This patch set provides several enhancements to mc13xxx MFD family
of devices by introducing mc34708 as a separate device.
This IC has dedicated pen detection feature, which allows better
touchscreen experience.
This is the fifth version of this code (v5).
Discussion regarding previous versions can be found here:
https://lkml.org/lkml/2018/4/12/351
https://www.mail-archive.com/[email protected]/msg1661934.html
https://www.mail-archive.com/[email protected]/msg1664296.html
https://lkml.org/lkml/2019/7/17/705
Sascha Hauer (3):
mfd: mc13xxx: Add mc34708 adc support
input: touchscreen mc13xxx: Make platform data optional
input: touchscreen mc13xxx: Add mc34708 support
drivers/input/touchscreen/mc13783_ts.c | 63 ++++++++++++++---
drivers/mfd/mc13xxx-core.c | 98 +++++++++++++++++++++++++-
include/linux/mfd/mc34708.h | 37 ++++++++++
3 files changed, 185 insertions(+), 13 deletions(-)
create mode 100644 include/linux/mfd/mc34708.h
--
2.20.1
From: Sascha Hauer <[email protected]>
The platform data once was optional, make it optional again. This
is a first step towards device tree support for the mc13xxx touchscreen
driver.
Signed-off-by: Sascha Hauer <[email protected]>
Signed-off-by: Lukasz Majewski <[email protected]>
Acked-by: Dmitry Torokhov <[email protected]>
---
Changes for v5:
- None
Changes for v4:
- Add Dmitry's Acked-by:
Changes for v3:
- None
Changes for v2:
- None
Changes from the original patch:
- Commit message typo fixes
---
drivers/input/touchscreen/mc13783_ts.c | 15 ++++++++-------
1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/drivers/input/touchscreen/mc13783_ts.c b/drivers/input/touchscreen/mc13783_ts.c
index ae0d978c83bf..edd49e44e0c9 100644
--- a/drivers/input/touchscreen/mc13783_ts.c
+++ b/drivers/input/touchscreen/mc13783_ts.c
@@ -35,7 +35,8 @@ struct mc13783_ts_priv {
struct mc13xxx *mc13xxx;
struct delayed_work work;
unsigned int sample[4];
- struct mc13xxx_ts_platform_data *touch;
+ u8 ato;
+ bool atox;
};
static irqreturn_t mc13783_ts_handler(int irq, void *data)
@@ -125,7 +126,7 @@ static void mc13783_ts_work(struct work_struct *work)
if (mc13xxx_adc_do_conversion(priv->mc13xxx,
mode, channel,
- priv->touch->ato, priv->touch->atox,
+ priv->ato, priv->atox,
priv->sample) == 0)
mc13783_ts_report_sample(priv);
}
@@ -169,6 +170,7 @@ static void mc13783_ts_close(struct input_dev *dev)
static int __init mc13783_ts_probe(struct platform_device *pdev)
{
struct mc13783_ts_priv *priv;
+ struct mc13xxx_ts_platform_data *pdata = dev_get_platdata(&pdev->dev);
struct input_dev *idev;
int ret = -ENOMEM;
@@ -180,11 +182,10 @@ static int __init mc13783_ts_probe(struct platform_device *pdev)
INIT_DELAYED_WORK(&priv->work, mc13783_ts_work);
priv->mc13xxx = dev_get_drvdata(pdev->dev.parent);
priv->idev = idev;
- priv->touch = dev_get_platdata(&pdev->dev);
- if (!priv->touch) {
- dev_err(&pdev->dev, "missing platform data\n");
- ret = -ENODEV;
- goto err_free_mem;
+
+ if (pdata) {
+ priv->atox = pdata->atox;
+ priv->ato = pdata->ato;
}
idev->name = MC13783_TS_NAME;
--
2.20.1
From: Sascha Hauer <[email protected]>
The mc34708 has an improved adc. The older variants will always convert
a fixed order of channels. The mc34708 can do up to eight conversions
in arbitrary channel order. Currently this extended feature is not
supported. We only support touchscreen conversions now, which will
be sampled in a data format compatible to the older chips in order
to keep the API between the mfd and the touchscreen driver.
Signed-off-by: Sascha Hauer <[email protected]>
Signed-off-by: Lukasz Majewski <[email protected]>
---
Changes for v5:
- Remove adc_do_conversion() callbacks from struct mc13xxx_variant
- Remove duplicated MC13XXX_ADC_WORKING #define
Changes for v4:
- None
Changes for v3:
- None
Changes for v2:
- Change the return code patch when the mc13xxx ADC is performing conversion
- Introduce new include/linux/mfd/mc34708.h header file for mc34708 specific
defines
Changes from the original patches:
- ADC conversion functions prototypes added to fix build error
- Adjustments to make checkpatch clean (-ENOSYS, line over 80 char)
This patch applies on top of Linux 5.3-rc8
SHA1: f74c2bb98776e2de508f4d607cd519873065118e
---
drivers/mfd/mc13xxx-core.c | 98 ++++++++++++++++++++++++++++++++++++-
include/linux/mfd/mc34708.h | 37 ++++++++++++++
2 files changed, 133 insertions(+), 2 deletions(-)
create mode 100644 include/linux/mfd/mc34708.h
diff --git a/drivers/mfd/mc13xxx-core.c b/drivers/mfd/mc13xxx-core.c
index 1abe7432aad8..b64c62366517 100644
--- a/drivers/mfd/mc13xxx-core.c
+++ b/drivers/mfd/mc13xxx-core.c
@@ -12,6 +12,7 @@
#include <linux/of_device.h>
#include <linux/platform_device.h>
#include <linux/mfd/core.h>
+#include <linux/mfd/mc34708.h>
#include "mc13xxx.h"
@@ -247,9 +248,9 @@ static irqreturn_t mc13xxx_handler_adcdone(int irq, void *data)
return IRQ_HANDLED;
}
-#define MC13XXX_ADC_WORKING (1 << 0)
+#define MC13XXX_ADC_WORKING BIT(0)
-int mc13xxx_adc_do_conversion(struct mc13xxx *mc13xxx, unsigned int mode,
+static int mc13xxx_adc_conversion(struct mc13xxx *mc13xxx, unsigned int mode,
unsigned int channel, u8 ato, bool atox,
unsigned int *sample)
{
@@ -358,6 +359,99 @@ int mc13xxx_adc_do_conversion(struct mc13xxx *mc13xxx, unsigned int mode,
return ret;
}
+
+static int mc34708_adc_conversion(struct mc13xxx *mc13xxx, unsigned int mode,
+ unsigned int channel, u8 ato, bool atox,
+ unsigned int *sample)
+{
+ int ret, i;
+ u32 adc0, adc3, adc1, old_adc0;
+ struct mc13xxx_adcdone_data adcdone_data = {
+ .mc13xxx = mc13xxx,
+ };
+
+ switch (mode) {
+ case MC13XXX_ADC_MODE_TS:
+ adc0 = MC34708_ADC0_TSEN | MC34708_ADC0_TSSTART |
+ MC34708_ADC0_TSSTOP(7);
+
+ adc1 = MC34708_ADC1_TSDLY1(0xf) |
+ MC34708_ADC1_TSDLY2(0xf) |
+ MC34708_ADC1_TSDLY3(0xf);
+
+ adc3 = MC34708_ADC3_TSSEL(0, MC34708_TS_X) |
+ MC34708_ADC3_TSSEL(1, MC34708_TS_Y) |
+ MC34708_ADC3_TSSEL(2, MC34708_TS_X) |
+ MC34708_ADC3_TSSEL(3, MC34708_TS_Y) |
+ MC34708_ADC3_TSSEL(4, MC34708_TS_X) |
+ MC34708_ADC3_TSSEL(5, MC34708_TS_R) |
+ MC34708_ADC3_TSSEL(6, MC34708_TS_Y) |
+ MC34708_ADC3_TSSEL(7, MC34708_TS_R);
+ break;
+
+ case MC13XXX_ADC_MODE_SINGLE_CHAN:
+ case MC13XXX_ADC_MODE_MULT_CHAN:
+ default:
+ return -EINVAL;
+ }
+
+ init_completion(&adcdone_data.done);
+
+ mc13xxx_lock(mc13xxx);
+
+ if (mc13xxx->adcflags & MC13XXX_ADC_WORKING) {
+ mc13xxx_unlock(mc13xxx);
+ return -EBUSY;
+ }
+
+ mc13xxx->adcflags |= MC13XXX_ADC_WORKING;
+
+ mc13xxx_reg_read(mc13xxx, MC13XXX_ADC0, &old_adc0);
+
+ mc13xxx_irq_request(mc13xxx, MC34708_IRQ_TSDONE,
+ mc13xxx_handler_adcdone, __func__, &adcdone_data);
+ mc13xxx_irq_ack(mc13xxx, MC34708_IRQ_TSDONE);
+
+ mc13xxx_reg_write(mc13xxx, MC34708_ADC3, adc3);
+ mc13xxx_reg_write(mc13xxx, MC13XXX_ADC1, adc1);
+ mc13xxx_reg_write(mc13xxx, MC13XXX_ADC0, adc0);
+
+ mc13xxx_unlock(mc13xxx);
+
+ ret = wait_for_completion_interruptible_timeout(&adcdone_data.done, HZ);
+
+ mc13xxx_lock(mc13xxx);
+
+ mc13xxx_irq_free(mc13xxx, MC34708_IRQ_TSDONE, &adcdone_data);
+
+ if (!ret) {
+ ret = -ETIMEDOUT;
+ goto out;
+ }
+
+ for (i = 0; i < 4; i++)
+ mc13xxx_reg_read(mc13xxx, MC34708_ADC4 + i, &sample[i]);
+
+out:
+ ret = mc13xxx_reg_write(mc13xxx, MC13XXX_ADC0, old_adc0);
+
+ mc13xxx->adcflags &= ~MC13XXX_ADC_WORKING;
+ mc13xxx_unlock(mc13xxx);
+
+ return ret;
+}
+
+int mc13xxx_adc_do_conversion(struct mc13xxx *mc13xxx, unsigned int mode,
+ unsigned int channel, u8 ato, bool atox,
+ unsigned int *sample)
+{
+ if (!strcmp(mc13xxx_get_chipname(mc13xxx), "mc34708"))
+ return mc34708_adc_conversion(mc13xxx, mode, channel, ato,
+ atox, sample);
+ else
+ return mc13xxx_adc_conversion(mc13xxx, mode, channel, ato,
+ atox, sample);
+}
EXPORT_SYMBOL_GPL(mc13xxx_adc_do_conversion);
static int mc13xxx_add_subdevice_pdata(struct mc13xxx *mc13xxx,
diff --git a/include/linux/mfd/mc34708.h b/include/linux/mfd/mc34708.h
new file mode 100644
index 000000000000..536e133f4f97
--- /dev/null
+++ b/include/linux/mfd/mc34708.h
@@ -0,0 +1,37 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2019
+ * Lukasz Majewski, DENX Software Engineering, [email protected]
+ */
+#ifndef __LINUX_MFD_MC34708_H
+#define __LINUX_MFD_MC34708_H
+
+#define MC34708_ADC3 46
+#define MC34708_ADC4 47
+
+#define MC34708_IRQ_TSDONE 1
+
+#define MC34708_ADC0_TSEN BIT(12)
+#define MC34708_ADC0_TSSTART BIT(13)
+#define MC34708_ADC0_TSCONT BIT(14)
+#define MC34708_ADC0_TSHOLD BIT(15)
+#define MC34708_ADC0_TSPENDETEN BIT(20)
+
+#define MC34708_ADC0_TSMASK (MC34708_ADC0_TSPENDETEN | \
+ MC34708_ADC0_TSEN | \
+ MC34708_ADC0_TSSTART | \
+ MC34708_ADC0_TSCONT | \
+ MC34708_ADC0_TSHOLD)
+
+#define MC34708_ADC0_TSSTOP(x) (((x) & 0x7) << 16)
+
+#define MC34708_ADC3_TSSEL(step, ch) ((ch) << (8 + 2 * (step)))
+#define MC34708_ADC1_TSDLY1(d) ((d) << 12)
+#define MC34708_ADC1_TSDLY2(d) ((d) << 16)
+#define MC34708_ADC1_TSDLY3(d) ((d) << 20)
+
+#define MC34708_TS_X 1
+#define MC34708_TS_Y 2
+#define MC34708_TS_R 3
+
+#endif /* __LINUX_MFD_MC34708_H */
--
2.20.1
Dear Lee,
> This patch set provides several enhancements to mc13xxx MFD family
> of devices by introducing mc34708 as a separate device.
>
> This IC has dedicated pen detection feature, which allows better
> touchscreen experience.
>
> This is the fifth version of this code (v5).
> Discussion regarding previous versions can be found here:
> https://lkml.org/lkml/2018/4/12/351
> https://www.mail-archive.com/[email protected]/msg1661934.html
> https://www.mail-archive.com/[email protected]/msg1664296.html
> https://lkml.org/lkml/2019/7/17/705
Gentle ping on this patch series. It is now 3 weeks without any reply...
>
> Sascha Hauer (3):
> mfd: mc13xxx: Add mc34708 adc support
> input: touchscreen mc13xxx: Make platform data optional
> input: touchscreen mc13xxx: Add mc34708 support
>
> drivers/input/touchscreen/mc13783_ts.c | 63 ++++++++++++++---
> drivers/mfd/mc13xxx-core.c | 98
> +++++++++++++++++++++++++- include/linux/mfd/mc34708.h |
> 37 ++++++++++ 3 files changed, 185 insertions(+), 13 deletions(-)
> create mode 100644 include/linux/mfd/mc34708.h
>
Best regards,
Lukasz Majewski
--
DENX Software Engineering GmbH, Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-59 Fax: (+49)-8142-66989-80 Email: [email protected]
On Mon, Sep 30, 2019 at 09:51:59AM +0200, Lukasz Majewski wrote:
> Dear Lee,
>
> > This patch set provides several enhancements to mc13xxx MFD family
> > of devices by introducing mc34708 as a separate device.
> >
> > This IC has dedicated pen detection feature, which allows better
> > touchscreen experience.
> >
> > This is the fifth version of this code (v5).
> > Discussion regarding previous versions can be found here:
> > https://lkml.org/lkml/2018/4/12/351
> > https://www.mail-archive.com/[email protected]/msg1661934.html
> > https://www.mail-archive.com/[email protected]/msg1664296.html
> > https://lkml.org/lkml/2019/7/17/705
>
> Gentle ping on this patch series. It is now 3 weeks without any reply...
It's the middle of the merge window, 2 weeks where we can't add any new
code to any of our trees. Please give people a chance...
greg k-h
On Mon, 30 Sep 2019, Lukasz Majewski wrote:
> Dear Lee,
>
> > This patch set provides several enhancements to mc13xxx MFD family
> > of devices by introducing mc34708 as a separate device.
> >
> > This IC has dedicated pen detection feature, which allows better
> > touchscreen experience.
> >
> > This is the fifth version of this code (v5).
> > Discussion regarding previous versions can be found here:
> > https://lkml.org/lkml/2018/4/12/351
> > https://www.mail-archive.com/[email protected]/msg1661934.html
> > https://www.mail-archive.com/[email protected]/msg1664296.html
> > https://lkml.org/lkml/2019/7/17/705
>
> Gentle ping on this patch series. It is now 3 weeks without any reply...
Please take note and follow the kernel release cycle.
These patches were sent after the release of -rc7 i.e. very late
in the release cycle and a point where most kernel maintainers stop
reviewing/applying patches and start to prepare for the impending
merge-window.
Also, there is no such thing as a gentle ping. If you genuinely think
your patches have unlikely("slipped though the gaps"), then post a
[RESEND] complete with a note alluding your reasons doing such.
> > Sascha Hauer (3):
> > mfd: mc13xxx: Add mc34708 adc support
> > input: touchscreen mc13xxx: Make platform data optional
> > input: touchscreen mc13xxx: Add mc34708 support
> >
> > drivers/input/touchscreen/mc13783_ts.c | 63 ++++++++++++++---
> > drivers/mfd/mc13xxx-core.c | 98
> > +++++++++++++++++++++++++- include/linux/mfd/mc34708.h |
> > 37 ++++++++++ 3 files changed, 185 insertions(+), 13 deletions(-)
> > create mode 100644 include/linux/mfd/mc34708.h
--
Lee Jones [李琼斯]
Linaro Services Technical Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
Hi Lee,
> On Mon, 30 Sep 2019, Lukasz Majewski wrote:
>
> > Dear Lee,
> >
> > > This patch set provides several enhancements to mc13xxx MFD family
> > > of devices by introducing mc34708 as a separate device.
> > >
> > > This IC has dedicated pen detection feature, which allows better
> > > touchscreen experience.
> > >
> > > This is the fifth version of this code (v5).
> > > Discussion regarding previous versions can be found here:
> > > https://lkml.org/lkml/2018/4/12/351
> > > https://www.mail-archive.com/[email protected]/msg1661934.html
> > > https://www.mail-archive.com/[email protected]/msg1664296.html
> > > https://lkml.org/lkml/2019/7/17/705
> >
> > Gentle ping on this patch series. It is now 3 weeks without any
> > reply...
>
> Please take note and follow the kernel release cycle.
>
> These patches were sent after the release of -rc7 i.e. very late
> in the release cycle and a point where most kernel maintainers stop
> reviewing/applying patches
Maybe something has changed recently, but I thought that the review is
an ongoing process (that the discussion about patches takes place all
the time to refine the code).
(and nobody expects maintainers pulling new patches after merge window).
> and start to prepare for the impending
> merge-window.
>
> Also, there is no such thing as a gentle ping.
I'm a bit puzzled now.
I do know that maintainers are overworked - and I do understand that.
However, NO reply for such a long time is at best confusing for
somebody who is willing to fix the kernel.
(Maybe this has changed too - but I do remember that there was a "rule
of thumb" to express any comment about the patch in 2 weeks time).
> If you genuinely think
> your patches have unlikely("slipped though the gaps"), then post a
> [RESEND] complete with a note alluding your reasons doing such.
>
Please forgive me, but are patches from this series eligible for pulling
(to v5.4 or even next release)?
> > > Sascha Hauer (3):
> > > mfd: mc13xxx: Add mc34708 adc support
> > > input: touchscreen mc13xxx: Make platform data optional
> > > input: touchscreen mc13xxx: Add mc34708 support
> > >
> > > drivers/input/touchscreen/mc13783_ts.c | 63 ++++++++++++++---
> > > drivers/mfd/mc13xxx-core.c | 98
> > > +++++++++++++++++++++++++- include/linux/mfd/mc34708.h
> > > | 37 ++++++++++ 3 files changed, 185 insertions(+), 13
> > > deletions(-) create mode 100644 include/linux/mfd/mc34708.h
>
Best regards,
Lukasz Majewski
--
DENX Software Engineering GmbH, Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-59 Fax: (+49)-8142-66989-80 Email: [email protected]
On Mon, 09 Sep 2019, Lukasz Majewski wrote:
> From: Sascha Hauer <[email protected]>
>
> The mc34708 has an improved adc. The older variants will always convert
> a fixed order of channels. The mc34708 can do up to eight conversions
> in arbitrary channel order. Currently this extended feature is not
> supported. We only support touchscreen conversions now, which will
> be sampled in a data format compatible to the older chips in order
> to keep the API between the mfd and the touchscreen driver.
Please take this opportunity to split all the ADC stuff out into an
ADC specific driver.
> Signed-off-by: Sascha Hauer <[email protected]>
> Signed-off-by: Lukasz Majewski <[email protected]>
>
> ---
> Changes for v5:
> - Remove adc_do_conversion() callbacks from struct mc13xxx_variant
> - Remove duplicated MC13XXX_ADC_WORKING #define
>
> Changes for v4:
> - None
>
> Changes for v3:
> - None
>
> Changes for v2:
> - Change the return code patch when the mc13xxx ADC is performing conversion
> - Introduce new include/linux/mfd/mc34708.h header file for mc34708 specific
> defines
>
> Changes from the original patches:
> - ADC conversion functions prototypes added to fix build error
> - Adjustments to make checkpatch clean (-ENOSYS, line over 80 char)
>
> This patch applies on top of Linux 5.3-rc8
> SHA1: f74c2bb98776e2de508f4d607cd519873065118e
> ---
> drivers/mfd/mc13xxx-core.c | 98 ++++++++++++++++++++++++++++++++++++-
> include/linux/mfd/mc34708.h | 37 ++++++++++++++
> 2 files changed, 133 insertions(+), 2 deletions(-)
> create mode 100644 include/linux/mfd/mc34708.h
--
Lee Jones [李琼斯]
Linaro Services Technical Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
On Tue, 01 Oct 2019, Lukasz Majewski wrote:
> > On Mon, 30 Sep 2019, Lukasz Majewski wrote:
> >
> > > Dear Lee,
> > >
> > > > This patch set provides several enhancements to mc13xxx MFD family
> > > > of devices by introducing mc34708 as a separate device.
> > > >
> > > > This IC has dedicated pen detection feature, which allows better
> > > > touchscreen experience.
> > > >
> > > > This is the fifth version of this code (v5).
> > > > Discussion regarding previous versions can be found here:
> > > > https://lkml.org/lkml/2018/4/12/351
> > > > https://www.mail-archive.com/[email protected]/msg1661934.html
> > > > https://www.mail-archive.com/[email protected]/msg1664296.html
> > > > https://lkml.org/lkml/2019/7/17/705
> > >
> > > Gentle ping on this patch series. It is now 3 weeks without any
> > > reply...
> >
> > Please take note and follow the kernel release cycle.
> >
> > These patches were sent after the release of -rc7 i.e. very late
> > in the release cycle and a point where most kernel maintainers stop
> > reviewing/applying patches
>
> Maybe something has changed recently, but I thought that the review is
> an ongoing process (that the discussion about patches takes place all
> the time to refine the code).
>
> (and nobody expects maintainers pulling new patches after merge window).
Nothing changed recently. Some maintainers will review just before
and during the merge windows, others use the time to prepare their
branches for submission and welcome the break from reviewing during
this time. Either way, please do not *expect* a fast turn-around
during in this period of the cycle.
> > and start to prepare for the impending
> > merge-window.
> >
> > Also, there is no such thing as a gentle ping.
>
> I'm a bit puzzled now.
> I do know that maintainers are overworked - and I do understand that.
>
> However, NO reply for such a long time is at best confusing for
> somebody who is willing to fix the kernel.
If you respect the merge cycle, you will come to expect a short delay
at the end and during the merge-window. Please also understand that
Maintainers also take vacation, attend conferences and have other work
to attend to.
> (Maybe this has changed too - but I do remember that there was a "rule
> of thumb" to express any comment about the patch in 2 weeks time).
Same rule of thumb applies. If after 2 weeks you have not received a
review, feel free to send a [RESEND]. However, bumping threads with
contentless pings is considered bad form.
As an aside, I (and others) conduct reviews in batches and in
chronological order (first come, first served) of how the mails are
represented in my inbox. Thus, if you bump a thread, it goes to the
back of the queue.
> > If you genuinely think
> > your patches have unlikely("slipped though the gaps"), then post a
> > [RESEND] complete with a note alluding your reasons doing such.
> >
>
> Please forgive me, but are patches from this series eligible for pulling
> (to v5.4 or even next release)?
v5.4-rc1 has already been released. No new functionality will hit
v5.4. We have around 6-7 weeks to agree on acceptance for this set
for them to be eligible for v5.5.
> > > > Sascha Hauer (3):
> > > > mfd: mc13xxx: Add mc34708 adc support
> > > > input: touchscreen mc13xxx: Make platform data optional
> > > > input: touchscreen mc13xxx: Add mc34708 support
> > > >
> > > > drivers/input/touchscreen/mc13783_ts.c | 63 ++++++++++++++---
> > > > drivers/mfd/mc13xxx-core.c | 98
> > > > +++++++++++++++++++++++++- include/linux/mfd/mc34708.h
> > > > | 37 ++++++++++ 3 files changed, 185 insertions(+), 13
> > > > deletions(-) create mode 100644 include/linux/mfd/mc34708.h
--
Lee Jones [李琼斯]
Linaro Services Technical Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog