Date: Fri, 9 Sep 2016 11:49:25 +0100 From: Dom <misc-freebsd@talk2dom.com> To: freebsd-virtualization@freebsd.org Subject: Re: Issues with GTX960 on CentOS7 using bhyve PCI passthru (FreeBSD 11-RC2) Message-ID: <2a0d0094-3263-de91-b3f8-38dda4a523ac@talk2dom.com> In-Reply-To: <22df7de9-1dfb-561c-125d-99af2b3da70b@talk2dom.com> References: <727839e2-4be1-5661-8c19-6d13f73ad36f@talk2dom.com> <62b2a72a-3662-5b3e-5029-f284a625d05b@freebsd.org> <22df7de9-1dfb-561c-125d-99af2b3da70b@talk2dom.com>
next in thread | previous in thread | raw e-mail | index | archive | help
It looks like there may not be an issue with MSI after all. The nvidia driver is issued an IRQ when first used, not at boot time. If I run the CUDA "deviceQuery" sample then this appears in dmesg: [ 67.207929] nvidia 0000:00:06.0: irq 29 for MSI/MSI-X [ 67.646207] NVRM: RmInitAdapter failed! (0x24:0x1f:1356) [ 67.646570] NVRM: rm_init_adapter failed for device bearing minor number 0 [ 67.647214] NVRM: nvidia_frontend_open: minor 0, module->open() failed, error -5 Maybe the IRQ is deallocated immediately so doesn't appear in the output of /proc/interrupts? I guess I'll need to research the NVRM error above now but at least this thread might be useful regarding the BAR allocation.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?2a0d0094-3263-de91-b3f8-38dda4a523ac>