On Thu, 4 Apr 2019 21:56:19 +0300
Sergei Shtylyov <sergei.shtylyov@xxxxxxxxxxxxxxxxxx> wrote:
> Hello!
>
> On 04/03/2019 12:20 PM, masonccyang@xxxxxxxxxxx wrote:
>
> >> >> > +static void rpc_spi_mem_set_prep_op_cfg(struct spi_device *spi,
> >> >> > + const struct spi_mem_op *op,
> >> >> > + u64 *offs, size_t *len)
> >> >> > +{
> >> >> > + struct rpc_spi *rpc = spi_controller_get_devdata(spi->controller);
> >> >> > +
> >> >> > + rpc->cmd = RPC_SMCMR_CMD(op->cmd.opcode);
> >> >> > + rpc->smenr = RPC_SMENR_CDE |
> >> >> > + RPC_SMENR_CDB(ilog2(op->cmd.buswidth));
> >> >> > + rpc->totalxferlen = 1;
> >> >> > + rpc->xfer_dir = SPI_MEM_NO_DATA;
> >> >> > + rpc->xferlen = 0;
> >> >> > + rpc->addr = 0;
> >> >> > +
> >> >> > + if (op->addr.nbytes) {
> >> >> > + rpc->smenr |= RPC_SMENR_ADB(ilog2(op->addr.buswidth));
> >> >> > + if (op->addr.nbytes == 4)
> >> >> > + rpc->smenr |= RPC_SMENR_ADE(0xf);
> >> >> > + else
> >> >> > + rpc->smenr |= RPC_SMENR_ADE(0x7);
> >> >> > +
> >> >> > + if (offs && len)
> >> >> > + rpc->addr = *offs;
> >> >> > + else
> >> >> > + rpc->addr = op->addr.val;
> >> >> > + rpc->totalxferlen += op->addr.nbytes;
> >> >> > + }
> >> >> > +
> >> >> > + if (op->dummy.nbytes) {
> >> >> > + rpc->smenr |= RPC_SMENR_DME;
> >> >> > + rpc->dummy = RPC_SMDMCR_DMCYC(op->dummy.nbytes);
> >> >>
> >> >> So you haven't fixed this bug? I repeat, the driver doesn't work right
> >> >> w/o this fixed!
> >> >
> >> > Do you mean
> >> >
> >> > rpc->dummy = RPC_SMDMCR_DMCYC(op->dummy.nbytes) * 8; ?
> >>
> >> Yes. Or some other code that amounts to it.
>
> Oops, I wasn't attentive enough, I was proposing:
>
> rpc->dummy = RPC_SMDMCR_DMCYC(op->dummy.nbytes * 8);
>
> > why not setting "op->dummy.nbytes = 7" from spi-nor.c protocol layer ?
>
> We want 8 dummy clocks, not 8 dummy bytes. And if you'd looked into
> spi_nor_read_sfdp(), you'd have seen that it requests 8 dummy clocks already.
>
> > driver may check device ID or something else to setup op->dummy.nbytes.
>
> The RDSFDP command is not chip specific.
>
> > I think it is not a good idea to *8 in spi host driver.
>
> >> > What is your flash part number?
> >>
> >> Spansion/Cypress S25FS512S. The datasheet says there must be 8 dummy cycles
> >> for the RSFDP command...
> >>
> >> > The problem you found is in 1 bit or 4 bits width ?
> >>
> >> 1-bit, I think.
> >>
> >> >>
> >> >> [...]
> >> >> > +static ssize_t rpc_spi_mem_dirmap_read(struct spi_mem_dirmap_desc *desc,
> >> >> > + u64 offs, size_t len, void *buf)
> >> >> > +{
> >> >> > + struct rpc_spi *rpc =
> >> >> > + spi_controller_get_devdata(desc->mem->spi->controller);
> >> >> > + int ret;
> >> >> > +
> >> >> > + if (offs + desc->info.offset + len > U32_MAX)
> >> >> > + return -EINVAL;
> >> >> > +
> >> >> > + if (len > 0x4000000)
> >> >> > + len = 0x4000000;
> >> >>
> >> >> Ugh...
> >> >
> >> > by mtd->size ?
> >>
> >> That'd be better, yes.
> >>
> >
> > Oops, it seems hard to get mtd->size info. from spi_mem_dirmap,
>
> It's in desc->info.length, no?
It's the lengths of the mapping which not necessarily mtd->size, but in
the SPI NOR case it is :-). Anyway, you should not assume
dirmapdesc->info.length == memory_device->size.
>
> > I would like to keep 0x4000000.
>
> I'm seeing Boris in the CC's... Boris, is it legitimate to limit
> a single dirmap read by the memory "window" size? Or should we try to
> serve any valid transfer length?
If by memory window you're talking about the memory region reserved in
the CPU address space, then no. It should not be limited to this size
if possible. Most HW have a way to configure an offset to apply to the
spi-mem operation, and in that case, the driver should change this
offset on the fly when one tries to access a region that's outside of
the currently configured window.
>
> >> >> > +
> >> >> > + ret = rpc_spi_set_freq(rpc, desc->mem->spi->max_speed_hz);
> >> >> > + if (ret)
> >> >> > + return ret;
> >> >> > +
> >> >> > + rpc_spi_mem_set_prep_op_cfg(desc->mem->spi,
> >> >> > + &desc->info.op_tmpl, &offs, &len);
> >> >> > +
> >> >> > + regmap_update_bits(rpc->regmap, RPC_CMNCR, RPC_CMNCR_MD, 0);
> >> >> > + regmap_write(rpc->regmap, RPC_DRCR, RPC_DRCR_RBURST(32) |
> >> >> > + RPC_DRCR_RBE);
> >> >> > +
> >> >> > + regmap_write(rpc->regmap, RPC_DRCMR, rpc->cmd);
> >> >> > + regmap_write(rpc->regmap, RPC_DREAR, RPC_DREAR_EAC(1));
> >> >>
> >> >> So you're not even trying to support flashes larger than the
> >> read dirmap?
> >> >> Now I don't think it's acceptable (and I have rewritten this code
> >> internally).
> >> >
> >> > what about the size comes form mtd->size ?
> >>
> >> I'm not talking about size here; we should use the full address.
> >> I'm attaching
> >> my patch...
> >
> > okay,got it!
> > what about just
> > - regmap_write(rpc->mfd->regmap, RPC_DREAR, RPC_DREAR_EAC(1));
> > + regmap_write(rpc->mfd->regmap, RPC_DREAR,
> > + RPC_DREAR_EAV(offs >> 25) | RPC_DREAR_EAC(1));
> >
> > because only > 64MByte size make RPC_DREAR_EAV() work.
>
> Boris, what's your opinion on this?
> Note that for the write dirmap we have just 256-byte buffer (reusing
> the read cache). Is it legitimate to limit the served length to 256 bytes?
I don't know what the HW is capable of, but drivers should use any try
they have to dynamically move the memory map window (make it point at a
different spi-mem offset on demand). Note that dirmap_read/write() are
allowed to return less than the amount of data requested, in that case
the caller should continue reading at the offset where things stopped.
This avoids having to implement a loop that splits things in several
accesses when the access cannot be done in one step.
