can: m_can: implement errata "Needless activation of MRAF irq"
authorEugen Hristev <eugen.hristev@microchip.com>
Mon, 4 Mar 2019 14:44:13 +0000 (14:44 +0000)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sun, 21 Jul 2019 07:04:22 +0000 (09:04 +0200)
commite086121ec28d44c7be58fbc112d67dc1b56313ee
tree5d2cf5f38c4467aa04c4e6069f118faf20fc576f
parent905610fde81cb58338c96f8c02389be531f0c337
can: m_can: implement errata "Needless activation of MRAF irq"

[ Upstream commit 3e82f2f34c930a2a0a9e69fdc2de2f2f1388b442 ]

During frame reception while the MCAN is in Error Passive state and the
Receive Error Counter has thevalue MCAN_ECR.REC = 127, it may happen
that MCAN_IR.MRAF is set although there was no Message RAM access
failure. If MCAN_IR.MRAF is enabled, an interrupt to the Host CPU is
generated.

Work around:
The Message RAM Access Failure interrupt routine needs to check whether

    MCAN_ECR.RP = '1' and MCAN_ECR.REC = '127'.

In this case, reset MCAN_IR.MRAF. No further action is required.
This affects versions older than 3.2.0

Errata explained on Sama5d2 SoC which includes this hardware block:
http://ww1.microchip.com/downloads/en/DeviceDoc/SAMA5D2-Family-Silicon-Errata-and-Data-Sheet-Clarification-DS80000803B.pdf
chapter 6.2

Reproducibility: If 2 devices with m_can are connected back to back,
configuring different bitrate on them will lead to interrupt storm on
the receiving side, with error "Message RAM access failure occurred".
Another way is to have a bad hardware connection. Bad wire connection
can lead to this issue as well.

This patch fixes the issue according to provided workaround.

Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com>
Reviewed-by: Ludovic Desroches <ludovic.desroches@microchip.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/net/can/m_can/m_can.c