RoshanESP commented on code in PR #1557:
URL: https://github.com/apache/mynewt-nimble/pull/1557#discussion_r1283998557
##########
nimble/host/src/ble_hs_mbuf.c:
##########
@@ -31,9 +31,9 @@ ble_hs_mbuf_gen_pkt(uint16_t leading_space)
int rc;
#if MYNEWT_VAL(BLE_CONTROLLER)
- om = os_msys_get_pkthdr(0, sizeof(struct ble_mbuf_hdr));
+ om = os_msys_get_pkthdr(leading_space, sizeof(struct ble_mbuf_hdr));
#else
- om = os_msys_get_pkthdr(0, 0);
+ om = os_msys_get_pkthdr(leading_space, 0);
Review Comment:
Hi @sjanc ,
Actually, the dsize param in os_msys_get_pkthdr() is used only to find the
pool and not to allocate the packet header.
https://github.com/apache/mynewt-nimble/blob/e718a2fb22a56141ddb6a63cab875489a6a08f25/porting/nimble/src/os_mbuf.c#L192-L209
In os_msys_get_pktdhr() dsize is used only to find the pool, by passing the
appropriate dsize, the possibility of smaller mbuf getting allocated can be
eliminated.
`while dsize param in os_msys_get_pkthdr() is useful only of one know exact
data size that will be used`.
I think passing dsize doesn't allocate the mbuf of exact size, rather it
allocates the mbuf that has data_size >= dsize.
As the ble_hs_mbuf_gen_pkt is static function and called only with
leading_space required for protocol headers(which is small enough), it is not
causing any problem.
If someone keeps the msys block_size small enough then this may cause error
as the check
`if (om->om_omp->omp_databuf_len < leading_space) ` will fail. So, passing
the leading_space to os_msys_get_pkthdr is safer.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
