1 files changed, 417 insertions, 23 deletions

diff --git a/db2/os/os_map.c b/db2/os/os_map.c
index b1553188dc..5f0fd790e6 100644
--- a/db2/os/os_map.c
+++ b/db2/os/os_map.c
@@ -1,47 +1,395 @@
 /*-
  * See the file LICENSE for redistribution information.
  *
- * Copyright (c) 1996, 1997
+ * Copyright (c) 1996, 1997, 1998
  *	Sleepycat Software.  All rights reserved.
  */
 
 #include "config.h"
 
 #ifndef lint
-static const char sccsid[] = "@(#)os_map.c	10.7 (Sleepycat) 10/25/97";
+static const char sccsid[] = "@(#)os_map.c	10.19 (Sleepycat) 5/3/98";
 #endif /* not lint */
 
 #ifndef NO_SYSTEM_INCLUDES
 #include <sys/types.h>
+#ifdef HAVE_MMAP
 #include <sys/mman.h>
+#endif
+
+#ifdef HAVE_SHMGET
+#include <sys/ipc.h>
+#include <sys/shm.h>
+#endif
 
 #include <errno.h>
+#include <string.h>
 #endif
 
 #include "db_int.h"
+#include "common_ext.h"
+
+#ifdef HAVE_MMAP
+static int __os_map __P((char *, int, size_t, int, int, int, void **));
+#endif
+#ifdef HAVE_SHMGET
+static int __os_shmget __P((char *, REGINFO *));
+#endif
 
 /*
- * __os_map --
- *	Map in some shared memory backed by a file descriptor.
+ * __db_mapanon_ok --
+ *	Return if this OS can support anonymous memory regions.
+ *
+ * PUBLIC: int __db_mapanon_ok __P((int));
+ */
+int
+__db_mapanon_ok(need_names)
+	int need_names;
+{
+	int ret;
+
+	ret = EINVAL;
+
+	/*
+	 * If we don't have spinlocks, we have to have a file descriptor
+	 * for fcntl(2) locking, which implies using mmap(2) to map in a
+	 * regular file.  Theoretically, we could probably find ways to
+	 * get a file descriptor to lock other types of shared regions,
+	 * but I don't see any reason to do so.
+	 *
+	 * If need_names is set, the application wants to share anonymous
+	 * memory among multiple processes, so we have to have a way to
+	 * name it.  This requires shmget(2), on UNIX systems.
+	 */
+#ifdef HAVE_SPINLOCKS
+#ifdef HAVE_SHMGET
+	ret = 0;
+#endif
+#ifdef HAVE_MMAP
+#ifdef MAP_ANON
+	if (!need_names)
+		ret = 0;
+#endif
+#ifdef MAP_ANONYMOUS
+	if (!need_names)
+		ret = 0;
+#endif
+#else
+	COMPQUIET(need_names, 0);
+#endif /* HAVE_MMAP */
+#endif /* HAVE_SPINLOCKS */
+
+	return (ret);
+}
+
+/*
+ * __db_mapinit --
+ *	Return if shared regions need to be initialized.
+ *
+ * PUBLIC: int __db_mapinit __P((void));
+ */
+int
+__db_mapinit()
+{
+	/*
+	 * Historically, some systems required that all of the bytes of the
+	 * region be written before it could be mmapped and accessed randomly.
+	 * We have the option of setting REGION_INIT_NEEDED at configuration
+	 * time if we're running on one of those systems.
+	 */
+#ifdef REGION_INIT_NEEDED
+	return (1);
+#else
+	return (0);
+#endif
+}
+
+/*
+ * __db_mapregion --
+ *	Attach to a shared memory region.
+ *
+ * PUBLIC: int __db_mapregion __P((char *, REGINFO *));
+ */
+int
+__db_mapregion(path, infop)
+	char *path;
+	REGINFO *infop;
+{
+	int called, ret;
+
+	called = 0;
+	ret = EINVAL;
+
+	/* If the user replaces the map call, call through their interface. */
+	if (__db_jump.j_map != NULL) {
+		F_SET(infop, REGION_HOLDINGSYS);
+		return (__db_jump.j_map(path, infop->fd, infop->size,
+		    1, F_ISSET(infop, REGION_ANONYMOUS), 0, &infop->addr));
+	}
+
+	if (F_ISSET(infop, REGION_ANONYMOUS)) {
+		/*
+		 * !!!
+		 * If we're creating anonymous regions:
+		 *
+		 * If it's private, we use mmap(2).  The problem with using
+		 * shmget(2) is that we may be creating a region of which the
+		 * application isn't aware, and if the application crashes
+		 * we'll have no way to remove the system resources for the
+		 * region.
+		 *
+		 * If it's not private, we use the shmget(2) interface if it's
+		 * available, because it allows us to name anonymous memory.
+		 * If shmget(2) isn't available, use the mmap(2) calls.
+		 *
+		 * In the case of anonymous memory, using mmap(2) means the
+		 * memory isn't named and only the single process and its
+		 * threads can access the region.
+		 */
+#ifdef	HAVE_MMAP
+#ifdef	MAP_ANON
+#define	HAVE_MMAP_ANONYMOUS	1
+#else
+#ifdef	MAP_ANONYMOUS
+#define	HAVE_MMAP_ANONYMOUS	1
+#endif
+#endif
+#endif
+#ifdef HAVE_MMAP_ANONYMOUS
+		if (!called && F_ISSET(infop, REGION_PRIVATE)) {
+			called = 1;
+			ret = __os_map(path,
+			    infop->fd, infop->size, 1, 1, 0, &infop->addr);
+		}
+#endif
+#ifdef HAVE_SHMGET
+		if (!called) {
+			called = 1;
+			ret = __os_shmget(path, infop);
+		}
+#endif
+#ifdef HAVE_MMAP
+		/*
+		 * If we're trying to join an unnamed anonymous region, fail --
+		 * that's not possible.
+		 */
+		if (!called) {
+			called = 1;
+
+			if (!F_ISSET(infop, REGION_CREATED)) {
+				__db_err(infop->dbenv,
+			    "cannot join region in unnamed anonymous memory");
+				return (EINVAL);
+			}
+
+			ret = __os_map(path,
+			    infop->fd, infop->size, 1, 1, 0, &infop->addr);
+		}
+#endif
+	} else {
+		/*
+		 * !!!
+		 * If we're creating normal regions, we use the mmap(2)
+		 * interface if it's available because it's POSIX 1003.1
+		 * standard and we trust it more than we do shmget(2).
+		 */
+#ifdef HAVE_MMAP
+		if (!called) {
+			called = 1;
+
+			/* Mmap(2) regions that aren't anonymous can grow. */
+			F_SET(infop, REGION_CANGROW);
+
+			ret = __os_map(path,
+			    infop->fd, infop->size, 1, 0, 0, &infop->addr);
+		}
+#endif
+#ifdef HAVE_SHMGET
+		if (!called) {
+			called = 1;
+			ret = __os_shmget(path, infop);
+		}
+#endif
+	}
+	return (ret);
+}
+
+/*
+ * __db_unmapregion --
+ *	Detach from the shared memory region.
+ *
+ * PUBLIC: int __db_unmapregion __P((REGINFO *));
+ */
+int
+__db_unmapregion(infop)
+	REGINFO *infop;
+{
+	int called, ret;
+
+	called = 0;
+	ret = EINVAL;
+
+	if (__db_jump.j_unmap != NULL)
+		return (__db_jump.j_unmap(infop->addr, infop->size));
+
+#ifdef HAVE_SHMGET
+	if (infop->segid != INVALID_SEGID) {
+		called = 1;
+		ret = shmdt(infop->addr) ? errno : 0;
+	}
+#endif
+#ifdef HAVE_MMAP
+	if (!called) {
+		called = 1;
+		ret = munmap(infop->addr, infop->size) ? errno : 0;
+	}
+#endif
+	return (ret);
+}
+
+/*
+ * __db_unlinkregion --
+ *	Remove the shared memory region.
+ *
+ * PUBLIC: int __db_unlinkregion __P((char *, REGINFO *));
+ */
+int
+__db_unlinkregion(name, infop)
+	char *name;
+	REGINFO *infop;
+{
+	int called, ret;
+
+	called = 0;
+	ret = EINVAL;
+
+	if (__db_jump.j_runlink != NULL)
+		return (__db_jump.j_runlink(name));
+
+#ifdef HAVE_SHMGET
+	if (infop->segid != INVALID_SEGID) {
+		called = 1;
+		ret = shmctl(infop->segid, IPC_RMID, NULL) ? errno : 0;
+	}
+#else
+	COMPQUIET(infop, NULL);
+#endif
+#ifdef HAVE_MMAP
+	if (!called) {
+		called = 1;
+		ret = 0;
+	}
+#endif
+	return (ret);
+}
+
+/*
+ * __db_mapfile --
+ *	Map in a shared memory file.
+ *
+ * PUBLIC: int __db_mapfile __P((char *, int, size_t, int, void **));
+ */
+int
+__db_mapfile(path, fd, len, is_rdonly, addr)
+	char *path;
+	int fd, is_rdonly;
+	size_t len;
+	void **addr;
+{
+	if (__db_jump.j_map != NULL)
+		return (__db_jump.j_map(path, fd, len, 0, 0, is_rdonly, addr));
+
+#ifdef HAVE_MMAP
+	return (__os_map(path, fd, len, 0, 0, is_rdonly, addr));
+#else
+	return (EINVAL);
+#endif
+}
+
+/*
+ * __db_unmapfile --
+ *	Unmap the shared memory file.
  *
- * PUBLIC: int __os_map __P((int, size_t, int, int, void **));
+ * PUBLIC: int __db_unmapfile __P((void *, size_t));
  */
 int
-__os_map(fd, len, is_private, is_rdonly, addr)
-	int fd, is_private, is_rdonly;
+__db_unmapfile(addr, len)
+	void *addr;
+	size_t len;
+{
+	if (__db_jump.j_unmap != NULL)
+		return (__db_jump.j_unmap(addr, len));
+
+#ifdef HAVE_MMAP
+	return (munmap(addr, len) ? errno : 0);
+#else
+	return (EINVAL);
+#endif
+}
+
+#ifdef HAVE_MMAP
+/*
+ * __os_map --
+ *	Call the mmap(2) function.
+ */
+static int
+__os_map(path, fd, len, is_region, is_anonymous, is_rdonly, addr)
+	char *path;
+	int fd, is_region, is_anonymous, is_rdonly;
 	size_t len;
 	void **addr;
 {
 	void *p;
 	int flags, prot;
 
-	flags = is_private ? MAP_PRIVATE : MAP_SHARED;
+	COMPQUIET(path, NULL);
+
+	/*
+	 * If it's read-only, it's private, and if it's not, it's shared.
+	 * Don't bother with an additional parameter.
+	 */
+	flags = is_rdonly ? MAP_PRIVATE : MAP_SHARED;
+
+	if (is_region && is_anonymous) {
+		/*
+		 * BSD derived systems use MAP_ANON; Digital Unix and HP/UX
+		 * use MAP_ANONYMOUS.
+		 */
+#ifdef MAP_ANON
+		flags |= MAP_ANON;
+#endif
+#ifdef MAP_ANONYMOUS
+		flags |= MAP_ANONYMOUS;
+#endif
+		fd = -1;
+	}
+#ifdef MAP_FILE
+	if (!is_region || !is_anonymous) {
+		/*
+		 * Historically, MAP_FILE was required for mapping regular
+		 * files, even though it was the default.  Some systems have
+		 * it, some don't, some that have it set it to 0.
+		 */
+		flags |= MAP_FILE;
+	}
+#endif
+
+	/*
+	 * I know of no systems that implement the flag to tell the system
+	 * that the region contains semaphores, but it's not an unreasonable
+	 * thing to do, and has been part of the design since forever.  I
+	 * don't think anyone will object, but don't set it for read-only
+	 * files, it doesn't make sense.
+	 */
 #ifdef MAP_HASSEMAPHORE
-	flags |= MAP_HASSEMAPHORE;
+	if (!is_rdonly)
+		flags |= MAP_HASSEMAPHORE;
 #endif
+
 	prot = PROT_READ | (is_rdonly ? 0 : PROT_WRITE);
 
-#ifndef MAP_FAILED			/* XXX: Mmap(2) failure return. */
+	/* MAP_FAILED was not defined in early mmap implementations. */
+#ifndef MAP_FAILED
 #define	MAP_FAILED	-1
 #endif
 	if ((p =
@@ -51,21 +399,67 @@ __os_map(fd, len, is_private, is_rdonly, addr)
 	*addr = p;
 	return (0);
 }
+#endif
 
+#ifdef HAVE_SHMGET
 /*
- * __os_unmap --
- *	Release the specified shared memory.
- *
- * PUBLIC: int __os_unmap __P((void *, size_t));
+ * __os_shmget --
+ *	Call the shmget(2) family of functions.
  */
-int
-__os_unmap(addr, len)
-	void *addr;
-	size_t len;
+static int
+__os_shmget(path, infop)
+	REGINFO *infop;
+	char *path;
 {
-	/*
-	 * !!!
-	 * The argument len is always the same length as was mapped.
-	 */
-	return (munmap(addr, len) ? errno : 0);
+	key_t key;
+	int shmflg;
+
+	if (F_ISSET(infop, REGION_CREATED)) {
+		/*
+		 * The return key from ftok(3) is not guaranteed to be unique.
+		 * The nice thing about the shmget(2) interface is that it
+		 * allows you to name anonymous pieces of memory.  The evil
+		 * thing about it is that the name space is separate from the
+		 * filesystem.
+		 */
+#ifdef __hp3000s900
+		{char mpe_path[MAXPATHLEN];
+		/*
+		 * MPE ftok() is broken as of 5.5pp4.  If the file path does
+		 * not start with '/' or '.', then ftok() tries to interpret
+		 * the file path in MPE syntax instead of POSIX HFS syntax.
+		 * The workaround is to prepend "./" to these paths.  See HP
+		 * SR 5003416081 for details.
+		 */
+		if (*path != '/' && *path != '.') {
+			if (strlen(path) + strlen("./") + 1 > sizeof(mpe_path))
+				return (ENAMETOOLONG);
+			mpe_path[0] = '.';
+			mpe_path[1] = '/';
+			(void)strcpy(mpe_path + 2, path);
+			path = mpe_path;
+		}
+		}
+#endif
+		if ((key = ftok(path, 1)) == (key_t)-1)
+			return (errno);
+
+		shmflg = IPC_CREAT | 0600;
+		if ((infop->segid = shmget(key, infop->size, shmflg)) == -1)
+			return (errno);
+	}
+
+	if ((infop->addr = shmat(infop->segid, NULL, 0)) == (void *)-1) {
+		/*
+		 * If we're trying to join the region and failing, assume
+		 * that there was a reboot and the region no longer exists.
+		 */
+		if (!F_ISSET(infop, REGION_CREATED))
+			errno = EAGAIN;
+		return (errno);
+	}
+
+	F_SET(infop, REGION_HOLDINGSYS);
+	return (0);
 }
+#endif