Conversation

Takashi Sakamoto (坂本 貴史)

Edited 1 month ago

Hm. A commit 29a43dc130ce (“PCI: Mark LSI FW643 to avoid bus reset”) should be reverted, since it would include some misunderstandings.

1
0
0

Takashi Sakamoto (坂本 貴史)

@gregkh Yup. I'm waiting for the release of -rc1 kernel to start the action.

I've been using the issued 1394 OHCI hardware in recent years to develop both Linux FireWire subsytem and IEC 61883-1/6 packet streaming engine in guest system to which the 1394 OHCI hardware is passedthough from host system. I have never experienced the reported issue.

In my opinion, the issue is in the PCIe host bridge or something like it in the reported Apple hardware which provides the secondary bus to the 1394 OHCI hardware. As long as I know, the reported Apple hardware has no IEEE 1394 interface, thus the patch author probably used some kind of specific hardware to connect the 1394 OHCI hardware. I guess that it should be [Apple thunderbolt display](https://lore.kernel.org/linux-pci/1372860295-8306-1-git-send-email-mika.westerberg@linux.intel.com/) or Apple thunderbolt 2 to FireWire adapter.

Anyway, it is better to ask the author what combination of hardware is used.
0
0
0