Hi Miquel, Sorry, previous email missed this mxic_nand_data_xfer() reply. This Flash Memory Controller implemented the Buffer read-write data transfer for SPI mode and raw NAND mode. That is each time driver write to the transmit of the TXD port and the data is shifted out, new data is received in RXD port. By transmitting the entire buffer without reading any data, driver are losing the received data. Actually the mxic_nand_data_xfer() is a copy of mxic_spi_data_xfer(). https://github.com/torvalds/linux/blame/master/drivers/spi/spi-mxic.c therefore, I plan to patch this part to MFD as a common code for both raw NAND and SPI. i.e,. In driver/mfd/mxic-mfd.c, we have mxic_mfd_data_xfer() and here call mxic_mfd_data_xfer() for raw NAND data read-write. > > > +static int mxic_nand_data_xfer(struct mxic_nand_ctlr *mxic, const void *txbuf, > > > + void *rxbuf, unsigned int len) > > > +{ > > > > There is not so much code shared, why not separating this function for > > rx and tx cases? > > > > > + unsigned int pos = 0; > > > + > > > + while (pos < len) { > > > + unsigned int nbytes = len - pos; > > > + u32 data = 0xffffffff; > > > + u32 sts; > > > + int ret; > > > + > > > + if (nbytes > 4) > > > + nbytes = 4; > > > + > > > + if (txbuf) > > > + memcpy(&data, txbuf + pos, nbytes); > > > + > > > + ret = readl_poll_timeout(mxic->mfd->regs + INT_STS, sts, > > > + sts & INT_TX_EMPTY, 0, USEC_PER_SEC); > > > > Using USEC_PER_SEC for a delay is weird > > > > > + if (ret) > > > + return ret; > > > + > > > + writel(data, mxic->mfd->regs + TXD(nbytes % 4)); > > > + > > > + if (rxbuf) { > > > + ret = readl_poll_timeout(mxic->mfd->regs + INT_STS, sts, > > > + sts & INT_TX_EMPTY, 0, > > > + USEC_PER_SEC); > > > + if (ret) > > > + return ret; > > > + > > > + ret = readl_poll_timeout(mxic->mfd->regs + INT_STS, sts, > > > + sts & INT_RX_NOT_EMPTY, 0, > > > + USEC_PER_SEC); > > > + if (ret) > > > + return ret; > > > + > > > + data = readl(mxic->mfd->regs + RXD); > > > + data >>= (8 * (4 - nbytes)); > > > > What is this? Are you trying to handle some endianness issue? > > yes, > > > > > > + memcpy(rxbuf + pos, &data, nbytes); > > > + WARN_ON(readl(mxic->mfd->regs + INT_STS) & > > > + INT_RX_NOT_EMPTY); > > > + } else { > > > + readl(mxic->mfd->regs + RXD); > > > + } > > > + WARN_ON(readl(mxic->mfd->regs + INT_STS) & INT_RX_NOT_EMPTY); > > > + > > > + pos += nbytes; > > > + } > > > + > > > + return 0; > > > +} thanks for your review. best regards, Mason -- CONFIDENTIALITY NOTE: This e-mail and any attachments may contain confidential information and/or personal data, which is protected by applicable laws. Please be reminded that duplication, disclosure, distribution, or use of this e-mail (and/or its attachments) or any part thereof is prohibited. If you receive this e-mail in error, please notify us immediately and delete this mail as well as its attachment(s) from your system. In addition, please be informed that collection, processing, and/or use of personal data is prohibited unless expressly permitted by personal data protection laws. Thank you for your attention and cooperation. Macronix International Co., Ltd. ===================================================================== ============================================================================ CONFIDENTIALITY NOTE: This e-mail and any attachments may contain confidential information and/or personal data, which is protected by applicable laws. Please be reminded that duplication, disclosure, distribution, or use of this e-mail (and/or its attachments) or any part thereof is prohibited. If you receive this e-mail in error, please notify us immediately and delete this mail as well as its attachment(s) from your system. In addition, please be informed that collection, processing, and/or use of personal data is prohibited unless expressly permitted by personal data protection laws. Thank you for your attention and cooperation. Macronix International Co., Ltd. =====================================================================