-- Oryginal message --
On 1/3/2018 1:12 PM, maciej.cze...@caviumnetworks.com wrote:
From: Maciej Czekaj <maciej.cze...@caviumnetworks.com>
This patch removes all references to old-style offload API
replacing them with new offload flags.
Signed-off-by: Maciej Czekaj <maciej.cze...@caviumnetworks.com>
<...>
dev_info->default_txconf = (struct rte_eth_txconf) {
.tx_free_thresh = NICVF_DEFAULT_TX_FREE_THRESH,
- .txq_flags =
- ETH_TXQ_FLAGS_NOMULTSEGS |
- ETH_TXQ_FLAGS_NOREFCOUNT |
- ETH_TXQ_FLAGS_NOMULTMEMP |
- ETH_TXQ_FLAGS_NOVLANOFFL |
- ETH_TXQ_FLAGS_NOXSUMSCTP,
+ .txq_flags = ETH_TXQ_FLAGS_IGNORE,
I am not sure about this, Shahafs may comment better, shouldn't application
decide to set "c" or not, instead of having this in default
configuration?
I think of it as a safeguard against a legacy application that is using
old-style txq_flags.
There is an assymetry in API, since
- rte_eth_tx_queue setup() and rte_eth_rx_queue_setup() convert flags
- rte_eth_dev_info_get() - does not convert them.
The scenario leading to issues is as follows:
1.Application reads default txq_flags from the rte_eth_dev_info_get().
- dev_info.txconf.offloads contains flags but it is ignored by legacy code
- dev_info.txq_flags may be:
a) txq_flags == 0
b) txq_flags == ETH_TXQ_FLAGS_IGNORE
c) txq_flags == new-style flags converted to old-style flags
2. Application uses default txq_flags field to rte_eth_tx_queue_setup().
Now, depending on txq_flags:
a) txq_flags == 0, ethdev layer __clears__ all offloads, see *
b) txq_flags == ETH_TXQ_FLAGS_IGNORE, ethdev layer converts offloads
to txq_flags, but leaves orignal offloads, so PMD can still use them
c) txq_flags == old-style flags, ethdev layer converts txq_flags to
offloads, destroying default offloads
* relevant code snippet from rte_eth_tx_queue_setup() with comments:
if (tx_conf->txq_flags & ETH_TXQ_FLAGS_IGNORE) {
// ==> converts offloads to txq_flags but LEAVES offloads, too
rte_eth_convert_txq_offloads(tx_conf->offloads,
&local_conf.txq_flags);
/* Keep the ignore flag. */
local_conf.txq_flags |= ETH_TXQ_FLAGS_IGNORE;
} else {
// ==> converts txq_flags to offloads but DESTROYS original
offloads
rte_eth_convert_txq_flags(tx_conf->txq_flags,
&local_conf.offloads);
}
So, out of 3 options:
a) does not work for legacy code
b) will work for legacy code
c) will work but defeats the purpose of removing old-style flags, since
dev_info() callback has to setup both old-style and new-style default flags
I chose b) as the simplest way to work-around the issue. I could post a
patch to ethdev API if you think it is important.
<...>
+ if ((conf_tx_offloads & tx_offload_capa) != conf_tx_offloads) {
+ PMD_INIT_LOG(ERR, "Some Tx offloads are not supported "
+ "requested 0x%lx supported 0x%lx\n",
+ conf_tx_offloads, tx_offload_capa);
This is broken for 32bits, using PRIx64 instead of "lx" makes your code more
portable.
Will fix in v2.
