[PATCH net-next v5 2/2] net/8390: apne.c - add 100 Mbit support to apne.c driver

Add Kconfig option, module parameter and PCMCIA reset code
required to support 100 Mbit PCMCIA ethernet cards on Amiga.

10 Mbit and 100 Mbit mode are supported by the same module.
A module parameter switches Amiga ISA IO accessors to word
access by changing isa_type at runtime. Additional code to
reset the PCMCIA hardware is also added to the driver probe.

Patch modified after patch "[PATCH RFC net-next] Amiga PCMCIA
100 MBit card support" submitted to netdev 2018/09/16 by Alex
Kazik [off-list ref].

CC: netdev@vger.kernel.org
Link: https://lore.kernel.org/r/1622958877-2026-1-git-send-email-schmitzmic@gmail.com
Tested-by: Alex Kazik <redacted>
Signed-off-by: Michael Schmitz <schmitzmic@gmail.com>

--
Changes from v4:

Geert Uytterhoeven:
- remove APNE100MBIT config option, always include 16 bit support
- change module parameter permissions
- try autoprobing for 16 bit mode early on in device probe

Changes from v3:

- change module parameter name to match Kconfig help

Finn Thain:
- fix coding style in new card reset code block
- allow reset of isa_type by module parameter

Changes from v1:

- fix module parameter name in Kconfig help text

Alex Kazik:
- change module parameter type to bool, fix module parameter
  permission

Changes from RFC:

Geert Uytterhoeven:
- change APNE_100MBIT to depend on APNE
- change '---help---' to 'help' (former no longer supported)
- fix whitespace errors
- fix module_param_named() arg count
- protect all added code by #ifdef CONFIG_APNE_100MBIT
---
 drivers/net/ethernet/8390/Kconfig |  4 ++++
 drivers/net/ethernet/8390/apne.c  | 33 +++++++++++++++++++++++++++++++++
 2 files changed, 37 insertions(+)

diff --git a/drivers/net/ethernet/8390/Kconfig b/drivers/net/ethernet/8390/Kconfig
index 9f4b302..cf0d8b3 100644
--- a/drivers/net/ethernet/8390/Kconfig
+++ b/drivers/net/ethernet/8390/Kconfig

@@ -143,6 +143,10 @@ config APNE
 	  To compile this driver as a module, choose M here: the module
 	  will be called apne.
 
+	  The driver also supports 10/100Mbit cards (e.g. Netgear FA411,
+	  CNet Singlepoint). To activate 100 Mbit support at runtime or
+	  from the kernel command line, use the apne.100mbit module parameter.
+
 config PCMCIA_PCNET
 	tristate "NE2000 compatible PCMCIA support"
 	depends on PCMCIA

diff --git a/drivers/net/ethernet/8390/apne.c b/drivers/net/ethernet/8390/apne.c
index fe6c834..8223e15 100644
--- a/drivers/net/ethernet/8390/apne.c
+++ b/drivers/net/ethernet/8390/apne.c

@@ -120,6 +120,10 @@ static u32 apne_msg_enable;
 module_param_named(msg_enable, apne_msg_enable, uint, 0444);
 MODULE_PARM_DESC(msg_enable, "Debug message level (see linux/netdevice.h for bitmap)");
 
+static bool apne_100_mbit;
+module_param_named(100_mbit, apne_100_mbit, bool, 0644);
+MODULE_PARM_DESC(100_mbit, "Enable 100 Mbit support");
+
 struct net_device * __init apne_probe(int unit)
 {
 	struct net_device *dev;

@@ -139,6 +143,11 @@ struct net_device * __init apne_probe(int unit)
 	if ( !(AMIGAHW_PRESENT(PCMCIA)) )
 		return ERR_PTR(-ENODEV);
 
+	if (apne_100_mbit)
+		isa_type = ISA_TYPE_AG16;
+	else
+		isa_type = ISA_TYPE_AG;
+
 	pr_info("Looking for PCMCIA ethernet card : ");
 
 	/* check if a card is inserted */

@@ -147,6 +156,20 @@ struct net_device * __init apne_probe(int unit)
 		return ERR_PTR(-ENODEV);
 	}
 
+	/* Reset card. Who knows what dain-bramaged state it was left in. */
+	{	unsigned long reset_start_time = jiffies;
+
+		outb(inb(IOBASE + NE_RESET), IOBASE + NE_RESET);
+
+		while ((inb(IOBASE + NE_EN0_ISR) & ENISR_RESET) == 0)
+			if (time_after(jiffies, reset_start_time + 2*HZ/100)) {
+				pr_info("Card not found (no reset ack).\n");
+				isa_type=ISA_TYPE_AG16;
+			}
+
+		outb(0xff, IOBASE + NE_EN0_ISR);		/* Ack all intr. */
+	}
+
 	dev = alloc_ei_netdev();
 	if (!dev)
 		return ERR_PTR(-ENOMEM);

@@ -590,6 +613,16 @@ static int init_pcmcia(void)
 #endif
 	u_long offset;
 
+	/* reset card (idea taken from CardReset by Artur Pogoda) */
+	if (isa_type == ISA_TYPE_AG16) {
+		u_char  tmp = gayle.intreq;
+
+		gayle.intreq = 0xff;
+		mdelay(1);
+		gayle.intreq = tmp;
+		mdelay(300);
+	}
+
 	pcmcia_reset();
 	pcmcia_program_voltage(PCMCIA_0V);
 	pcmcia_access_speed(PCMCIA_SPEED_250NS);

-- 
2.7.4

On Sat, 19 Jun 2021, Michael Schmitz wrote:

There's a missing line break here.

+
+		outb(inb(IOBASE + NE_RESET), IOBASE + NE_RESET);
+
+		while ((inb(IOBASE + NE_EN0_ISR) & ENISR_RESET) == 0)
+			if (time_after(jiffies, reset_start_time + 2*HZ/100)) {

You could use msecs_to_jiffies(20) here.

+				pr_info("Card not found (no reset ack).\n");
+				isa_type=ISA_TYPE_AG16;

Whitespace is needed around the '='.

+			}

Missing a break statement?

Extra whitespace.

+		gayle.intreq = 0xff;
+		mdelay(1);
+		gayle.intreq = tmp;
+		mdelay(300);
+	}
+
 	pcmcia_reset();
 	pcmcia_program_voltage(PCMCIA_0V);
 	pcmcia_access_speed(PCMCIA_SPEED_250NS);

Hi Finn,

thanks for reviewing again!

Am 19.06.2021 um 12:56 schrieb Finn Thain:

quoted

+	/* Reset card. Who knows what dain-bramaged state it was left in. */
+	{	unsigned long reset_start_time = jiffies;

There's a missing line break here.

Straight copy from apne_probe1() below, but you're right about style of 
course.

quoted

+
+		outb(inb(IOBASE + NE_RESET), IOBASE + NE_RESET);
+
+		while ((inb(IOBASE + NE_EN0_ISR) & ENISR_RESET) == 0)
+			if (time_after(jiffies, reset_start_time + 2*HZ/100)) {

You could use msecs_to_jiffies(20) here.

quoted

+				pr_info("Card not found (no reset ack).\n");
+				isa_type=ISA_TYPE_AG16;

Whitespace is needed around the '='.

quoted

+			}

Missing a break statement?

Ouch.

quoted

+
+		outb(0xff, IOBASE + NE_EN0_ISR);		/* Ack all intr. */
+	}
+
 	dev = alloc_ei_netdev();
 	if (!dev)
 		return ERR_PTR(-ENOMEM);

@@ -590,6 +613,16 @@ static int init_pcmcia(void)
 #endif
 	u_long offset;

+	/* reset card (idea taken from CardReset by Artur Pogoda) */
+	if (isa_type == ISA_TYPE_AG16) {
+		u_char  tmp = gayle.intreq;
+

Extra whitespace.

quoted

+		gayle.intreq = 0xff;
+		mdelay(1);
+		gayle.intreq = tmp;
+		mdelay(300);
+	}
+
 	pcmcia_reset();
 	pcmcia_program_voltage(PCMCIA_0V);
 	pcmcia_access_speed(PCMCIA_SPEED_250NS);

Thanks, will fix in v6 (and correct the commit message). Might split off 
the autoprobe bit so that can be easily dropped if need be.

Cheers,

	Michael

Hi Michael,

On Sat, Jun 19, 2021 at 2:35 AM Michael Schmitz [off-list ref] wrote:

Add Kconfig option, module parameter and PCMCIA reset code
required to support 100 Mbit PCMCIA ethernet cards on Amiga.

10 Mbit and 100 Mbit mode are supported by the same module.
A module parameter switches Amiga ISA IO accessors to word
access by changing isa_type at runtime. Additional code to
reset the PCMCIA hardware is also added to the driver probe.

Patch modified after patch "[PATCH RFC net-next] Amiga PCMCIA
100 MBit card support" submitted to netdev 2018/09/16 by Alex
Kazik [off-list ref].

CC: netdev@vger.kernel.org
Link: https://lore.kernel.org/r/1622958877-2026-1-git-send-email-schmitzmic@gmail.com
Tested-by: Alex Kazik <redacted>
Signed-off-by: Michael Schmitz <schmitzmic@gmail.com>

Thanks for your patch!

Note that this patch has a hard dependency on "[PATCH v5 1/2] m68k:
io_mm.h - add APNE 100 MBit support" in the series, so it must not
be applied to the netdev tree yet.

According to the recent discussion about that, "at runtime" is not
really possible?  So that limits it to kernel command line (for the
builtin case)
or module parameter (for the modular case).

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

Hi Geert,

thanks for your review!

Am 19.06.2021 um 21:08 schrieb Geert Uytterhoeven:

Hi Michael,

On Sat, Jun 19, 2021 at 2:35 AM Michael Schmitz [off-list ref] wrote:

quoted

Add Kconfig option, module parameter and PCMCIA reset code
required to support 100 Mbit PCMCIA ethernet cards on Amiga.

10 Mbit and 100 Mbit mode are supported by the same module.
A module parameter switches Amiga ISA IO accessors to word
access by changing isa_type at runtime. Additional code to
reset the PCMCIA hardware is also added to the driver probe.

Patch modified after patch "[PATCH RFC net-next] Amiga PCMCIA
100 MBit card support" submitted to netdev 2018/09/16 by Alex
Kazik [off-list ref].

CC: netdev@vger.kernel.org
Link: https://lore.kernel.org/r/1622958877-2026-1-git-send-email-schmitzmic@gmail.com
Tested-by: Alex Kazik <redacted>
Signed-off-by: Michael Schmitz <schmitzmic@gmail.com>

Thanks for your patch!

Note that this patch has a hard dependency on "[PATCH v5 1/2] m68k:
io_mm.h - add APNE 100 MBit support" in the series, so it must not
be applied to the netdev tree yet.

Hmm - so we ought to protect the new code by

#ifdef ARCH_HAVE_16BIT_PCMCIA

and set that in the m68k machine Kconfig in the first patch?

(It's almost, but not quite like a config option :-)

quoted

--- a/drivers/net/ethernet/8390/Kconfig
+++ b/drivers/net/ethernet/8390/Kconfig

@@ -143,6 +143,10 @@ config APNE
          To compile this driver as a module, choose M here: the module
          will be called apne.

+         The driver also supports 10/100Mbit cards (e.g. Netgear FA411,
+         CNet Singlepoint). To activate 100 Mbit support at runtime or
+         from the kernel command line, use the apne.100mbit module parameter.

According to the recent discussion about that, "at runtime" is not
really possible?  So that limits it to kernel command line (for the
builtin case)
or module parameter (for the modular case).

True - I'll reword that.

Cheers,

	Michael

quoted

+
 config PCMCIA_PCNET
        tristate "NE2000 compatible PCMCIA support"
        depends on PCMCIA

diff --git a/drivers/net/ethernet/8390/apne.c b/drivers/net/ethernet/8390/apne.c
index fe6c834..8223e15 100644
--- a/drivers/net/ethernet/8390/apne.c
+++ b/drivers/net/ethernet/8390/apne.c

@@ -120,6 +120,10 @@ static u32 apne_msg_enable;
 module_param_named(msg_enable, apne_msg_enable, uint, 0444);
 MODULE_PARM_DESC(msg_enable, "Debug message level (see linux/netdevice.h for bitmap)");

+static bool apne_100_mbit;
+module_param_named(100_mbit, apne_100_mbit, bool, 0644);
+MODULE_PARM_DESC(100_mbit, "Enable 100 Mbit support");

Gr{oetje,eeting}s,

                        Geert

Hi Michael,

On Sat, Jun 19, 2021 at 9:31 PM Michael Schmitz [off-list ref] wrote:

Am 19.06.2021 um 21:08 schrieb Geert Uytterhoeven:

quoted

On Sat, Jun 19, 2021 at 2:35 AM Michael Schmitz [off-list ref] wrote:

quoted

Add Kconfig option, module parameter and PCMCIA reset code
required to support 100 Mbit PCMCIA ethernet cards on Amiga.

10 Mbit and 100 Mbit mode are supported by the same module.
A module parameter switches Amiga ISA IO accessors to word
access by changing isa_type at runtime. Additional code to
reset the PCMCIA hardware is also added to the driver probe.

Patch modified after patch "[PATCH RFC net-next] Amiga PCMCIA
100 MBit card support" submitted to netdev 2018/09/16 by Alex
Kazik [off-list ref].

CC: netdev@vger.kernel.org
Link: https://lore.kernel.org/r/1622958877-2026-1-git-send-email-schmitzmic@gmail.com
Tested-by: Alex Kazik <redacted>
Signed-off-by: Michael Schmitz <schmitzmic@gmail.com>

Thanks for your patch!

Note that this patch has a hard dependency on "[PATCH v5 1/2] m68k:
io_mm.h - add APNE 100 MBit support" in the series, so it must not
be applied to the netdev tree yet.

Hmm - so we ought to protect the new code by

#ifdef ARCH_HAVE_16BIT_PCMCIA

and set that in the m68k machine Kconfig in the first patch?

(It's almost, but not quite like a config option :-)

No, we just manage dependencies, so either:
  1. Patch 2 cannot go in until patch 1 is upstream,
  2. One subsystem maintainer gives an Acked-by for one patch,so
    the other subsystem maintainer can apply both patches.

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

Hi Geert,

On 20/06/21 11:47 pm, Geert Uytterhoeven wrote:

quoted

quoted

Thanks for your patch!

Note that this patch has a hard dependency on "[PATCH v5 1/2] m68k:
io_mm.h - add APNE 100 MBit support" in the series, so it must not
be applied to the netdev tree yet.

Hmm - so we ought to protect the new code by

#ifdef ARCH_HAVE_16BIT_PCMCIA

and set that in the m68k machine Kconfig in the first patch?

(It's almost, but not quite like a config option :-)

No, we just manage dependencies, so either:
   1. Patch 2 cannot go in until patch 1 is upstream,
   2. One subsystem maintainer gives an Acked-by for one patch,so
     the other subsystem maintainer can apply both patches.

I haven't had any review from netdev yet - option 1 looks more feasible 
(if you want to carry a patch that's useless without its follow-up). And 
with the autoprobe failing (if I understood Alex right), I'll have to 
think of something else or drop that patch again.

Is there anything (aside from linking to the e-mail thread) that I can 
do to help netdev maintainers locate the first patch? The commit ID 
won't remain the same once accepted upstream, am I right?

Cheers,

     Michael

Gr{oetje,eeting}s,

                         Geert