Date: Fri, 13 Sep 2019 23:08:22 -0600 From: Warner Losh <imp@bsdimp.com> To: Sebastian Huber <sebastian.huber@embedded-brains.de> Cc: FreeBSD Hackers <freebsd-hackers@freebsd.org> Subject: Re: Problems with port of NVMe support Message-ID: <CANCZdfrGov3bu631aNjqgce-Ki1tHAQ2VRTrSx%2Bo3uwP0-Xf6w@mail.gmail.com> In-Reply-To: <dd216882-2375-7751-ca94-57caec3763c5@embedded-brains.de> References: <dd216882-2375-7751-ca94-57caec3763c5@embedded-brains.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, Sep 13, 2019, 12:02 AM Sebastian Huber < sebastian.huber@embedded-brains.de> wrote: > Hello, > > I port currently the NVMe support from FreeBSD to RTEMS. Basically, it > works. I can issue administration commands and some IO commands (e.g. > FLUSH). I use this SSD for testing: > > $ nvmecontrol identify nvme0 > Controller Capabilities/Features > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D > Vendor ID: 8086 > Subsystem Vendor ID: 8086 > Serial Number: PHM27390006S280AGN > Model Number: INTEL SSDPE21D280GA > Firmware Version: E2010325 > Recommended Arb Burst: 0 > IEEE OUI Identifier: e4 d2 5c > Multi-Path I/O Capabilities: Not Supported > Max Data Transfer Size: 131072 > Controller ID: 0x0000 > Version: 0.0.0 > > My problem is now, that every WRITE transfer ends up in a DATA TRANSFER > ERROR: > > nvme0: WRITE sqid:1 cid:127 nsid:1 lba:0 len:8 > nvme0: DATA TRANSFER ERROR (00/04) sqid:1 cid:127 cdw0:0 > nvme0: async event occurred (type 0x0, info 0x04, page 0x01) > > Every READ transfer ends up in a UNRECOVERED READ ERROR: > > nvme0: READ sqid:1 cid:127 nsid:1 lba:0 len:8 > nvme0: UNRECOVERED READ ERROR (02/81) sqid:1 cid:127 cdw0:0 > nvme0: async event occurred (type 0x0, info 0x04, page 0x01) > > I see the transfers in the error log: > > $ nvmecontrol logpage -p 1 nvme0 > > > Error Information Log > > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > > > Entry 01 > > > =3D=3D=3D=3D=3D=3D=3D=3D=3D > > > Error count: 2546 > > > Submission queue ID: 1 > > > Command ID: 32512 > > > Status: > > > Phase tag: 0 > > > Status code: 129 > > > Status code type: 2 > > > More: 1 > > > DNR: 1 > > > Error location: 0 > > > LBA: 0 > > > Namespace ID: 1 > > > Vendor specific info: 0 > > > Transport type: 0 > > > Command specific info:0 > > > Transport specific: 0 > > > Entry 02 > > > =3D=3D=3D=3D=3D=3D=3D=3D=3D > > > Error count: 2545 > > > Submission queue ID: 1 > > > Command ID: 32512 > > > Status: > > > Phase tag: 0 > > > Status code: 4 > > > Status code type: 0 > > > More: 1 > > > DNR: 1 > > > Error location: 0 > > > LBA: 0 > > > Namespace ID: 1 > > > Vendor specific info: 0 > > > Transport type: 0 > > > Command specific info:0 > > > Transport specific: 0 > > From the NVMe specification I didn't get any hints what the error > reason could be. Maybe someone has a hint for me. The platform is a > T4240 PowerPC (big-endian). > Maybe the physaddr in the SG list for these commands doesn't translate right? Warner --=20 > Sebastian Huber, embedded brains GmbH > > Address : Dornierstr. 4, D-82178 Puchheim, Germany > Phone : +49 89 189 47 41-16 > Fax : +49 89 189 47 41-09 > E-Mail : sebastian.huber@embedded-brains.de > PGP : Public key available on request. > > Diese Nachricht ist keine gesch=C3=A4ftliche Mitteilung im Sinne des EHUG= . > _______________________________________________ > freebsd-hackers@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-hackers > To unsubscribe, send any mail to "freebsd-hackers-unsubscribe@freebsd.org= " >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CANCZdfrGov3bu631aNjqgce-Ki1tHAQ2VRTrSx%2Bo3uwP0-Xf6w>