[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <5B8DA87D05A7694D9FA63FD143655C1B9DCB0429@hasmsx108.ger.corp.intel.com>
Date: Thu, 5 Sep 2019 13:00:18 +0000
From: "Winkler, Tomas" <tomas.winkler@...el.com>
To: Paul Menzel <pmenzel@...gen.mpg.de>
CC: "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"Usyskin, Alexander" <alexander.usyskin@...el.com>
Subject: RE: Could not read FW version, FW version command failed -5
> Dear Tomas,
>
>
> Testing Fedora 30 with Linux 5.2.11 on an old Dell OptiPlex 980, Linux
> log the message below.
>
> [ 15.964298] mei mei::55213584-9a29-4916-badf-0fb7ed682aeb:01: Could
> not read FW version
> [ 15.964301] mei mei::55213584-9a29-4916-badf-0fb7ed682aeb:01: FW
> version command failed -5
>
> The prints happen in `drivers/misc/mei/bus-fixup.c`.
>
> ret = 0;
> bytes_recv = __mei_cl_recv(cldev->cl, buf, sizeof(buf), 0,
> MKHI_RCV_TIMEOUT);
> if (bytes_recv < 0 || (size_t)bytes_recv < MKHI_FWVER_LEN(1)) {
> /*
> * Should be at least one version block,
> * error out if nothing found
> */
> dev_err(&cldev->dev, "Could not read FW version\n");
> return -EIO;
> }
>
> The Web is full of these messages, so it does seem like a normal
> condition on certain devices. Do you know the reason? Do older Intel
> MEs support this?
Wow, this quite an old system ~ 2009, the FW version has a bit different format there, it was overlooked, will look at that ASAP, thanks for the report.
Tomas
Powered by blists - more mailing lists