kernel/drivers/mtd/nand/spi, branch linux-rolling-stable

mtd: spinand: Disable continuous read during probe

2026-03-04T12:21:29Z

[ Upstream commit b4af7d194dc879353829f3c56988a68fbba1fbdd ] Macronix serial NAND devices with continuous read support do not clear the configuration register on soft reset and lack a hardware reset pin. When continuous read is interrupted (e.g., during reboot), the feature remains enabled at the device level. With continuous read enabled, the OOB area becomes inaccessible and all reads are instead directed to the main area. As a result, during partition allocation as part of MTD device registration, the first two bytes of the main area for the master block are read and indicate that the block is bad. This process repeats for every subsequent block for the partition. All reads and writes that reference the BBT find no good blocks and fail. The only paths for recovery from this state are triggering the continuous read feature by way of raw MTD reads or through a NAND device power drain. Disable continuous read explicitly during spinand probe to ensure quiescent feature state. Fixes: 631cfdd0520d ("mtd: spi-nand: Add continuous read support") Cc: stable@vger.kernel.org Signed-off-by: David LaPorte Reviewed-by: Gunnar Kudrjavets Reviewed-by: Mikhail Kshevetskiy Signed-off-by: Miquel Raynal Signed-off-by: Sasha Levin

Revert "mtd: spinand: esmt: fix id code for F50D1G41LB"

2025-12-16T08:08:28Z

This reverts commit dd26402642a0899fde59ea6b0852fad3d799b4cc. The issue George met is due to the limit of QPIC, not the issue of the flash chip. QPIC only supports 4 bytes ID. So the fifth byte is always 0. If we use spi-gpio, the fifth byte can be read correctly. Signed-off-by: Ziyang Huang Signed-off-by: Miquel Raynal

Merge tag 'mtd/for-6.19' of git://git.kernel.org/pub/scm/linux/kernel/git/mtd/linux

2025-12-04T19:07:46Z

Pull mtd updates from Miquel Raynal: "Core MTD changes: - We must ignore error -ENOENT from parsers on subpartitions which is a legitimate return value - PM support is added to the intel-dg driver Raw NAND changes: - The major change in this is the support for the Allwinner H616 NAND controller, which lead to numerous changes and cleanups in the driver. - Another notable change in this driver is the use of field_get() and field_prep(), but since the global support for this helpers is going to be merged in the same release as we start using these helpers, we undefine them in the first place to avoid warnings. - Marvell drivers layout handling changes have also landed, they fix previous definitions and abuses that have been made previously, which implied to relax the ECC parameters validation in the core a bit. - The Cadence NAND controller driver gets NV-DDR interface support. SPI NAND changes: - Support for FudanMicro FM25S01BI3 and ESMT F50L1G41LC is added. SPI NOR changes: - Fix SMPT parsing for S25FS-S flash family. They report variable dummy cycles for reads. This results in the default of 0 being used. This works for other Infineon chips, but not for the S25FS-S family. They need 8 dummy cycles. Add fixup hooks to specify that. Also add fixup hooks to fix incorrect map ID data in SFDP. - Add support for a bunch of Winbond flashes. Their block protection information is not discoverable, so they need to have an entry in the flash tables to describe that. - Some cleanups for Micron flash support. - Add support for Micron mt35xu01gbba. - Some SPI controllers like the Intel one on the PCI bus do not support the read CR opcode (0x35). Do not use the opcode if the controller does not support it. Aside from these main changes, there is the usual load of API updates, kdoc fixes, potential memory leaks fixes, etc" * tag 'mtd/for-6.19' of git://git.kernel.org/pub/scm/linux/kernel/git/mtd/linux: (51 commits) mtd: sm_ftl: Fix typo in comment in sm_read_lba mtd: sm_ftl: Replace deprecated strncpy with sysfs_emit in sm_attr_show mtd: lpddr_cmds: fix signed shifts in lpddr_cmds mtd: docg3: fix kernel-doc warnings mtd: spinand: add support for FudanMicro FM25S01BI3 mtd: rawnand: renesas: Handle devm_pm_runtime_enable() errors mtd: nand: realtek-ecc: Fix Kconfig dependencies mtd: rawnand: sunxi: #undef field_{get,prep}() before local definition mailmap: update Pratyush Yadav's email address mtd: spi-nor: core: Check read CR support mtd: spi-nor: micron-st: add TODO for fixing mt35xu02gcba mtd: spi-nor: micron-st: add mt35xu01gbba support mtd: spi-nor: micron-st: use SFDP of mt35xu512aba mtd: spi-nor: micron-st: move set_octal_dtr to late_init() mtd: spi-nor: micron-st: rename the die_late_init functions mtd: spinand: esmt: add support for F50L1G41LC mtd: rawnand: lpc32xx_slc: Convert to use devm_gpiod_get_optional() mtd: mtdpart: ignore error -ENOENT from parsers on subpartitions mtd: maps: pcmciamtd: fix potential memory leak in pcmciamtd_detach() mtd: spi-nor: spansion: SMPT fixups for S25FS-S ...

mtd: spinand: add support for FudanMicro FM25S01BI3

2025-11-29T12:46:40Z

Add support for FudanMicro FM25S01BI3 SPI NAND. Link: https://www.fmsh.com/nvm/FM25S01BI3_ds_eng.pdf Signed-off-by: Mikhail Zhilkin Signed-off-by: Miquel Raynal

mtd: spinand: esmt: add support for F50L1G41LC

2025-11-17T10:53:08Z

This adds support for ESMT F50L1G41LC, which appears to be an updated version of the already supported F50L1G41LB. Add esmt_8c SPI_NAND manufacturer to account for the newly used vendor ID with support for the ESMT F50L1G41LC chip. Link: https://github.com/openwrt/openwrt/pull/15214#issuecomment-3514824435 Signed-off-by: Daniel Golle Signed-off-by: Miquel Raynal

mtd: spinand: fmsh: remove QE bit for FM25S01A flash

2025-11-03T10:33:04Z

According to datasheet (http://eng.fmsh.com/nvm/FM25S01A_ds_eng.pdf) there is no QE (Quad Enable) bit for FM25S01A flash, so remove it. Fixes: 5f284dc15ca86 ("mtd: spinand: add support for FudanMicro FM25S01A") Signed-off-by: Mikhail Kshevetskiy Tested-by: Tianling Shen Signed-off-by: Miquel Raynal

Merge tag 'mtd/for-6.18' of git://git.kernel.org/pub/scm/linux/kernel/git/mtd/linux

2025-10-04T22:50:37Z

Pull MTD updates from Miquel Raynal: "MTD core: - Bad blocks increment is skipped if the block is already known bad (improves user statistics relevance) - Expose the OOB layout via debugfs Raw NAND: - Add support for Loongson-2K1000 and Loongson-2K0500 NAND controllers, including extra features, such as chip select and 6-byte NAND ID reading support - Drop the s3c2410 driver SPI NAND: - Important SPI NAND continuous read improvements and fixes - Add support for FudanMicro FM25S01A - Add support for continuous reads in Gigadevice vendor driver ECC: - Add support for the Realtek ECC engine SPI NOR: - Some flashes can't perform reads or writes with start or end being an odd number in Octal DTR mode. File systems like UBIFS can request such reads or writes, causing the transaction to error out. Pad the read or write transactions with extra bytes to avoid this problem. And the usual amount of various miscellaneous fixes" * tag 'mtd/for-6.18' of git://git.kernel.org/pub/scm/linux/kernel/git/mtd/linux: (46 commits) mtd: rawnand: sunxi: drop unused module alias mtd: rawnand: stm32_fmc2: drop unused module alias mtd: rawnand: rockchip: drop unused module alias mtd: rawnand: pl353: drop unused module alias mtd: rawnand: omap2: drop unused module alias mtd: rawnand: atmel: drop unused module alias mtd: onenand: omap2: drop unused module alias mtd: hyperbus: hbmc-am654: drop unused module alias mtd: jedec_probe: use struct_size() helper for cfiq allocation mtd: cfi: use struct_size() helper for cfiq allocation mtd: nand: raw: gpmi: fix clocks when CONFIG_PM=N mtd: rawnand: omap2: fix device leak on probe failure mtd: rawnand: atmel: Fix error handling path in atmel_nand_controller_add_nands mtd: nand: realtek-ecc: Add Realtek external ECC engine support dt-bindings: mtd: Add realtek,rtl9301-ecc mtd: spinand: repeat reading in regular mode if continuous reading fails mtd: spinand: try a regular dirmap if creating a dirmap for continuous reading fails mtd: spinand: fix direct mapping creation sizes mtd: rawnand: fsmc: Default to autodetect buswidth mtd: nand: move nand_check_erased_ecc_chunk() to nand/core ...

mtd: spinand: repeat reading in regular mode if continuous reading fails

2025-09-29T15:53:08Z

Continuous reading may result in multiple flash pages reading in one operation. Unfortunately, not all spinand controllers support such large reading. They will read less data. Unfortunately, the operation can't be continued. In this case: * disable continuous reading on this (not good enough) spi controller * repeat reading in regular mode. Signed-off-by: Mikhail Kshevetskiy Signed-off-by: Miquel Raynal

mtd: spinand: try a regular dirmap if creating a dirmap for continuous reading fails

2025-09-29T15:53:08Z

Continuous reading may result in multiple flash pages reading in one operation. Typically only one flash page has read/written (a little bit more than 2-4 Kb), but continuous reading requires the spi controller to read up to 512 Kb in one operation without toggling CS in beetween. Roughly speaking spi controllers can be divided on 2 categories: * spi controllers without dirmap acceleration support * spi controllers with dirmap acceleration support Firt of them will have issues with continuous reading if restriction on the transfer length is implemented in the adjust_op_size() handler. Second group often supports acceleration of single page only reading. Thus enabling of continuous reading can break flash reading. This patch tries to create dirmap for continuous reading first and fallback to regular reading if spi controller refuses to create it. Signed-off-by: Mikhail Kshevetskiy Signed-off-by: Miquel Raynal

mtd: spinand: fix direct mapping creation sizes

2025-09-29T15:53:07Z

Continuous mode is only supported for data reads, thus writing requires only single flash page mapping. Signed-off-by: Mikhail Kshevetskiy Signed-off-by: Miquel Raynal