mlx4_core: Fix integer overflow issues around MTT table
Fix some issues around int variables used in data structures related
to memory registration.
Handle int overflow in mlx4_init_icm_table by using a u64 intermediate
variable and changing struct mlx4_icm_table num_obj field to be u32.
Change some more fields/variables to use u32 instead of int to prevent
a case where the variable becomes negative when bit 31 is set.
Also subtract log_mtts_per_seg from the exponent when computing
num_mtt, since its added later on in that very same code area.
This and the previous commit fixes some issues which actually prevent
commit db5a7a65c058 ("mlx4_core: Scale size of MTT table with system
RAM") from working. Now, when the number of MTTs is scaled with the
size of the RAM we can map up to 8TB.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Jack Morgenstein <jackm@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
diff --git a/drivers/net/ethernet/mellanox/mlx4/profile.c b/drivers/net/ethernet/mellanox/mlx4/profile.c
index 9ee4725..8e0c3cc 100644
--- a/drivers/net/ethernet/mellanox/mlx4/profile.c
+++ b/drivers/net/ethernet/mellanox/mlx4/profile.c
@@ -76,7 +76,7 @@
u64 size;
u64 start;
int type;
- int num;
+ u32 num;
int log_num;
};
@@ -105,7 +105,7 @@
si_meminfo(&si);
request->num_mtt =
roundup_pow_of_two(max_t(unsigned, request->num_mtt,
- min(1UL << 31,
+ min(1UL << (31 - log_mtts_per_seg),
si.totalram >> (log_mtts_per_seg - 1))));
profile[MLX4_RES_QP].size = dev_cap->qpc_entry_sz;