Hi Moritz,
On Mon, 8 Apr 2019 11:05:49 -0700
Moritz Fischer mdf@xxxxxxxxxx wrote:
>Hi Anatolij,
>
>looks mostly good, a couple of nits inline.
Thanks for review.
...
>> +++ b/drivers/fpga/ftdi-fifo-fpp.c
>> @@ -0,0 +1,594 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/*
>> + * Altera FPGA firmware upload via FPP using FT232H Bitbang/FT245-FIFO.
>> + *
>> + * Copyright (C) 2017 - 2018 DENX Software Engineering
>Maybe bump that to 2019 ;-)
I'll update it in next version.
...
>> + gpiod_set_raw_value_cansleep(priv->nconfig, 1);
>> + gpiod_set_raw_value_cansleep(priv->conf_done, 0);
>> + msleep(50);
>
>You could use a constant here for the 50 ...
Do you mean something like below?
const unsigned int wait_50ms = 50;
...
msleep(wait_50ms);
...
>and here ...
>> + gpiod_set_value_cansleep(priv->nconfig, 1);
>> + msleep(50);
>and here ...
>> + gpiod_set_value_cansleep(priv->nconfig, 0);
>> +
>> + /* Wait for CONF_DONE to get low */
>Does the retry number come from datasheet?
No, it is an empirical value.
...
>> + if (!retries) {
>> + dev_warn(dev, "CONF_DONE wait timeout\n");
>
>warn or err?
err fits better here, I'll update to dev_err().
...
>> + struct fpp_fpga_mgr_priv *priv = mgr->priv;
>> + struct device *dev = &priv->pdev->dev;
>> + int retries = MAX_RETRIES;
>> + int ret;
>> + u8 mask, status;
>
>Can you swap the two lines, reverse xmas-tree please.
Will do in v5.
Thanks,
Anatolij
