vda, xmalloc'ing can use *alot* of RAM. Don't forget that there are

modules out there that are half a megabyte! mmap()ing is way nicer
   for small mem boxes.
This commit is contained in:
Bernhard Reutner-Fischer 2008-05-27 09:06:05 +00:00
parent 9474830006
commit 9cf0f62b15

View File

@ -955,6 +955,7 @@ arch_apply_relocation(struct obj_file *f,
case R_386_PLT32:
case R_386_PC32:
case R_386_GOTOFF:
*loc += v - dot;
break;
@ -973,9 +974,6 @@ arch_apply_relocation(struct obj_file *f,
case R_386_GOT32:
goto bb_use_got;
case R_386_GOTOFF:
*loc += v - got;
break;
#elif defined(__microblaze__)
@ -4213,7 +4211,6 @@ int insmod_main(int argc ATTRIBUTE_UNUSED, char **argv)
static int insmod_ng_main(int argc ATTRIBUTE_UNUSED, char **argv)
#endif
{
long ret;
size_t len;
int optlen;
void *map;
@ -4234,6 +4231,11 @@ static int insmod_ng_main(int argc ATTRIBUTE_UNUSED, char **argv)
#if 0
/* Any special reason why mmap? It isn't performace critical... */
/* yes, xmalloc'ing can use *alot* of RAM. Don't forget that there are
* modules out there that are half a megabyte! mmap()ing is way nicer
* for small mem boxes, i guess.
*/
int fd;
struct stat st;
unsigned long len;
@ -4255,11 +4257,9 @@ static int insmod_ng_main(int argc ATTRIBUTE_UNUSED, char **argv)
map = xmalloc_open_read_close(filename, &len);
#endif
ret = syscall(__NR_init_module, map, len, options);
if (ret != 0) {
if (syscall(__NR_init_module, map, len, options) != 0)
bb_error_msg_and_die("cannot insert '%s': %s",
filename, moderror(errno));
}
return 0;
}