Re: [PATCH v3 2/2] hw/block/nvme: add simple copy command

[Klaus Jensen]

On Dec  9 07:13, Keith Busch wrote:
> On Tue, Dec 08, 2020 at 09:33:39AM +0100, Klaus Jensen wrote:
> > +static uint16_t nvme_copy(NvmeCtrl *n, NvmeRequest *req)
> > +{
> 
> <snip>
> 
> > +    for (i = 0; i < nr; i++) {
> > +        uint32_t _nlb = le16_to_cpu(range[i].nlb) + 1;
> > +        if (_nlb > le16_to_cpu(ns->id_ns.mssrl)) {
> > +            return NVME_CMD_SIZE_LIMIT | NVME_DNR;
> > +        }
> > +
> > +        nlb += _nlb;
> > +    }
> > +
> > +    if (nlb > le32_to_cpu(ns->id_ns.mcl)) {
> > +        return NVME_CMD_SIZE_LIMIT | NVME_DNR;
> > +    }
> > +
> > +    bounce = bouncep = g_malloc(nvme_l2b(ns, nlb));
> > +
> > +    for (i = 0; i < nr; i++) {
> > +        uint64_t slba = le64_to_cpu(range[i].slba);
> > +        uint32_t nlb = le16_to_cpu(range[i].nlb) + 1;
> > +
> > +        status = nvme_check_bounds(ns, slba, nlb);
> > +        if (status) {
> > +            trace_pci_nvme_err_invalid_lba_range(slba, nlb, 
> > ns->id_ns.nsze);
> > +            goto free_bounce;
> > +        }
> > +
> > +        if (NVME_ERR_REC_DULBE(ns->features.err_rec)) {
> > +            status = nvme_check_dulbe(ns, slba, nlb);
> > +            if (status) {
> > +                goto free_bounce;
> > +            }
> > +        }
> > +    }
> 
> Only comment I have is that these two for-loops look like they can be
> collaped into one, which also simplifies how you account for the bounce
> buffer when error'ing out. 
> 

Yeah. And the shadowing of nlb is not good either. I'll fix it up.

signature.asc
Description: PGP signature