Use a more generic form for __section that requires quotes to avoid
complications with clang and gcc differences.

Remove the quote operator # from compiler_attributes.h __section macro.

Convert all unquoted __section(foo) uses to quoted __section("foo").
Also convert __attribute__((section("foo"))) uses to __section("foo")
even if the __attribute__ has multiple list entry forms.

Conversion done using the script at:

    https://lore.kernel.org/lkml/75393e5ddc272dc7403de74d645e6c6e0f4e70eb.camel@perches.com/2-convert_section.pl

Signed-off-by: Joe Perches <joe@perches.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@gooogle.com>
Reviewed-by: Miguel Ojeda <ojeda@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

HyperBus changes

* DMA support for TI's AM654 HyperBus controller driver.
* HyperBus frontend driver for Renesas RPC-IF driver.

NAND core changes:
* Use the new generic ECC object
* Create helpers to set/extract the ECC requirements
* Create a helper to extract the ECC configuration
* Add a NAND page I/O request type
* Introduce the ECC engine framework

Raw NAND core changes:
* Don't overwrite the error code from nand_set_ecc_soft_ops()
* Introduce nand_set_ecc_on_host_ops()
* Use the NAND framework user_conf object for ECC flags
* Use the ECC framework user input parsing bits
* Use the ECC framework nand_ecc_is_strong_enough() helper
* Use the ECC framework OOB layouts
* Make use of the ECC framework
* Use nanddev_get/set_ecc_requirements() when relevant
* Use the new ECC engine type enumeration
* Separate the ECC engine type and the ECC byte placement
* Move the nand_ecc_algo enum to the generic NAND layer
* Rename the ECC algorithm enumeration items
* Add a kernel doc to the ECC algorithm enumeration
* DT bindings:
  - Document boolean NAND ECC properties
  - Document nand-ecc-engine
  - Document nand-ecc-placement

Raw NAND drivers changes:
* Ams-Delta: Fix non-OF build warning
* Atmel:
  - Check return values for nand_read_data_op
  - Simplify with dev_err_probe()
  - Get rid of the legacy interface implementation
  - Convert the driver to exec_op()
  - Use nand_prog_page_end_op()
  - Use nand_{write,read}_data_op()
  - Drop redundant nand_read_page_op()
  - Enable the NFC controller at probe time
  - Disable clk on error handling path in probe
* Cadence: remove a redundant dev_err call
* Gpmi:
  - Simplify with dev_err_probe()
* Marvell:
  - Fix and update kerneldoc
  - Simplify with dev_err_probe()
  - Fix and update kerneldoc
  - Simplify with dev_err_probe()
  - Support panic_write for mtdoops
* Onenand:
  - Simplify the return expression of onenand_transfer_auto_oob
  - Simplify with dev_err_probe()
* Oxnas: cleanup/simplify code
* Pasemi: Make pasemi_device_ready() static
* Qcom: Simplify with dev_err_probe()
* Stm32_fmc2: fix a buffer overflow
* Vf610: Remove unused function vf610_nfc_transfer_size()

SPI-NAND changes:
* Use nanddev_get_ecc_conf() when relevant
* Gigadevice:
  - Add support for GD5F4GQ4xC
  - Add QE Bit
  - Use only one dummy byte in QUADIO
* Macronix:
  - Add support for MX31UF1GE4BC
  - Add support for MX31LF1GE4BC

Add the HyperFLash driver for the Renesas RPC-IF.  It's the "front end"
driver using the "back end" APIs in the main driver to talk to the real
hardware.

Signed-off-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Link: https://lore.kernel.org/r/78abb851-2beb-fe7d-87e5-ce58ee877d35@gmail.com

Instead of storing the ECC flags in chip->ecc.options, use
nanddev->ecc.user_conf.flags.

There is currently only one to save: NAND_ECC_MAXIMIZE.

Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200827085208.16276-21-miquel.raynal@bootlin.com

Many helpers are generic to all NAND chips, they should not be
raw-NAND specific, so use the generic ones.

To avoid moving all the raw NAND core "history" into the generic NAND
layer, we keep a part of this parsing in the raw NAND core to ensure
backward compatibility.

Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200827085208.16276-20-miquel.raynal@bootlin.com

No need to have our own in the raw NAND core.

Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200827085208.16276-18-miquel.raynal@bootlin.com

Provide per device private pointer that can be used by controller
drivers to store device specific private data.

Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
Link: https://lore.kernel.org/r/20200924081214.16934-2-vigneshr@ti.com

Embed a generic NAND ECC high-level object in the nand_device
structure to carry all the ECC engine configuration/data.

Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200827085208.16276-16-miquel.raynal@bootlin.com

Despite its current name, the eccreq field actually encodes both the
NAND requirements and the final ECC configuration. That works fine when
using on-die ECC since those 2 concepts match perfectly, but it starts
being a problem as soon as we use on-host ECC engines, where we're not
guaranteed to have a perfect match.

Let's hide the ECC requirements access behind helpers so we can later
split those 2 concepts. As the structures have not been clarified yet,
these helpers access the same internal variable as
nanddev_get_ecc_conf() for now.

Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200827085208.16276-14-miquel.raynal@bootlin.com