From: Kees Cook <keescook@xxxxxxxxxxxx>
commit 439a8468242b313486e69b8cc3b45ddcfa898fbf upstream.
Partially revert commit 5f501d555653 ("binfmt_elf: reintroduce using
MAP_FIXED_NOREPLACE"), which applied the ET_DYN "total_mapping_size"
logic also to ET_EXEC.
At least ia64 has ET_EXEC PT_LOAD segments that are not virtual-address
contiguous (but _are_ file-offset contiguous). This would result in a
giant mapping attempting to cover the entire span, including the virtual
address range hole, and well beyond the size of the ELF file itself,
causing the kernel to refuse to load it. For example:
$ readelf -lW /usr/bin/gcc
...
Program Headers:
Type Offset VirtAddr PhysAddr FileSiz MemSiz ...
...
LOAD 0x000000 0x4000000000000000 0x4000000000000000 0x00b5a0 0x00b5a0 ...
LOAD 0x00b5a0 0x600000000000b5a0 0x600000000000b5a0 0x0005ac 0x000710 ...
...
^^^^^^^^ ^^^^^^^^^^^^^^^^^^ ^^^^^^^^ ^^^^^^^^
File offset range : 0x000000-0x00bb4c
0x00bb4c bytes
Virtual address range : 0x4000000000000000-0x600000000000bcb0
0x200000000000bcb0 bytes
Remove the total_mapping_size logic for ET_EXEC, which reduces the
ET_EXEC MAP_FIXED_NOREPLACE coverage to only the first PT_LOAD (better
than nothing), and retains it for ET_DYN.
Ironically, this is the reverse of the problem that originally caused
problems with MAP_FIXED_NOREPLACE: overlapping PT_LOAD segments. Future
work could restore full coverage if load_elf_binary() were to perform
mappings in a separate phase from the loading (where it could resolve
both overlaps and holes).
Cc: Eric Biederman <ebiederm@xxxxxxxxxxxx>
Cc: Alexander Viro <viro@xxxxxxxxxxxxxxxxxx>
Cc: linux-fsdevel@xxxxxxxxxxxxxxx
Cc: linux-mm@xxxxxxxxx
Reported-by: matoro <matoro_bugzilla_kernel@xxxxxxxxx>
Fixes: 5f501d555653 ("binfmt_elf: reintroduce using MAP_FIXED_NOREPLACE")
Link: https://lore.kernel.org/r/a3edd529-c42d-3b09-135c-7e98a15b150f@xxxxxxxxxxxxx
Tested-by: matoro <matoro_mailinglist_kernel@xxxxxxxxx>
Link: https://lore.kernel.org/lkml/ce8af9c13bcea9230c7689f3c1e0e2cd@xxxxxxxxx
Tested-By: John Paul Adrian Glaubitz <glaubitz@xxxxxxxxxxxxxxxxxxx>
Link: https://lore.kernel.org/lkml/49182d0d-708b-4029-da5f-bc18603440a6@xxxxxxxxxxxxxxxxxxx
Cc: stable@xxxxxxxxxxxxxxx
Signed-off-by: Kees Cook <keescook@xxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
fs/binfmt_elf.c | 25 ++++++++++++++++++-------
1 file changed, 18 insertions(+), 7 deletions(-)
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -1135,14 +1135,25 @@ out_free_interp:
* is then page aligned.
*/
load_bias = ELF_PAGESTART(load_bias - vaddr);
- }
- /*
- * Calculate the entire size of the ELF mapping (total_size).
- * (Note that load_addr_set is set to true later once the
- * initial mapping is performed.)
- */
- if (!load_addr_set) {
+ /*
+ * Calculate the entire size of the ELF mapping
+ * (total_size), used for the initial mapping,
+ * due to load_addr_set which is set to true later
+ * once the initial mapping is performed.
+ *
+ * Note that this is only sensible when the LOAD
+ * segments are contiguous (or overlapping). If
+ * used for LOADs that are far apart, this would
+ * cause the holes between LOADs to be mapped,
+ * running the risk of having the mapping fail,
+ * as it would be larger than the ELF file itself.
+ *
+ * As a result, only ET_DYN does this, since
+ * some ET_EXEC (e.g. ia64) may have large virtual
+ * memory holes between LOADs.
+ *
+ */
total_size = total_mapping_size(elf_phdata,
elf_ex->e_phnum);
if (!total_size) {
Patches currently in stable-queue which might be from keescook@xxxxxxxxxxxx are
queue-5.16/selftests-seccomp-fix-seccomp-failure-by-adding-miss.patch
queue-5.16/binfmt_elf-avoid-total_mapping_size-for-et_exec.patch
queue-5.16/ucounts-fix-systemd-limitnproc-with-private-users-regression.patch
