laforge has submitted this change. ( https://gerrit.osmocom.org/c/libosmocore/+/29591 )

Change subject: msgb: assert msgb->lXh to be not NULL ......................................................................

msgb: assert msgb->lXh to be not NULL

When any of l1h, l2h, l2h or l4h is set to NULL (which is the default for newly allocated message buffers). Then the msgb_lXhlen() functions will return the address value of msgb->tail. This can lead to unexpected results at a later point. We should have an OSMO_ASSERT to catch the problem early.

Change-Id: I1795c559f190713ebbabfbabf3453ab77da46a49 Related: OS#5645 --- M include/osmocom/core/msgb.h 1 file changed, 4 insertions(+), 0 deletions(-)

Approvals: laforge: Looks good to me, approved fixeria: Looks good to me, approved pespin: Looks good to me, but someone else must approve Jenkins Builder: Verified

diff --git a/include/osmocom/core/msgb.h b/include/osmocom/core/msgb.h index 117fcb0..fbf1742 100644 --- a/include/osmocom/core/msgb.h +++ b/include/osmocom/core/msgb.h @@ -144,6 +144,7 @@ */ static inline unsigned int msgb_l1len(const struct msgb *msgb) { + OSMO_ASSERT(msgb->l1h); return msgb->tail - (uint8_t *)msgb_l1(msgb); }

@@ -156,6 +157,7 @@ */ static inline unsigned int msgb_l2len(const struct msgb *msgb) { + OSMO_ASSERT(msgb->l2h); return msgb->tail - (uint8_t *)msgb_l2(msgb); }

@@ -168,6 +170,7 @@ */ static inline unsigned int msgb_l3len(const struct msgb *msgb) { + OSMO_ASSERT(msgb->l3h); return msgb->tail - (uint8_t *)msgb_l3(msgb); }

@@ -180,6 +183,7 @@ */ static inline unsigned int msgb_l4len(const struct msgb *msgb) { + OSMO_ASSERT(msgb->l4h); return msgb->tail - (uint8_t *)msgb_sms(msgb); }