Hi Linus,
This patch does:
a) cleans up the show_mem() function on various architectures wrt
page_count() macro and the 'free' variable
b) corrects the comment above paging_init() about 0-8M page tables
c) changes the name of kflushd to bdflush. All kernel data structures
around this thread are called bdflush* so kflushd is a wrong name and
bdflush is right.
d) corrects the comments in fs/file.c about __get_free_page() usage; it
should say "vmalloc or kmalloc" for these two are used for array and fd
set allocation
e) corrects the comment in asm-i386/spinlock.h about semaphore.S
f) adds a free_mm() macro to kernel/fork.c to be symmetric to allocate_mm
g) optimizes the vmlist_lock usage by dropping it in vfree() before
freeing the 'tmp' element which was unlinked from vmlist but _after_ the
pages have been freed.
(another thing I wanted to do, but was in doubt, was to force the old lock
nfsservctl request to do do_exit(0) like we do in sys_bdlflush() for old
2.2.x userspace update to self-terminate. This way current rpc.lockd would
not spew out an error but quietly (and successfully, i.e. errcode=0) die,
just like update)
diff -urN -X dontdiff linux/arch/alpha/mm/init.c work/arch/alpha/mm/init.c
--- linux/arch/alpha/mm/init.c Mon Oct 16 23:38:41 2000
+++ work/arch/alpha/mm/init.c Fri Nov 24 12:40:52 2000
@@ -185,7 +185,7 @@
reserved++;
else if (PageSwapCache(mem_map+i))
cached++;
- else if (!atomic_read(&mem_map[i].count))
+ else if (!page_count(mem_map+i))
free++;
else
shared += atomic_read(&mem_map[i].count) - 1;
diff -urN -X dontdiff linux/arch/i386/mm/init.c work/arch/i386/mm/init.c
--- linux/arch/i386/mm/init.c Mon Oct 23 22:42:33 2000
+++ work/arch/i386/mm/init.c Fri Nov 24 12:37:56 2000
@@ -198,7 +198,7 @@
void show_mem(void)
{
- int i,free = 0, total = 0, reserved = 0;
+ int i, total = 0, reserved = 0;
int shared = 0, cached = 0;
int highmem = 0;
@@ -214,9 +214,7 @@
reserved++;
else if (PageSwapCache(mem_map+i))
cached++;
- else if (!page_count(mem_map+i))
- free++;
- else
+ else if (page_count(mem_map+i))
shared += page_count(mem_map+i) - 1;
}
printk("%d pages of RAM\n", total);
@@ -437,7 +435,7 @@
}
/*
- * paging_init() sets up the page tables - note that the first 4MB are
+ * paging_init() sets up the page tables - note that the first 8MB are
* already mapped by head.S.
*
* This routines also unmaps the page at virtual kernel address 0, so
diff -urN -X dontdiff linux/arch/ia64/mm/init.c work/arch/ia64/mm/init.c
--- linux/arch/ia64/mm/init.c Tue Oct 10 01:54:56 2000
+++ work/arch/ia64/mm/init.c Fri Nov 24 12:38:20 2000
@@ -245,7 +245,7 @@
void
show_mem (void)
{
- int i,free = 0,total = 0,reserved = 0;
+ int i, total = 0, reserved = 0;
int shared = 0, cached = 0;
printk("Mem-info:\n");
@@ -258,9 +258,7 @@
reserved++;
else if (PageSwapCache(mem_map+i))
cached++;
- else if (!page_count(mem_map + i))
- free++;
- else
+ else if (page_count(mem_map + i))
shared += page_count(mem_map + i) - 1;
}
printk("%d pages of RAM\n", total);
diff -urN -X dontdiff linux/arch/ppc/mm/init.c work/arch/ppc/mm/init.c
--- linux/arch/ppc/mm/init.c Thu Nov 9 03:01:34 2000
+++ work/arch/ppc/mm/init.c Fri Nov 24 12:41:53 2000
@@ -293,7 +293,7 @@
reserved++;
else if (PageSwapCache(mem_map+i))
cached++;
- else if (!atomic_read(&mem_map[i].count))
+ else if (!page_count(mem_map+i))
free++;
else
shared += atomic_read(&mem_map[i].count) - 1;
diff -urN -X dontdiff linux/arch/s390/mm/init.c work/arch/s390/mm/init.c
--- linux/arch/s390/mm/init.c Mon Oct 16 20:58:51 2000
+++ work/arch/s390/mm/init.c Fri Nov 24 12:42:08 2000
@@ -192,7 +192,7 @@
void show_mem(void)
{
- int i,free = 0,total = 0,reserved = 0;
+ int i, total = 0, reserved = 0;
int shared = 0, cached = 0;
printk("Mem-info:\n");
@@ -205,9 +205,7 @@
reserved++;
else if (PageSwapCache(mem_map+i))
cached++;
- else if (!atomic_read(&mem_map[i].count))
- free++;
- else
+ else if (page_count(mem_map+i))
shared += atomic_read(&mem_map[i].count) - 1;
}
printk("%d pages of RAM\n",total);
diff -urN -X dontdiff linux/arch/sh/mm/init.c work/arch/sh/mm/init.c
--- linux/arch/sh/mm/init.c Mon Oct 16 20:58:51 2000
+++ work/arch/sh/mm/init.c Fri Nov 24 12:42:13 2000
@@ -150,7 +150,7 @@
void show_mem(void)
{
- int i,free = 0,total = 0,reserved = 0;
+ int i, total = 0, reserved = 0;
int shared = 0, cached = 0;
printk("Mem-info:\n");
@@ -163,9 +163,7 @@
reserved++;
else if (PageSwapCache(mem_map+i))
cached++;
- else if (!page_count(mem_map+i))
- free++;
- else
+ else if (page_count(mem_map+i))
shared += page_count(mem_map+i) - 1;
}
printk("%d pages of RAM\n",total);
diff -urN -X dontdiff linux/fs/buffer.c work/fs/buffer.c
--- linux/fs/buffer.c Sun Nov 12 02:31:12 2000
+++ work/fs/buffer.c Fri Nov 24 12:35:53 2000
@@ -2481,9 +2481,9 @@
return;
}
- /* kflushd can wakeup us before we have a chance to
+ /* bdflush can wakeup us before we have a chance to
go to sleep so we must be smart in handling
- this wakeup event from kflushd to avoid deadlocking in SMP
+ this wakeup event from bdflush to avoid deadlocking in SMP
(we are not holding any lock anymore in these two paths). */
__set_current_state(TASK_UNINTERRUPTIBLE);
add_wait_queue(&bdflush_done, &wait);
@@ -2646,7 +2646,7 @@
tsk->session = 1;
tsk->pgrp = 1;
- strcpy(tsk->comm, "kflushd");
+ strcpy(tsk->comm, "bdflush");
bdflush_tsk = tsk;
/* avoid getting signals */
diff -urN -X dontdiff linux/fs/file.c work/fs/file.c
--- linux/fs/file.c Fri Jul 7 20:20:04 2000
+++ work/fs/file.c Fri Nov 24 12:32:26 2000
@@ -16,7 +16,7 @@
/*
- * Allocate an fd array, using __get_free_page() if possible.
+ * Allocate an fd array, using kmalloc or vmalloc.
* Note: the array isn't cleared at allocation time.
*/
struct file ** alloc_fd_array(int num)
@@ -125,7 +125,7 @@
}
/*
- * Allocate an fdset array, using __get_free_page() if possible.
+ * Allocate an fdset array, using kmalloc or vmalloc.
* Note: the array isn't cleared at allocation time.
*/
fd_set * alloc_fdset(int num)
diff -urN -X dontdiff linux/include/asm-i386/spinlock.h work/include/asm-i386/spinlock.h
--- linux/include/asm-i386/spinlock.h Sun Nov 19 04:56:59 2000
+++ work/include/asm-i386/spinlock.h Fri Nov 24 12:42:55 2000
@@ -141,7 +141,7 @@
* Changed to use the same technique as rw semaphores. See
* semaphore.h for details. -ben
*/
-/* the spinlock helpers are in arch/i386/kernel/semaphore.S */
+/* the spinlock helpers are in arch/i386/kernel/semaphore.c */
static inline void read_lock(rwlock_t *rw)
{
diff -urN -X dontdiff linux/kernel/fork.c work/kernel/fork.c
--- linux/kernel/fork.c Wed Nov 15 23:13:40 2000
+++ work/kernel/fork.c Fri Nov 24 12:35:23 2000
@@ -196,6 +196,7 @@
}
#define allocate_mm() (kmem_cache_alloc(mm_cachep, SLAB_KERNEL))
+#define free_mm(mm) (kmem_cache_free(mm_cachep, (mm)))
static struct mm_struct * mm_init(struct mm_struct * mm)
{
@@ -206,7 +207,7 @@
mm->pgd = pgd_alloc();
if (mm->pgd)
return mm;
- kmem_cache_free(mm_cachep, mm);
+ free_mm(mm);
return NULL;
}
@@ -236,7 +237,7 @@
if (mm == &init_mm) BUG();
pgd_free(mm->pgd);
destroy_context(mm);
- kmem_cache_free(mm_cachep, mm);
+ free_mm(mm);
}
/*
diff -urN -X dontdiff linux/mm/vmalloc.c work/mm/vmalloc.c
--- linux/mm/vmalloc.c Thu Nov 9 01:03:09 2000
+++ work/mm/vmalloc.c Fri Nov 24 12:30:34 2000
@@ -215,8 +215,8 @@
if (tmp->addr == addr) {
*p = tmp->next;
vmfree_area_pages(VMALLOC_VMADDR(tmp->addr), tmp->size);
- kfree(tmp);
write_unlock(&vmlist_lock);
+ kfree(tmp);
return;
}
}
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Nov 30 2000 - 21:00:11 EST