Fix arm mmap when using mmap2 syscall. Fixes bug #1303

1 files changed, 41 insertions, 0 deletions

diff --git a/test/mmap/mmap2.c b/test/mmap/mmap2.c
new file mode 100644
index 000000000..b9a8f9ac4
--- /dev/null
+++ b/test/mmap/mmap2.c
@@ -0,0 +1,41 @@
+/* When trying to map /dev/mem with offset 0xFFFFF000 on the ARM platform, mmap
+ * returns -EOVERFLOW.
+ *
+ * Since off_t is defined as a long int and the sign bit is set in the address,
+ * the shift operation shifts in ones instead of zeroes
+ * from the left. This results the offset sent to the kernel function becomes
+ * 0xFFFFFFFF instead of 0x000FFFFF with MMAP2_PAGE_SHIFT set to 12.
+ */
+
+#include <unistd.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <sys/mman.h>
+
+#define FATAL do { fprintf(stderr, "Error at line %d, file %s (%d) [%s]\n", \
+  __LINE__, __FILE__, errno, strerror(errno)); exit(1); } while(0)
+
+#define MAP_SIZE 4096UL
+#define MAP_MASK (MAP_SIZE - 1)
+
+int main(int argc, char **argv) {
+    void* map_base = 0;
+    int fd;
+    off_t target = 0xfffff000;
+    if((fd = open("/dev/mem", O_RDWR | O_SYNC)) == -1) FATAL;
+    printf("/dev/mem opened.\n");
+    fflush(stdout);
+
+   /* Map one page */
+    map_base = mmap(0, MAP_SIZE, PROT_READ | PROT_WRITE, MAP_SHARED,
+                        fd, target & ~MAP_MASK);
+    if(map_base == (void *) -1) FATAL;
+    printf("Memory mapped at address %p.\n", map_base);
+    fflush(stdout);
+    if(munmap(map_base, MAP_SIZE) == -1) FATAL;
+    close(fd);
+    return 0;
+}