d\n", header->subcache_num);
+        ap_rprintf(r, "CacheIndexesPerSubcaches: %d\n", header->index_num);
+        if (non_empty_subcaches) {
+            apr_time_t average_expiry = (apr_time_t)(expiry_total / (double)non_empty_subcaches);
+            if (now < average_expiry) {
+                ap_rprintf(r, "CacheTimeLeftOldestAvg: %d\n", (int)apr_time_sec(average_expiry - now));
+                ap_rprintf(r, "CacheTimeLeftOldestMin: %d\n", (int)apr_time_sec(min_expiry - now));
+                ap_rprintf(r, "CacheTimeLeftOldestMax: %d\n", (int)apr_time_sec(max_expiry - now));
+            }
+        }
+
+        ap_rprintf(r, "CacheIndexUsage: %d%%\n", index_pct);
+        ap_rprintf(r, "CacheUsage: %d%%\n", cache_pct);
+        ap_rprintf(r, "CacheStoreCount: %lu\n", header->stat_stores);
+        ap_rprintf(r, "CacheReplaceCount: %lu\n", header->stat_replaced);
+        ap_rprintf(r, "CacheExpireCount: %lu\n", header->stat_expiries);
+        ap_rprintf(r, "CacheDiscardCount: %lu\n", header->stat_scrolled);
+        ap_rprintf(r, "CacheRetrieveHitCount: %lu\n", header->stat_retrieves_hit);
+        ap_rprintf(r, "CacheRetrieveMissCount: %lu\n", header->stat_retrieves_miss);
+        ap_rprintf(r, "CacheRemoveHitCount: %lu\n", header->stat_removes_hit);
+        ap_rprintf(r, "CacheRemoveMissCount: %lu\n", header->stat_removes_miss);
+    }
+    ap_log_rerror(APLOG_MARK, APLOG_DEBUG, 0, r, APLOGNO(00841) "leaving shmcb_status");
+}
+
+static apr_status_t socache_shmcb_iterate(ap_socache_instance_t *instance,
+                                          server_rec *s, void *userctx,
+                                          ap_socache_iterator_t *iterator,
+                                          apr_pool_t *pool)
+{
+    SHMCBHeader *header = instance->header;
+    unsigned int loop;
+    apr_time_t now = apr_time_now();
+    apr_status_t rv = APR_SUCCESS;
+    apr_size_t buflen = 0;
+    unsigned char *buf = NULL;
+
+    /* Perform the iteration inside the mutex to avoid corruption or invalid
+     * pointer arithmetic. The rest of our logic uses read-only header data so
+     * doesn't need the lock. */
+    /* Iterate over the subcaches */
+    for (loop = 0; loop < header->subcache_num && rv == APR_SUCCESS; loop++) {
+        SHMCBSubcache *subcache = SHMCB_SUBCACHE(header, loop);
+        rv = shmcb_subcache_iterate(instance, s, userctx, header, subcache,
+                                    iterator, &buf, &buflen, pool, now);
+    }
+    return rv;
+}
+
+/*
+ * Subcache-level cache operations
+ */
+
+static void shmcb_subcache_expire(server_rec *s, SHMCBHeader *header,
+                                  SHMCBSubcache *subcache, apr_time_t now)
+{
+    unsigned int loop = 0, freed = 0, expired = 0;
+    unsigned int new_idx_pos = subcache->idx_pos;
+    SHMCBIndex *idx = NULL;
+
+    while (loop < subcache->idx_used) {
+        idx = SHMCB_INDEX(subcache, new_idx_pos);
+        if (idx->removed)
+            freed++;
+        else if (idx->expires <= now)
+            expired++;
+        else
+            /* not removed and not expired yet, we're done iterating */
+            break;
+        loop++;
+        new_idx_pos = SHMCB_CYCLIC_INCREMENT(new_idx_pos, 1, header->index_num);
+    }
+    if (!loop)
+        /* Nothing to do */
+        return;
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00842)
+                 "expiring %u and reclaiming %u removed socache entries",
+                 expired, freed);
+    if (loop == subcache->idx_used) {
+        /* We're expiring everything, piece of cake */
+        subcache->idx_used = 0;
+        subcache->data_used = 0;
+    } else {
+        /* There remain other indexes, so we can use idx to adjust 'data' */
+        unsigned int diff = SHMCB_CYCLIC_SPACE(subcache->data_pos,
+                                               idx->data_pos,
+                                               header->subcache_data_size);
+        /* Adjust the indexes */
+        subcache->idx_used -= loop;
+        subcache->idx_pos = new_idx_pos;
+        /* Adjust the data area */
+        subcache->data_used -= diff;
+        subcache->data_pos = idx->data_pos;
+    }
+    header->stat_expiries += expired;
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00843)
+                 "we now have %u socache entries", subcache->idx_used);
+}
+
+static int shmcb_subcache_store(server_rec *s, SHMCBHeader *header,
+                                SHMCBSubcache *subcache,
+                                unsigned char *data, unsigned int data_len,
+                                const unsigned char *id, unsigned int id_len,
+                                apr_time_t expiry)
+{
+    unsigned int data_offset, new_idx, id_offset;
+    SHMCBIndex *idx;
+    unsigned int total_len = id_len + data_len;
+
+    /* Sanity check the input */
+    if (total_len > header->subcache_data_size) {
+        ap_log_error(APLOG_MARK, APLOG_ERR, 0, s, APLOGNO(00844)
+                     "inserting socache entry larger (%d) than subcache data area (%d)",
+                     total_len, header->subcache_data_size);
+        return -1;
+    }
+
+    /* First reclaim space from removed and expired records. */
+    shmcb_subcache_expire(s, header, subcache, apr_time_now());
+
+    /* Loop until there is enough space to insert
+     * XXX: This should first compress out-of-order expiries and
+     * removed records, and then force-remove oldest-first
+     */
+    if (header->subcache_data_size - subcache->data_used < total_len
+        || subcache->idx_used == header->index_num) {
+        unsigned int loop = 0;
+
+        idx = SHMCB_INDEX(subcache, subcache->idx_pos);
+        ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00845)
+                     "about to force-expire, subcache: idx_used=%d, "
+                     "data_used=%d", subcache->idx_used, subcache->data_used);
+        do {
+            SHMCBIndex *idx2;
+
+            /* Adjust the indexes by one */
+            subcache->idx_pos = SHMCB_CYCLIC_INCREMENT(subcache->idx_pos, 1,
+                                                       header->index_num);
+            subcache->idx_used--;
+            if (!subcache->idx_used) {
+                /* There's nothing left */
+                subcache->data_used = 0;
+                break;
+            }
+            /* Adjust the data */
+            idx2 = SHMCB_INDEX(subcache, subcache->idx_pos);
+            subcache->data_used -= SHMCB_CYCLIC_SPACE(idx->data_pos, idx2->data_pos,
+                                                      header->subcache_data_size);
+            subcache->data_pos = idx2->data_pos;
+            /* Stats */
+            header->stat_scrolled++;
+            /* Loop admin */
+            idx = idx2;
+            loop++;
+        } while (header->subcache_data_size - subcache->data_used < total_len);
+
+        ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00846)
+                     "finished force-expire, subcache: idx_used=%d, "
+                     "data_used=%d", subcache->idx_used, subcache->data_used);
+    }
+
+    /* HERE WE ASSUME THAT THE NEW ENTRY SHOULD GO ON THE END! I'M NOT
+     * CHECKING WHETHER IT SHOULD BE GENUINELY "INSERTED" SOMEWHERE.
+     *
+     * We aught to fix that.  httpd (never mind third party modules)
+     * does not promise to perform any processing in date order
+     * (c.f. FAQ "My log entries are not in date order!")
+     */
+    /* Insert the id */
+    id_offset = SHMCB_CYCLIC_INCREMENT(subcache->data_pos, subcache->data_used,
+                                       header->subcache_data_size);
+    shmcb_cyclic_ntoc_memcpy(header->subcache_data_size,
+                             SHMCB_DATA(header, subcache), id_offset,
+                             id, id_len);
+    subcache->data_used += id_len;
+    /* Insert the data */
+    data_offset = SHMCB_CYCLIC_INCREMENT(subcache->data_pos, subcache->data_used,
+                                         header->subcache_data_size);
+    shmcb_cyclic_ntoc_memcpy(header->subcache_data_size,
+                             SHMCB_DATA(header, subcache), data_offset,
+                             data, data_len);
+    subcache->data_used += data_len;
+    /* Insert the index */
+    new_idx = SHMCB_CYCLIC_INCREMENT(subcache->idx_pos, subcache->idx_used,
+                                     header->index_num);
+    idx = SHMCB_INDEX(subcache, new_idx);
+    idx->expires = expiry;
+    idx->data_pos = id_offset;
+    idx->data_used = total_len;
+    idx->id_len = id_len;
+    idx->removed = 0;
+    subcache->idx_used++;
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00847)
+                 "insert happened at idx=%d, data=(%u:%u)", new_idx,
+                 id_offset, data_offset);
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00848)
+                 "finished insert, subcache: idx_pos/idx_used=%d/%d, "
+                 "data_pos/data_used=%d/%d",
+                 subcache->idx_pos, subcache->idx_used,
+                 subcache->data_pos, subcache->data_used);
+    return 0;
+}
+
+static int shmcb_subcache_retrieve(server_rec *s, SHMCBHeader *header,
+                                   SHMCBSubcache *subcache,
+                                   const unsigned char *id, unsigned int idlen,
+                                   unsigned char *dest, unsigned int *destlen)
+{
+    unsigned int pos;
+    unsigned int loop = 0;
+    apr_time_t now = apr_time_now();
+
+    pos = subcache->idx_pos;
+
+    while (loop < subcache->idx_used) {
+        SHMCBIndex *idx = SHMCB_INDEX(subcache, pos);
+
+        /* Only consider 'idx' if the id matches, and the "removed"
+         * flag isn't set, and the record is not expired.
+         * Check the data length too to avoid a buffer overflow
+         * in case of corruption, which should be impossible,
+         * but it's cheap to be safe. */
+        if (!idx->removed
+            && idx->id_len == idlen
+            && (idx->data_used - idx->id_len) <= *destlen
+            && shmcb_cyclic_memcmp(header->subcache_data_size,
+                                   SHMCB_DATA(header, subcache),
+                                   idx->data_pos, id, idx->id_len) == 0) {
+            ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00849)
+                         "match at idx=%d, data=%d", pos, idx->data_pos);
+            if (idx->expires > now) {
+                unsigned int data_offset;
+
+                /* Find the offset of the data segment, after the id */
+                data_offset = SHMCB_CYCLIC_INCREMENT(idx->data_pos,
+                                                     idx->id_len,
+                                                     header->subcache_data_size);
+
+                *destlen = idx->data_used - idx->id_len;
+
+                /* Copy out the data */
+                shmcb_cyclic_cton_memcpy(header->subcache_data_size,
+                                         dest, SHMCB_DATA(header, subcache),
+                                         data_offset, *destlen);
+
+                return 0;
+            }
+            else {
+                /* Already stale, quietly remove and treat as not-found */
+                idx->removed = 1;
+                header->stat_expiries++;
+                ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00850)
+                             "shmcb_subcache_retrieve discarding expired entry");
+                return -1;
+            }
+        }
+        /* Increment */
+        loop++;
+        pos = SHMCB_CYCLIC_INCREMENT(pos, 1, header->index_num);
+    }
+
+    ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00851)
+                 "shmcb_subcache_retrieve found no match");
+    return -1;
+}
+
+static int shmcb_subcache_remove(server_rec *s, SHMCBHeader *header,
+                                 SHMCBSubcache *subcache,
+                                 const unsigned char *id,
+                                 unsigned int idlen)
+{
+    unsigned int pos;
+    unsigned int loop = 0;
+
+    pos = subcache->idx_pos;
+    while (loop < subcache->idx_used) {
+        SHMCBIndex *idx = SHMCB_INDEX(subcache, pos);
+
+        /* Only consider 'idx' if the id matches, and the "removed"
+         * flag isn't set. */
+        if (!idx->removed && idx->id_len == idlen
+            && shmcb_cyclic_memcmp(header->subcache_data_size,
+                                   SHMCB_DATA(header, subcache),
+                                   idx->data_pos, id, idx->id_len) == 0) {
+            ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00852)
+                         "possible match at idx=%d, data=%d", pos, idx->data_pos);
+
+            /* Found the matching entry, remove it quietly. */
+            idx->removed = 1;
+            ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00853)
+                         "shmcb_subcache_remove removing matching entry");
+            return 0;
+        }
+        /* Increment */
+        loop++;
+        pos = SHMCB_CYCLIC_INCREMENT(pos, 1, header->index_num);
+    }
+
+    return -1; /* failure */
+}
+
+
+static apr_status_t shmcb_subcache_iterate(ap_socache_instance_t *instance,
+                                           server_rec *s,
+                                           void *userctx,
+                                           SHMCBHeader *header,
+                                           SHMCBSubcache *subcache,
+                                           ap_socache_iterator_t *iterator,
+                                           unsigned char **buf,
+                                           apr_size_t *buf_len,
+                                           apr_pool_t *pool,
+                                           apr_time_t now)
+{
+    unsigned int pos;
+    unsigned int loop = 0;
+    apr_status_t rv;
+
+    pos = subcache->idx_pos;
+    while (loop < subcache->idx_used) {
+        SHMCBIndex *idx = SHMCB_INDEX(subcache, pos);
+
+        /* Only consider 'idx' if the "removed" flag isn't set. */
+        if (!idx->removed) {
+
+            ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00854)
+                         "iterating idx=%d, data=%d", pos, idx->data_pos);
+            if (idx->expires > now) {
+                unsigned char *id = *buf;
+                unsigned char *dest;
+                unsigned int data_offset, dest_len;
+                apr_size_t buf_req;
+
+                /* Find the offset of the data segment, after the id */
+                data_offset = SHMCB_CYCLIC_INCREMENT(idx->data_pos,
+                                                     idx->id_len,
+                                                     header->subcache_data_size);
+
+                dest_len = idx->data_used - idx->id_len;
+
+                buf_req = APR_ALIGN_DEFAULT(idx->id_len + 1)
+                        + APR_ALIGN_DEFAULT(dest_len + 1);
+
+                if (buf_req > *buf_len) {
+                     /* Grow to ~150% of this buffer requirement on resize
+                      * always using APR_ALIGN_DEFAULT sized pages
+                      */
+                     *buf_len = buf_req + APR_ALIGN_DEFAULT(buf_req / 2);
+                     *buf = apr_palloc(pool, *buf_len);
+                     id = *buf;
+                }
+
+                dest = *buf + APR_ALIGN_DEFAULT(idx->id_len + 1);
+
+                /* Copy out the data, because it's potentially cyclic */
+                shmcb_cyclic_cton_memcpy(header->subcache_data_size, id,
+                                         SHMCB_DATA(header, subcache),
+                                         idx->data_pos, idx->id_len);
+                id[idx->id_len] = '\0';
+
+                shmcb_cyclic_cton_memcpy(header->subcache_data_size, dest,
+                                         SHMCB_DATA(header, subcache),
+                                         data_offset, dest_len);
+                dest[dest_len] = '\0';
+
+                rv = iterator(instance, s, userctx, id, idx->id_len,
+                              dest, dest_len, pool);
+                ap_log_error(APLOG_MARK, APLOG_DEBUG, rv, s, APLOGNO(00855)
+                             "shmcb entry iterated");
+                if (rv != APR_SUCCESS)
+                    return rv;
+            }
+            else {
+                /* Already stale, quietly remove and treat as not-found */
+                idx->removed = 1;
+                header->stat_expiries++;
+                ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(00856)
+                             "shmcb_subcache_iterate discarding expired entry");
+            }
+        }
+        /* Increment */
+        loop++;
+        pos = SHMCB_CYCLIC_INCREMENT(pos, 1, header->index_num);
+    }
+
+    return APR_SUCCESS;
+}
+
+static const ap_socache_provider_t socache_shmcb = {
+    "shmcb",
+    AP_SOCACHE_FLAG_NOTMPSAFE,
+    socache_shmcb_create,
+    socache_shmcb_init,
+    socache_shmcb_destroy,
+    socache_shmcb_store,
+    socache_shmcb_retrieve,
+    socache_shmcb_remove,
+    socache_shmcb_status,
+    socache_shmcb_iterate
+};
+
+static void register_hooks(apr_pool_t *p)
+{
+    ap_register_provider(p, AP_SOCACHE_PROVIDER_GROUP, "shmcb",
+                         AP_SOCACHE_PROVIDER_VERSION,
+                         &socache_shmcb);
+
+    /* Also register shmcb under the default provider name. */
+    ap_register_provider(p, AP_SOCACHE_PROVIDER_GROUP,
+                         AP_SOCACHE_DEFAULT_PROVIDER,
+                         AP_SOCACHE_PROVIDER_VERSION,
+                         &socache_shmcb);
+}
+
+AP_DECLARE_MODULE(socache_shmcb) = {
+    STANDARD20_MODULE_STUFF,
+    NULL, NULL, NULL, NULL, NULL,
+    register_hooks
+};