From 1dce27c5aa6770e9d195f2bb7db1db3d4dde5591 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Feb 2012 17:49:42 +0000 Subject: Wrap accesses to the fd_sets in struct fdtable Wrap accesses to the fd_sets in struct fdtable (for recording open files and close-on-exec flags) so that we can move away from using fd_sets since we abuse the fd_set structs by not allocating the full-sized structure under normal circumstances and by non-core code looking at the internals of the fd_sets. The first abuse means that use of FD_ZERO() on these fd_sets is not permitted, since that cannot be told about their abnormal lengths. This introduces six wrapper functions for setting, clearing and testing close-on-exec flags and fd-is-open flags: void __set_close_on_exec(int fd, struct fdtable *fdt); void __clear_close_on_exec(int fd, struct fdtable *fdt); bool close_on_exec(int fd, const struct fdtable *fdt); void __set_open_fd(int fd, struct fdtable *fdt); void __clear_open_fd(int fd, struct fdtable *fdt); bool fd_is_open(int fd, const struct fdtable *fdt); Note that I've prepended '__' to the names of the set/clear functions because they require the caller to hold a lock to use them. Note also that I haven't added wrappers for looking behind the scenes at the the array. Possibly that should exist too. Signed-off-by: David Howells Link: http://lkml.kernel.org/r/20120216174942.23314.1364.stgit@warthog.procyon.org.uk Signed-off-by: H. Peter Anvin Cc: Al Viro --- fs/file.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs/file.c') diff --git a/fs/file.c b/fs/file.c index 4c6992d8f3ba..114fea0a2cec 100644 --- a/fs/file.c +++ b/fs/file.c @@ -366,7 +366,7 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp) * is partway through open(). So make sure that this * fd is available to the new process. */ - FD_CLR(open_files - i, new_fdt->open_fds); + __clear_open_fd(open_files - i, new_fdt); } rcu_assign_pointer(*new_fds++, f); } @@ -460,11 +460,11 @@ repeat: if (start <= files->next_fd) files->next_fd = fd + 1; - FD_SET(fd, fdt->open_fds); + __set_open_fd(fd, fdt); if (flags & O_CLOEXEC) - FD_SET(fd, fdt->close_on_exec); + __set_close_on_exec(fd, fdt); else - FD_CLR(fd, fdt->close_on_exec); + __clear_close_on_exec(fd, fdt); error = fd; #if 1 /* Sanity check */ -- cgit v1.2.3 From 1fd36adcd98c14d2fd97f545293c488775cb2823 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Feb 2012 17:49:54 +0000 Subject: Replace the fd_sets in struct fdtable with an array of unsigned longs Replace the fd_sets in struct fdtable with an array of unsigned longs and then use the standard non-atomic bit operations rather than the FD_* macros. This: (1) Removes the abuses of struct fd_set: (a) Since we don't want to allocate a full fd_set the vast majority of the time, we actually, in effect, just allocate a just-big-enough array of unsigned longs and cast it to an fd_set type - so why bother with the fd_set at all? (b) Some places outside of the core fdtable handling code (such as SELinux) want to look inside the array of unsigned longs hidden inside the fd_set struct for more efficient iteration over the entire set. (2) Eliminates the use of FD_*() macros in the kernel completely. (3) Permits the __FD_*() macros to be deleted entirely where not exposed to userspace. Signed-off-by: David Howells Link: http://lkml.kernel.org/r/20120216174954.23314.48147.stgit@warthog.procyon.org.uk Signed-off-by: H. Peter Anvin Cc: Al Viro --- fs/file.c | 46 ++++++++++++++++++++++------------------------ 1 file changed, 22 insertions(+), 24 deletions(-) (limited to 'fs/file.c') diff --git a/fs/file.c b/fs/file.c index 114fea0a2cec..2d479dd8484e 100644 --- a/fs/file.c +++ b/fs/file.c @@ -40,7 +40,7 @@ int sysctl_nr_open_max = 1024 * 1024; /* raised later */ */ static DEFINE_PER_CPU(struct fdtable_defer, fdtable_defer_list); -static void *alloc_fdmem(unsigned int size) +static void *alloc_fdmem(size_t size) { /* * Very large allocations can stress page reclaim, so fall back to @@ -142,7 +142,7 @@ static void copy_fdtable(struct fdtable *nfdt, struct fdtable *ofdt) static struct fdtable * alloc_fdtable(unsigned int nr) { struct fdtable *fdt; - char *data; + void *data; /* * Figure out how many fds we actually want to support in this fdtable. @@ -172,14 +172,15 @@ static struct fdtable * alloc_fdtable(unsigned int nr) data = alloc_fdmem(nr * sizeof(struct file *)); if (!data) goto out_fdt; - fdt->fd = (struct file **)data; - data = alloc_fdmem(max_t(unsigned int, + fdt->fd = data; + + data = alloc_fdmem(max_t(size_t, 2 * nr / BITS_PER_BYTE, L1_CACHE_BYTES)); if (!data) goto out_arr; - fdt->open_fds = (fd_set *)data; - data += nr / BITS_PER_BYTE; - fdt->close_on_exec = (fd_set *)data; + fdt->open_fds = data; + data += nr / BITS_PER_LONG; + fdt->close_on_exec = data; fdt->next = NULL; return fdt; @@ -275,11 +276,11 @@ static int count_open_files(struct fdtable *fdt) int i; /* Find the last open fd */ - for (i = size/(8*sizeof(long)); i > 0; ) { - if (fdt->open_fds->fds_bits[--i]) + for (i = size / BITS_PER_LONG; i > 0; ) { + if (fdt->open_fds[--i]) break; } - i = (i+1) * 8 * sizeof(long); + i = (i + 1) * BITS_PER_LONG; return i; } @@ -306,8 +307,8 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp) newf->next_fd = 0; new_fdt = &newf->fdtab; new_fdt->max_fds = NR_OPEN_DEFAULT; - new_fdt->close_on_exec = (fd_set *)&newf->close_on_exec_init; - new_fdt->open_fds = (fd_set *)&newf->open_fds_init; + new_fdt->close_on_exec = newf->close_on_exec_init; + new_fdt->open_fds = newf->open_fds_init; new_fdt->fd = &newf->fd_array[0]; new_fdt->next = NULL; @@ -350,10 +351,8 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp) old_fds = old_fdt->fd; new_fds = new_fdt->fd; - memcpy(new_fdt->open_fds->fds_bits, - old_fdt->open_fds->fds_bits, open_files/8); - memcpy(new_fdt->close_on_exec->fds_bits, - old_fdt->close_on_exec->fds_bits, open_files/8); + memcpy(new_fdt->open_fds, old_fdt->open_fds, open_files / 8); + memcpy(new_fdt->close_on_exec, old_fdt->close_on_exec, open_files / 8); for (i = open_files; i != 0; i--) { struct file *f = *old_fds++; @@ -379,11 +378,11 @@ struct files_struct *dup_fd(struct files_struct *oldf, int *errorp) memset(new_fds, 0, size); if (new_fdt->max_fds > open_files) { - int left = (new_fdt->max_fds-open_files)/8; - int start = open_files / (8 * sizeof(unsigned long)); + int left = (new_fdt->max_fds - open_files) / 8; + int start = open_files / BITS_PER_LONG; - memset(&new_fdt->open_fds->fds_bits[start], 0, left); - memset(&new_fdt->close_on_exec->fds_bits[start], 0, left); + memset(&new_fdt->open_fds[start], 0, left); + memset(&new_fdt->close_on_exec[start], 0, left); } rcu_assign_pointer(newf->fdt, new_fdt); @@ -419,8 +418,8 @@ struct files_struct init_files = { .fdtab = { .max_fds = NR_OPEN_DEFAULT, .fd = &init_files.fd_array[0], - .close_on_exec = (fd_set *)&init_files.close_on_exec_init, - .open_fds = (fd_set *)&init_files.open_fds_init, + .close_on_exec = init_files.close_on_exec_init, + .open_fds = init_files.open_fds_init, }, .file_lock = __SPIN_LOCK_UNLOCKED(init_task.file_lock), }; @@ -443,8 +442,7 @@ repeat: fd = files->next_fd; if (fd < fdt->max_fds) - fd = find_next_zero_bit(fdt->open_fds->fds_bits, - fdt->max_fds, fd); + fd = find_next_zero_bit(fdt->open_fds, fdt->max_fds, fd); error = expand_files(files, fd); if (error < 0) -- cgit v1.2.3 From f044db4cb4bf16893812d35b5fbeaaf3e30c9215 Mon Sep 17 00:00:00 2001 From: Bobby Powers Date: Wed, 22 Feb 2012 00:29:47 -0500 Subject: fs: Fix close_on_exec pointer in alloc_fdtable alloc_fdtable allocates space for the open_fds and close_on_exec bitfields together, as 2 * nr / BITS_PER_BYTE. close_on_exec needs to point to open_fds + nr / BITS_PER_BYTE, not open_fds + nr / BITS_PER_LONG, as introducted in 1fd36adc: Replace the fd_sets in struct fdtable with an array of unsigned longs. Signed-off-by: Bobby Powers Link: http://lkml.kernel.org/r/1329888587-3087-1-git-send-email-bobbypowers@gmail.com Acked-by: David Howells Signed-off-by: H. Peter Anvin --- fs/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/file.c') diff --git a/fs/file.c b/fs/file.c index 2d479dd8484e..171f6b41b4cd 100644 --- a/fs/file.c +++ b/fs/file.c @@ -179,7 +179,7 @@ static struct fdtable * alloc_fdtable(unsigned int nr) if (!data) goto out_arr; fdt->open_fds = data; - data += nr / BITS_PER_LONG; + data += nr / BITS_PER_BYTE; fdt->close_on_exec = data; fdt->next = NULL; -- cgit v1.2.3