| commit | fd6c211eab32d489a304bb3ae63774887a3d6024 | |
| author | Harald Welte <laforge@gnumonks.org> | |
| Thu, 30 May 2019 13:21:45 +0000 (30 15:21 +0200) | ||
| committer | Vadim Yanitskiy <axilirator@gmail.com> | |
| Fri, 31 May 2019 15:35:31 +0000 (31 22:35 +0700) | ||
| tree | acb7e2b8cdc50b65d0a0f717c7879d25df95c863 | treesnapshot (tar.gz zip) |
| parent | 1a66c2991cbd90e5bdc51bfcfcfcd851b4abb2a3 | commitdiff |
trxcon: Suppress POWERON to TRX if we're already powered on.
The existing logic unconditionally wants to send a POWERON command
on TRXC whenever L1CTL_FBSB_REQ is received. That may cause some
problems when sending subsequent L1CTL_FBSB_REQ, e.g. due to signal loss.
Sending POWEROFF when transceiver is not powered on is normal though.
This can happen if trxcon is restarted while fake_trx was running.
The existing FSM state could unfortunately not been used, as it's a
mixture between the TRX connection state and the command/response state.
The current solution is just a work around. We definitely need to
introduce separate state machines for transceiver and its TRXC
interface.
Change-Id: I834e8897b95a2490811319697fc7cab6076db480
The existing logic unconditionally wants to send a POWERON command
on TRXC whenever L1CTL_FBSB_REQ is received. That may cause some
problems when sending subsequent L1CTL_FBSB_REQ, e.g. due to signal loss.
Sending POWEROFF when transceiver is not powered on is normal though.
This can happen if trxcon is restarted while fake_trx was running.
The existing FSM state could unfortunately not been used, as it's a
mixture between the TRX connection state and the command/response state.
The current solution is just a work around. We definitely need to
introduce separate state machines for transceiver and its TRXC
interface.
Change-Id: I834e8897b95a2490811319697fc7cab6076db480
| src/host/trxcon/l1ctl.c | diffblobblamehistory | |
| src/host/trxcon/trx_if.c | diffblobblamehistory | |
| src/host/trxcon/trx_if.h | diffblobblamehistory |