patch-2.3.40 linux/mm/highmem.c
Next file: linux/mm/slab.c
Previous file: linux/mm/filemap.c
Back to the patch index
Back to the overall index
- Lines: 22
- Date:
Thu Jan 20 09:51:42 2000
- Orig file:
v2.3.39/linux/mm/highmem.c
- Orig date:
Fri Jan 7 19:13:23 2000
diff -u --recursive --new-file v2.3.39/linux/mm/highmem.c linux/mm/highmem.c
@@ -128,7 +128,7 @@
flush_tlb_all();
}
-static unsigned long map_new_virtual(struct page *page)
+static inline unsigned long map_new_virtual(struct page *page)
{
unsigned long vaddr;
int count;
@@ -170,12 +170,6 @@
}
vaddr = PKMAP_ADDR(last_pkmap_nr);
pkmap_page_table[last_pkmap_nr] = mk_pte(page, kmap_prot);
- /*
- * Subtle! For some reason if we dont do this TLB flush then
- * we get data corruption and weird behavior in dbench runs.
- * But invlpg this should not be necessery ... Any ideas?
- */
- __flush_tlb_one(vaddr);
pkmap_count[last_pkmap_nr] = 1;
page->virtual = vaddr;
FUNET's LINUX-ADM group, linux-adm@nic.funet.fi
TCL-scripts by Sam Shen (who was at: slshen@lbl.gov)