K64f non-blocking powerup #7206
Conversation
Changed K64F/K66F power up to return without waiting for link up i.e. for the ethernet cable to be connected. This is needed for non-blocking use of driver e.g. for using the driver from event queue.
Currently, if all TX descriptors are in use and IP stack calls K64F/K66F ethernet driver link out, link out drops the packet. Added 10ms delay to link out to wait for a descriptor to become available before dropping the packet.
|
Thank you. Can we make changes to the PHY driver in the HAL layer and calls these functions instead of moving the PHY implementation to this driver. |
| status_t result = kStatus_Success; | ||
| uint32_t instance = ENET_GetInstance(base); | ||
|
|
||
| #if TARGET_K66F |
There was a problem hiding this comment.
Looks like we would like to split the PHY_Init function into 2. Can this be done inside fsl_phy.c and then call those functions rather than move the implementation from fsl_phy.c to inside this driver.
There was a problem hiding this comment.
Yes, that sounds good. Will you make a commit about the fsl_phy.c changes? Or should I make those?
There was a problem hiding this comment.
I have attached a patch, can you take a look and update PR.
There was a problem hiding this comment.
Thanks, I'll apply the patch.
| prev_state.speed = (phy_speed_t)STATE_UNKNOWN; | ||
| prev_state.duplex = (phy_duplex_t)STATE_UNKNOWN; | ||
|
|
||
| phy_task_handle = mbed::mbed_event_queue()->call_every(PHY_TASK_PERIOD_MS, mbed::callback(this, &Kinetis_EMAC::phy_task)); | ||
| mbed::mbed_event_queue()->call(mbed::callback(this, &Kinetis_EMAC::phy_task)); |
There was a problem hiding this comment.
The first call_every() callback is triggered after 200ms. Now the first call() callback is triggered right after the power up and then every 200ms.
PHY init and autonegotation is split into own functions.
|
Updated the patch to the pull request. |
|
@ARMmbed/mbed-os-ipcore Thoughts? |
|
/morph build |
Build : SUCCESSBuild number : 2383 Triggering tests/morph test |
Test : SUCCESSBuild number : 2172 |
Exporter Build : FAILUREBuild number : 2016 |
|
Relaunching since something went wrong with the nodes. |
The failure is due to build node been unresponsive. re-triggering /morph export-build |
Exporter Build : SUCCESSBuild number : 2024 |
|
@mikaleppanen Do we need similar changes for the LPC546XX driver as well? |
|
@mmahadevan108 Yes, it would be a good change to LPC546XX as well. |
Description
Changed K64F/K66F power up to return without waiting for link up i.e. for the ethernet cable to be connected. This is needed for non-blocking use of driver e.g. for using the driver from event queue.
Added delay to K64F/K66F ethernet TX when all descriptors are in use. Previously, if all TX descriptors were in use and IP stack called K64F/K66F ethernet driver link out, link out dropped the packet. Added 10ms delay to link out to wait for a descriptor to become available before dropping the packet.
Pull request type