remove the legacy ide driver v2

remove the legacy ide driver v2

[Christoph Hellwig]

Hi all,

we've been trying to get rid of the legacy ide driver for a while now,
and finally scheduled a removal for 2021, which is three month old now.

In general distros and most defconfigs have switched to libata long ago,
but there are a few exceptions.  This series first switches over all
remaining defconfigs to use libata and then removes the legacy ide
driver.

libata mostly covers all hardware supported by the legacy ide driver.
There are three mips drivers that are not supported, but the linux-mips
list could not identify any users of those.  There also are two m68k
drivers that do not have libata equivalents, which might or might not
have users, so we'll need some input and possibly help from the m68k
community here.

This series is against Jens' for-5.14/libata branch.

Changes since v1:
 - dropped various already merged patches
 - add a new module option to allow disabling DMA just for the quirky
   cypress devices found on some stone age ARM and Alpha systems

[PATCH 1/6] pata_cypress: add a module option to disable BM-DMA

[Christoph Hellwig]

Multiple users report that they need to disable DMA on this driver,
so provide an option to do so without affecting all of libata.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/ata/pata_cypress.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/ata/pata_cypress.c b/drivers/ata/pata_cypress.c
index e1486fe298ae..5b3a7a8ebef6 100644
--- a/drivers/ata/pata_cypress.c
+++ b/drivers/ata/pata_cypress.c
@@ -41,6 +41,10 @@ enum {
 	CY82_INDEX_TIMEOUT	= 0x32
 };
 
+static bool enable_dma = true;
+module_param(enable_dma, bool, 0);
+MODULE_PARM_DESC(enable_dma, "Enable bus master DMA operations");
+
 /**
  *	cy82c693_set_piomode	-	set initial PIO mode data
  *	@ap: ATA interface
@@ -124,14 +128,16 @@ static struct ata_port_operations cy82c693_port_ops = {
 
 static int cy82c693_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
 {
-	static const struct ata_port_info info = {
+	static struct ata_port_info info = {
 		.flags = ATA_FLAG_SLAVE_POSS,
 		.pio_mask = ATA_PIO4,
-		.mwdma_mask = ATA_MWDMA2,
 		.port_ops = &cy82c693_port_ops
 	};
 	const struct ata_port_info *ppi[] = { &info, &ata_dummy_port_info };
 
+	if (enable_dma)
+		info.mwdma_mask = ATA_MWDMA2;
+
 	/* Devfn 1 is the ATA primary. The secondary is magic and on devfn2.
 	   For the moment we don't handle the secondary. FIXME */
 
-- 
2.30.2

Re: [PATCH 1/6] pata_cypress: add a module option to disable BM-DMA

[Russell King (Oracle)]

On Wed, Jun 16, 2021 at 03:46:53PM +0200, Christoph Hellwig wrote:
> Multiple users report that they need to disable DMA on this driver,
> so provide an option to do so without affecting all of libata.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Thanks!

Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!

[PATCH 2/6] alpha: use libata instead of the legacy ide driver

[Christoph Hellwig]

Switch the alpha defconfig from the legacy ide driver to libata.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 arch/alpha/configs/defconfig | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/arch/alpha/configs/defconfig b/arch/alpha/configs/defconfig
index 724c4075df40..dd2dd9f0861f 100644
--- a/arch/alpha/configs/defconfig
+++ b/arch/alpha/configs/defconfig
@@ -25,19 +25,18 @@ CONFIG_PNP=y
 CONFIG_ISAPNP=y
 CONFIG_BLK_DEV_FD=y
 CONFIG_BLK_DEV_LOOP=m
-CONFIG_IDE=y
-CONFIG_BLK_DEV_IDECD=y
-CONFIG_IDE_GENERIC=y
-CONFIG_BLK_DEV_GENERIC=y
-CONFIG_BLK_DEV_ALI15X3=y
-CONFIG_BLK_DEV_CMD64X=y
-CONFIG_BLK_DEV_CY82C693=y
 CONFIG_SCSI=y
 CONFIG_BLK_DEV_SD=y
 CONFIG_BLK_DEV_SR=y
 CONFIG_SCSI_AIC7XXX=m
 CONFIG_AIC7XXX_CMDS_PER_DEVICE=253
 # CONFIG_AIC7XXX_DEBUG_ENABLE is not set
+CONFIG_ATA=y
+# CONFIG_SATA_PMP is not set
+CONFIG_PATA_ALI=y
+CONFIG_PATA_CMD64X=y
+CONFIG_PATA_CYPRESS=y
+CONFIG_ATA_GENERIC=y
 CONFIG_NETDEVICES=y
 CONFIG_DUMMY=m
 CONFIG_NET_ETHERNET=y
-- 
2.30.2

[PATCH 3/6] ARM: disable CONFIG_IDE in footbridge_defconfig

[Christoph Hellwig]

footbridge_defconfig enables CONFIG_IDE but no actual host controller
driver, so just drop it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 arch/arm/configs/footbridge_defconfig | 1 -
 1 file changed, 1 deletion(-)

diff --git a/arch/arm/configs/footbridge_defconfig b/arch/arm/configs/footbridge_defconfig
index 2aa3ebeb89d7..7a32de51f0fa 100644
--- a/arch/arm/configs/footbridge_defconfig
+++ b/arch/arm/configs/footbridge_defconfig
@@ -64,7 +64,6 @@ CONFIG_PARIDE_ON26=m
 CONFIG_BLK_DEV_LOOP=m
 CONFIG_BLK_DEV_NBD=m
 CONFIG_BLK_DEV_RAM=y
-CONFIG_IDE=y
 CONFIG_NETDEVICES=y
 CONFIG_NET_ETHERNET=y
 CONFIG_NET_VENDOR_3COM=y
-- 
2.30.2

[PATCH 4/6] ARM: disable CONFIG_IDE in pxa_defconfig

[Christoph Hellwig]

pxa_defconfig already enables libata including the pata_pcmcia driver, so
drop the legacy ide driver and idecs host driver.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 arch/arm/configs/pxa_defconfig | 2 --
 1 file changed, 2 deletions(-)

diff --git a/arch/arm/configs/pxa_defconfig b/arch/arm/configs/pxa_defconfig
index 875a3c28a267..363f1b1b08e3 100644
--- a/arch/arm/configs/pxa_defconfig
+++ b/arch/arm/configs/pxa_defconfig
@@ -215,8 +215,6 @@ CONFIG_IIO=m
 CONFIG_AD5446=m
 CONFIG_EEPROM_AT24=m
 CONFIG_SENSORS_LIS3_SPI=m
-CONFIG_IDE=m
-CONFIG_BLK_DEV_IDECS=m
 CONFIG_SCSI=y
 CONFIG_BLK_DEV_SD=m
 CONFIG_CHR_DEV_ST=m
-- 
2.30.2

[PATCH 5/6] m68k: use libata instead of the legacy ide driver

[Christoph Hellwig]

Switch the m68 defconfigs from the deprecated ide subsystem to use libata
instead.  The gayle and buddha and falcon drivers are enabled for libata,
while support for the q40 and macide drivers is lost.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 arch/m68k/configs/amiga_defconfig | 10 +++++-----
 arch/m68k/configs/atari_defconfig |  8 ++++----
 arch/m68k/configs/mac_defconfig   |  8 ++++----
 arch/m68k/configs/multi_defconfig | 14 +++++++-------
 arch/m68k/configs/q40_defconfig   |  8 ++++----
 5 files changed, 24 insertions(+), 24 deletions(-)

diff --git a/arch/m68k/configs/amiga_defconfig b/arch/m68k/configs/amiga_defconfig
index 59b727b69357..4fe26d54627e 100644
--- a/arch/m68k/configs/amiga_defconfig
+++ b/arch/m68k/configs/amiga_defconfig
@@ -323,11 +323,6 @@ CONFIG_BLK_DEV_RAM=y
 CONFIG_CDROM_PKTCDVD=m
 CONFIG_ATA_OVER_ETH=m
 CONFIG_DUMMY_IRQ=m
-CONFIG_IDE=y
-CONFIG_IDE_GD_ATAPI=y
-CONFIG_BLK_DEV_IDECD=y
-CONFIG_BLK_DEV_GAYLE=y
-CONFIG_BLK_DEV_BUDDHA=y
 CONFIG_RAID_ATTRS=m
 CONFIG_SCSI=y
 CONFIG_BLK_DEV_SD=y
@@ -344,6 +339,11 @@ CONFIG_GVP11_SCSI=y
 CONFIG_SCSI_A4000T=y
 CONFIG_SCSI_ZORRO7XX=y
 CONFIG_SCSI_ZORRO_ESP=y
+CONFIG_ATA=y
+# CONFIG_ATA_VERBOSE_ERROR is not set
+# CONFIG_ATA_BMDMA is not set
+CONFIG_PATA_GAYLE=y
+CONFIG_PATA_BUDDHA=y
 CONFIG_MD=y
 CONFIG_MD_LINEAR=m
 CONFIG_BLK_DEV_DM=m
diff --git a/arch/m68k/configs/atari_defconfig b/arch/m68k/configs/atari_defconfig
index 9cc9f1a06516..21b2990fe9af 100644
--- a/arch/m68k/configs/atari_defconfig
+++ b/arch/m68k/configs/atari_defconfig
@@ -324,10 +324,6 @@ CONFIG_BLK_DEV_RAM=y
 CONFIG_CDROM_PKTCDVD=m
 CONFIG_ATA_OVER_ETH=m
 CONFIG_DUMMY_IRQ=m
-CONFIG_IDE=y
-CONFIG_IDE_GD_ATAPI=y
-CONFIG_BLK_DEV_IDECD=y
-CONFIG_BLK_DEV_FALCON_IDE=y
 CONFIG_RAID_ATTRS=m
 CONFIG_SCSI=y
 CONFIG_BLK_DEV_SD=y
@@ -339,6 +335,10 @@ CONFIG_SCSI_SAS_ATTRS=m
 CONFIG_ISCSI_TCP=m
 CONFIG_ISCSI_BOOT_SYSFS=m
 CONFIG_ATARI_SCSI=y
+CONFIG_ATA=y
+# CONFIG_ATA_VERBOSE_ERROR is not set
+# CONFIG_ATA_BMDMA is not set
+CONFIG_PATA_FALCON=y
 CONFIG_MD=y
 CONFIG_MD_LINEAR=m
 CONFIG_BLK_DEV_DM=m
diff --git a/arch/m68k/configs/mac_defconfig b/arch/m68k/configs/mac_defconfig
index 406d3f2a16ea..b03300df13fc 100644
--- a/arch/m68k/configs/mac_defconfig
+++ b/arch/m68k/configs/mac_defconfig
@@ -315,10 +315,6 @@ CONFIG_BLK_DEV_RAM=y
 CONFIG_CDROM_PKTCDVD=m
 CONFIG_ATA_OVER_ETH=m
 CONFIG_DUMMY_IRQ=m
-CONFIG_IDE=y
-CONFIG_IDE_GD_ATAPI=y
-CONFIG_BLK_DEV_IDECD=y
-CONFIG_BLK_DEV_PLATFORM=y
 CONFIG_RAID_ATTRS=m
 CONFIG_SCSI=y
 CONFIG_BLK_DEV_SD=y
@@ -331,6 +327,10 @@ CONFIG_ISCSI_TCP=m
 CONFIG_ISCSI_BOOT_SYSFS=m
 CONFIG_MAC_SCSI=y
 CONFIG_SCSI_MAC_ESP=y
+CONFIG_ATA=y
+# CONFIG_ATA_VERBOSE_ERROR is not set
+# CONFIG_ATA_BMDMA is not set
+CONFIG_PATA_PLATFORM=y
 CONFIG_MD=y
 CONFIG_MD_LINEAR=m
 CONFIG_BLK_DEV_DM=m
diff --git a/arch/m68k/configs/multi_defconfig b/arch/m68k/configs/multi_defconfig
index f0992435e9ef..e2c8368e2231 100644
--- a/arch/m68k/configs/multi_defconfig
+++ b/arch/m68k/configs/multi_defconfig
@@ -344,13 +344,6 @@ CONFIG_BLK_DEV_RAM=y
 CONFIG_CDROM_PKTCDVD=m
 CONFIG_ATA_OVER_ETH=m
 CONFIG_DUMMY_IRQ=m
-CONFIG_IDE=y
-CONFIG_IDE_GD_ATAPI=y
-CONFIG_BLK_DEV_IDECD=y
-CONFIG_BLK_DEV_PLATFORM=y
-CONFIG_BLK_DEV_GAYLE=y
-CONFIG_BLK_DEV_BUDDHA=y
-CONFIG_BLK_DEV_FALCON_IDE=y
 CONFIG_RAID_ATTRS=m
 CONFIG_SCSI=y
 CONFIG_BLK_DEV_SD=y
@@ -374,6 +367,13 @@ CONFIG_MVME147_SCSI=y
 CONFIG_MVME16x_SCSI=y
 CONFIG_BVME6000_SCSI=y
 CONFIG_SUN3X_ESP=y
+CONFIG_ATA=y
+# CONFIG_ATA_VERBOSE_ERROR is not set
+# CONFIG_ATA_BMDMA is not set
+CONFIG_PATA_FALCON=y
+CONFIG_PATA_GAYLE=y
+CONFIG_PATA_BUDDHA=y
+CONFIG_PATA_PLATFORM=y
 CONFIG_MD=y
 CONFIG_MD_LINEAR=m
 CONFIG_BLK_DEV_DM=m
diff --git a/arch/m68k/configs/q40_defconfig b/arch/m68k/configs/q40_defconfig
index b893163d9f06..514e2e8cddbd 100644
--- a/arch/m68k/configs/q40_defconfig
+++ b/arch/m68k/configs/q40_defconfig
@@ -314,10 +314,6 @@ CONFIG_BLK_DEV_RAM=y
 CONFIG_CDROM_PKTCDVD=m
 CONFIG_ATA_OVER_ETH=m
 CONFIG_DUMMY_IRQ=m
-CONFIG_IDE=y
-CONFIG_IDE_GD_ATAPI=y
-CONFIG_BLK_DEV_IDECD=y
-CONFIG_BLK_DEV_FALCON_IDE=y
 CONFIG_RAID_ATTRS=m
 CONFIG_SCSI=y
 CONFIG_BLK_DEV_SD=y
@@ -328,6 +324,10 @@ CONFIG_SCSI_CONSTANTS=y
 CONFIG_SCSI_SAS_ATTRS=m
 CONFIG_ISCSI_TCP=m
 CONFIG_ISCSI_BOOT_SYSFS=m
+CONFIG_ATA=y
+# CONFIG_ATA_VERBOSE_ERROR is not set
+# CONFIG_ATA_BMDMA is not set
+CONFIG_PATA_FALCON=y
 CONFIG_MD=y
 CONFIG_MD_LINEAR=m
 CONFIG_BLK_DEV_DM=m
-- 
2.30.2

Re: [PATCH 5/6] m68k: use libata instead of the legacy ide driver

[Geert Uytterhoeven]

Hi Christoph,

On Wed, Jun 16, 2021 at 3:50 PM Christoph Hellwig <hch@lst.de> wrote:
> Switch the m68 defconfigs from the deprecated ide subsystem to use libata
> instead.  The gayle and buddha and falcon drivers are enabled for libata,
> while support for the q40 and macide drivers is lost.

I guess you forgot to update the last sentence, too?
These are now handled by the falcon and generic platform libata drivers.

> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>  arch/m68k/configs/amiga_defconfig | 10 +++++-----
>  arch/m68k/configs/atari_defconfig |  8 ++++----
>  arch/m68k/configs/mac_defconfig   |  8 ++++----
>  arch/m68k/configs/multi_defconfig | 14 +++++++-------
>  arch/m68k/configs/q40_defconfig   |  8 ++++----
>  5 files changed, 24 insertions(+), 24 deletions(-)

With the above fixed:
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH 5/6] m68k: use libata instead of the legacy ide driver

[Christoph Hellwig]

On Wed, Jun 16, 2021 at 04:03:38PM +0200, Geert Uytterhoeven wrote:
> Hi Christoph,
> 
> On Wed, Jun 16, 2021 at 3:50 PM Christoph Hellwig <hch@lst.de> wrote:
> > Switch the m68 defconfigs from the deprecated ide subsystem to use libata
> > instead.  The gayle and buddha and falcon drivers are enabled for libata,
> > while support for the q40 and macide drivers is lost.
> 
> I guess you forgot to update the last sentence, too?
> These are now handled by the falcon and generic platform libata drivers.

Indeed.

> With the above fixed:
> Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>

Thanks!

Re: remove the legacy ide driver v2

[John Paul Adrian Glaubitz]

Hi Christoph!

On 6/16/21 3:46 PM, Christoph Hellwig wrote:
> libata mostly covers all hardware supported by the legacy ide driver.
> There are three mips drivers that are not supported, but the linux-mips
> list could not identify any users of those.  There also are two m68k
> drivers that do not have libata equivalents, which might or might not
> have users, so we'll need some input and possibly help from the m68k
> community here.

I think the m68k drivers are being addressed at the moment. But Geert will
know more about the current status.

Adrian

-- 
 .''`.  John Paul Adrian Glaubitz
: :' :  Debian Developer - glaubitz@debian.org
`. `'   Freie Universitaet Berlin - glaubitz@physik.fu-berlin.de
  `-    GPG: 62FF 8A75 84E0 2956 9546  0006 7426 3B37 F5B5 F913

Re: remove the legacy ide driver v2

[Christoph Hellwig]

On Wed, Jun 16, 2021 at 03:51:07PM +0200, John Paul Adrian Glaubitz wrote:
> Hi Christoph!
> 
> On 6/16/21 3:46 PM, Christoph Hellwig wrote:
> > libata mostly covers all hardware supported by the legacy ide driver.
> > There are three mips drivers that are not supported, but the linux-mips
> > list could not identify any users of those.  There also are two m68k
> > drivers that do not have libata equivalents, which might or might not
> > have users, so we'll need some input and possibly help from the m68k
> > community here.
> 
> I think the m68k drivers are being addressed at the moment. But Geert will
> know more about the current status.

Actually they are, and the patches are in the libata tree.  I just forgot
to update the cover letter.

Re: remove the legacy ide driver v2

[Jens Axboe]

On 6/16/21 7:46 AM, Christoph Hellwig wrote:
> Hi all,
> 
> we've been trying to get rid of the legacy ide driver for a while now,
> and finally scheduled a removal for 2021, which is three month old now.
> 
> In general distros and most defconfigs have switched to libata long ago,
> but there are a few exceptions.  This series first switches over all
> remaining defconfigs to use libata and then removes the legacy ide
> driver.
> 
> libata mostly covers all hardware supported by the legacy ide driver.
> There are three mips drivers that are not supported, but the linux-mips
> list could not identify any users of those.  There also are two m68k
> drivers that do not have libata equivalents, which might or might not
> have users, so we'll need some input and possibly help from the m68k
> community here.
> 
> This series is against Jens' for-5.14/libata branch.

Unless someone complains loudly, I plan on queueing this up end this
week.

-- 
Jens Axboe

Re: remove the legacy ide driver v2

[Jens Axboe]

On 6/16/21 8:52 AM, Jens Axboe wrote:
> On 6/16/21 7:46 AM, Christoph Hellwig wrote:
>> Hi all,
>>
>> we've been trying to get rid of the legacy ide driver for a while now,
>> and finally scheduled a removal for 2021, which is three month old now.
>>
>> In general distros and most defconfigs have switched to libata long ago,
>> but there are a few exceptions.  This series first switches over all
>> remaining defconfigs to use libata and then removes the legacy ide
>> driver.
>>
>> libata mostly covers all hardware supported by the legacy ide driver.
>> There are three mips drivers that are not supported, but the linux-mips
>> list could not identify any users of those.  There also are two m68k
>> drivers that do not have libata equivalents, which might or might not
>> have users, so we'll need some input and possibly help from the m68k
>> community here.
>>
>> This series is against Jens' for-5.14/libata branch.
> 
> Unless someone complains loudly, I plan on queueing this up end this
> week.

This has now been applied for 5.14, thanks.

-- 
Jens Axboe