From: hch@lst.de (Christoph Hellwig) Date: Thu, 2 Nov 2017 20:44:08 +0100 Subject: [PATCH 2/2] nvme: Check admin passthru command effects In-Reply-To: <20171101180400.6619-2-keith.busch@intel.com> References: <20171101180400.6619-1-keith.busch@intel.com> <20171101180400.6619-2-keith.busch@intel.com> Message-ID: <20171102194408.GB5270@lst.de> On Wed, Nov 01, 2017@12:04:00PM -0600, Keith Busch wrote: > The NVMe standard provides a command effects log page so the host may > be aware of special requirements it may need to do in response to a > particular command. For example, the command may need to run with IO > quiesced to prevent timeouts or undefined behavior, or it may change > the logical block formats that determine how the host needs to construct > future IO commands. > > This patch saves the nvme command effects log page if the controller > supports it, and performs appropriate actions before and after an admin > passthrough command is completed. If the controller does not support the > command effects log page, the driver will define the effects for known > opcodes. The nvme format and santize are the only commands in this patch > with known effects. > > Signed-off-by: Keith Busch > --- > drivers/nvme/host/core.c | 116 +++++++++++++++++++++++++++++++++++++++++++++++ > drivers/nvme/host/nvme.h | 1 + > include/linux/nvme.h | 19 ++++++++ > 3 files changed, 136 insertions(+) > > diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c > index bae987b5dc79..a3444a67b1bc 100644 > --- a/drivers/nvme/host/core.c > +++ b/drivers/nvme/host/core.c > @@ -72,6 +72,9 @@ static DEFINE_IDA(nvme_instance_ida); > static dev_t nvme_chr_devt; > static struct class *nvme_class; > > +static void nvme_ns_remove(struct nvme_ns *ns); > +static int nvme_revalidate_disk(struct gendisk *disk); > + > static __le32 nvme_get_log_dw10(u8 lid, size_t size) > { > return cpu_to_le32((((size / 4) - 1) << 16) | lid); > @@ -990,12 +993,96 @@ static int nvme_submit_io(struct nvme_ns *ns, struct nvme_user_io __user *uio) > metadata, meta_len, io.slba, NULL, 0); > } > > +static u32 nvme_get_effects(struct nvme_ctrl *ctrl, struct nvme_ns *ns, > + u8 opcode) > +{ > + if (ctrl->effects) { > + if (ns) > + return le32_to_cpu(ctrl->effects->iocs[opcode]); > + return le32_to_cpu(ctrl->effects->acs[opcode]); > + } > + if (ns) > + return 0; This looks like we should simply have two different versions for admin vs I/O commands. > +static u32 nvme_passthru_start(struct nvme_ctrl *ctrl, struct nvme_ns *ns, > + u8 opcode) > +{ > + u32 effects = nvme_get_effects(ctrl, ns, opcode); > + > + /* > + * XXX: IO command effects are not handled. > + */ > + if (ns && (effects & ~NVME_CMD_EFFECTS_CSUPP)) { > + dev_warn(ctrl->device, > + "IO command;%02x has unhandled effects:%08x\n", > + opcode, effects); > + return 0; > + } Maybe key it odd here: if (ns) { /* * XXX: IO command effects are not handled. */ effects = le32_to_cpu(ctrl->effects->iocs[opcode]); if (ns && (effects & ~NVME_CMD_EFFECTS_CSUPP)) { dev_warn(ctrl->device, "IO command;%02x has unhandled effects:%08x\n", opcode, effects); return 0; } } else { effects = le32_to_cpu(ctrl->effects->acs[opcode]); if (!effects) effects = nvme_known_admin_effects(ctrl, opcode); } Otherwise this looks fine to me.