| Message ID | 20140119195843.7ae9753d@armhf (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Sun, Jan 19, 2014 at 07:58:43PM +0100, Jean-Francois Moine wrote: > This patch adds the optional treatment of the tda998x IRQ. > > The interrupt function is used to know the display connection status > without polling and to speedup reading the EDID. > > The interrupt number may be defined either in the DT or at encoder set > config time for non-DT boards. > > Signed-off-by: Jean-Francois Moine <moinejf@free.fr> > --- > v3 > - remarks from Russell King > - move the setting of the wq_edid_wait flag before asking the > device to read the EDID > Thanks about this bug fix, but I don't see the other problem: > there is no reason for the read_edid_block function to be > called more than once... It will happen whenever DRM asks the connector for the modes. However, since the read process is triggered each time, I think this is fine. > - use '0' as 'no irq' > - remove the useless delay after irq init Some further comments... > @@ -720,6 +787,10 @@ tda998x_encoder_set_config(struct drm_encoder *encoder, void *params) > priv->audio_port = p->audio_cfg; > priv->audio_format = p->audio_format; > } > + > + priv->irq = p->irq; > + if (p->irq) > + tda_irq_init(priv); If we're going to do it this way, this should probably release the IRQ if there was one before re-claiming it, just in case this function gets called more than once by some driver using it. The alternative is, as I said before, to use the infrastructure which is already there, namely setting the interrupt via struct i2c_client's irq member. Yes, that doesn't satisfy Sebastian's comment about using a GPIO, but there's no sign of GPIO usage in here at the moment anyway. So we might as well use what's already provided. That would then avoid the need to deal with IRQ changes etc in tda998x_encoder_set_config(). In any case, I've tested this patch in both non-IRQ and IRQ modes, and it appears to work - but I'd like the suggestion above before I provide any attributations. Thanks.
On 01/22/14 23:27, Russell King - ARM Linux wrote: > On Sun, Jan 19, 2014 at 07:58:43PM +0100, Jean-Francois Moine wrote: >> This patch adds the optional treatment of the tda998x IRQ. >> >> The interrupt function is used to know the display connection status >> without polling and to speedup reading the EDID. >> >> The interrupt number may be defined either in the DT or at encoder set >> config time for non-DT boards. >> >> Signed-off-by: Jean-Francois Moine <moinejf@free.fr> >> --- [...] >> @@ -720,6 +787,10 @@ tda998x_encoder_set_config(struct drm_encoder *encoder, void *params) >> priv->audio_port = p->audio_cfg; >> priv->audio_format = p->audio_format; >> } >> + >> + priv->irq = p->irq; >> + if (p->irq) >> + tda_irq_init(priv); > > If we're going to do it this way, this should probably release the IRQ if > there was one before re-claiming it, just in case this function gets called > more than once by some driver using it. > > The alternative is, as I said before, to use the infrastructure which is > already there, namely setting the interrupt via struct i2c_client's > irq member. Yes, that doesn't satisfy Sebastian's comment about using > a GPIO, but there's no sign of GPIO usage in here at the moment anyway. > So we might as well use what's already provided. Russell, I am fine with using an irq instead of gpio here. I remember you telling me on a similar patch, that from the gpio you can derive the irq but not the other way round. Anyway, I also remember reading discussions about DT gpios vs interrupts, and IIRC the outcome was that passing interrupts is fine, too. We usually have both interrupt-controller; and gpio-controller; set on DT gpio controllers, so let's stick with irq. And: Thanks for reviewing this again, I am still too busy to keep up with it. Sebastian
On Fri, Jan 24, 2014 at 06:29:12PM +0100, Sebastian Hesselbarth wrote: > On 01/22/14 23:27, Russell King - ARM Linux wrote: >> On Sun, Jan 19, 2014 at 07:58:43PM +0100, Jean-Francois Moine wrote: >>> This patch adds the optional treatment of the tda998x IRQ. >>> >>> The interrupt function is used to know the display connection status >>> without polling and to speedup reading the EDID. >>> >>> The interrupt number may be defined either in the DT or at encoder set >>> config time for non-DT boards. >>> >>> Signed-off-by: Jean-Francois Moine <moinejf@free.fr> >>> --- > [...] >>> @@ -720,6 +787,10 @@ tda998x_encoder_set_config(struct drm_encoder *encoder, void *params) >>> priv->audio_port = p->audio_cfg; >>> priv->audio_format = p->audio_format; >>> } >>> + >>> + priv->irq = p->irq; >>> + if (p->irq) >>> + tda_irq_init(priv); >> >> If we're going to do it this way, this should probably release the IRQ if >> there was one before re-claiming it, just in case this function gets called >> more than once by some driver using it. >> >> The alternative is, as I said before, to use the infrastructure which is >> already there, namely setting the interrupt via struct i2c_client's >> irq member. Yes, that doesn't satisfy Sebastian's comment about using >> a GPIO, but there's no sign of GPIO usage in here at the moment anyway. >> So we might as well use what's already provided. > > Russell, > > I am fine with using an irq instead of gpio here. I remember you telling > me on a similar patch, that from the gpio you can derive the irq but > not the other way round. Anyway, I also remember reading discussions > about DT gpios vs interrupts, and IIRC the outcome was that passing > interrupts is fine, too. Sebastian, You can derive the irq from a gpio (using gpio_to_irq()), but you can't derive a gpio from an irq. It's annoying that i2c_client and the i2c board data do not allow a GPIO to be specified, but only an IRQ, but that's a separate problem which when solved will also get solved here. So, I'm not too worried about the "should this be an IRQ or should it be a GPIO" question here. For us on ARM, it's less of a problem because we can just deal with GPIOs or IRQs at the DT level, and so we don't care what's in i2c_client.
diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c index 98142db..013a67c 100644 --- a/drivers/gpu/drm/i2c/tda998x_drv.c +++ b/drivers/gpu/drm/i2c/tda998x_drv.c @@ -19,6 +19,7 @@ #include <linux/hdmi.h> #include <linux/module.h> +#include <linux/of_irq.h> #include <drm/drmP.h> #include <drm/drm_crtc_helper.h> @@ -44,6 +45,11 @@ struct tda998x_priv { u8 audio_format; u8 audio_frame[6]; u32 audio_port; + + int irq; + wait_queue_head_t wq_edid; + volatile int wq_edid_wait; + struct drm_encoder *encoder; }; #define to_tda998x_priv(x) ((struct tda998x_priv *)to_encoder_slave(x)->slave_priv) @@ -310,11 +316,16 @@ struct tda998x_priv { /* CEC registers: (not paged) */ +#define REG_CEC_INTSTATUS 0xee /* read */ +# define CEC_INTSTATUS_CEC (1 << 0) +# define CEC_INTSTATUS_HDMI (1 << 1) #define REG_CEC_FRO_IM_CLK_CTRL 0xfb /* read/write */ # define CEC_FRO_IM_CLK_CTRL_GHOST_DIS (1 << 7) # define CEC_FRO_IM_CLK_CTRL_ENA_OTP (1 << 6) # define CEC_FRO_IM_CLK_CTRL_IMCLK_SEL (1 << 1) # define CEC_FRO_IM_CLK_CTRL_FRO_DIV (1 << 0) +#define REG_CEC_RXSHPDINTENA 0xfc /* read/write */ +#define REG_CEC_RXSHPDINT 0xfd /* read */ #define REG_CEC_RXSHPDLEV 0xfe /* read */ # define CEC_RXSHPDLEV_RXSENS (1 << 0) # define CEC_RXSHPDLEV_HPD (1 << 1) @@ -528,6 +539,62 @@ tda998x_reset(struct tda998x_priv *priv) reg_write(priv, REG_MUX_VP_VIP_OUT, 0x24); } +/* + * only 2 interrupts may occur: screen plug/unplug and EDID read + */ +static irqreturn_t tda998x_irq_thread(int irq, void *data) +{ + struct tda998x_priv *priv = data; + u8 sta, cec, lvl, flag0, flag1, flag2; + + if (!priv) + return IRQ_HANDLED; + sta = cec_read(priv, REG_CEC_INTSTATUS); + cec = cec_read(priv, REG_CEC_RXSHPDINT); + lvl = cec_read(priv, REG_CEC_RXSHPDLEV); + flag0 = reg_read(priv, REG_INT_FLAGS_0); + flag1 = reg_read(priv, REG_INT_FLAGS_1); + flag2 = reg_read(priv, REG_INT_FLAGS_2); + DRM_DEBUG_DRIVER( + "tda irq sta %02x cec %02x lvl %02x f0 %02x f1 %02x f2 %02x\n", + sta, cec, lvl, flag0, flag1, flag2); + if ((flag2 & INT_FLAGS_2_EDID_BLK_RD) && priv->wq_edid_wait) { + priv->wq_edid_wait = 0; + wake_up(&priv->wq_edid); + } else if (cec != 0) { /* level change */ + if (priv->encoder && priv->encoder->dev) + drm_helper_hpd_irq_event(priv->encoder->dev); + } + return IRQ_HANDLED; +} + +static void tda_irq_init(struct tda998x_priv *priv) +{ + int ret; + + /* init read EDID waitqueue */ + init_waitqueue_head(&priv->wq_edid); + + /* clear pending interrupts */ + reg_read(priv, REG_INT_FLAGS_0); + reg_read(priv, REG_INT_FLAGS_1); + reg_read(priv, REG_INT_FLAGS_2); + + ret = request_threaded_irq(priv->irq, NULL, + tda998x_irq_thread, + IRQF_TRIGGER_FALLING | IRQF_ONESHOT, + "tda998x", priv); + if (ret) { + dev_err(&priv->hdmi->dev, "failed to request IRQ#%u: %d\n", + priv->irq, ret); + return; + } + + /* enable HPD irq */ + cec_write(priv, REG_CEC_RXSHPDINTENA, + CEC_RXSHPDLEV_HPD | CEC_RXSHPDLEV_RXSENS); +} + static uint8_t tda998x_cksum(uint8_t *buf, size_t bytes) { uint8_t sum = 0; @@ -720,6 +787,10 @@ tda998x_encoder_set_config(struct drm_encoder *encoder, void *params) priv->audio_port = p->audio_cfg; priv->audio_format = p->audio_format; } + + priv->irq = p->irq; + if (p->irq) + tda_irq_init(priv); } static void @@ -990,9 +1061,6 @@ read_edid_block(struct drm_encoder *encoder, uint8_t *buf, int blk) uint8_t offset, segptr; int ret, i; - /* enable EDID read irq: */ - reg_set(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD); - offset = (blk & 1) ? 128 : 0; segptr = blk / 2; @@ -1002,23 +1070,36 @@ read_edid_block(struct drm_encoder *encoder, uint8_t *buf, int blk) reg_write(priv, REG_DDC_SEGM, segptr); /* enable reading EDID: */ + priv->wq_edid_wait = 1; reg_write(priv, REG_EDID_CTRL, 0x1); /* flag must be cleared by sw: */ reg_write(priv, REG_EDID_CTRL, 0x0); /* wait for block read to complete: */ - for (i = 100; i > 0; i--) { - ret = reg_read(priv, REG_INT_FLAGS_2); - if (ret < 0) - return ret; - if (ret & INT_FLAGS_2_EDID_BLK_RD) - break; - msleep(1); + if (priv->irq != 0) { + i = wait_event_timeout(priv->wq_edid, + !priv->wq_edid_wait, + msecs_to_jiffies(100)); + if (i < 0) { + dev_err(encoder->dev->dev, "read edid wait err %d\n", i); + return i; + } + } else { + for (i = 10; i > 0; i--) { + msleep(10); + ret = reg_read(priv, REG_INT_FLAGS_2); + if (ret < 0) + return ret; + if (ret & INT_FLAGS_2_EDID_BLK_RD) + break; + } } - if (i == 0) + if (i == 0) { + dev_err(encoder->dev->dev, "read edid timeout\n"); return -ETIMEDOUT; + } ret = reg_read_range(priv, REG_EDID_DATA_0, buf, EDID_LENGTH); if (ret != EDID_LENGTH) { @@ -1027,8 +1108,6 @@ read_edid_block(struct drm_encoder *encoder, uint8_t *buf, int blk) return ret; } - reg_clear(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD); - return 0; } @@ -1046,6 +1125,9 @@ do_get_edid(struct drm_encoder *encoder) if (priv->rev == TDA19988) reg_clear(priv, REG_TX4, TX4_PD_RAM); + /* enable EDID read irq: */ + reg_set(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD); + /* base block fetch */ if (read_edid_block(encoder, block, 0)) goto fail; @@ -1118,7 +1200,13 @@ static int tda998x_encoder_create_resources(struct drm_encoder *encoder, struct drm_connector *connector) { - DBG(""); + struct tda998x_priv *priv = to_tda998x_priv(encoder); + + if (priv->irq != 0) + connector->polled = DRM_CONNECTOR_POLL_HPD; + else + connector->polled = DRM_CONNECTOR_POLL_CONNECT | + DRM_CONNECTOR_POLL_DISCONNECT; return 0; } @@ -1137,6 +1225,8 @@ tda998x_encoder_destroy(struct drm_encoder *encoder) { struct tda998x_priv *priv = to_tda998x_priv(encoder); drm_i2c_encoder_destroy(encoder); + if (priv->irq != 0) + free_irq(priv->irq, priv); if (priv->cec) i2c_unregister_device(priv->cec); kfree(priv); @@ -1195,6 +1285,7 @@ tda998x_encoder_init(struct i2c_client *client, priv->cec = i2c_new_dummy(client->adapter, 0x34); if (!priv->cec) return -ENODEV; + priv->encoder = &encoder_slave->base; priv->dpms = DRM_MODE_DPMS_OFF; encoder_slave->slave_priv = priv; @@ -1259,6 +1350,13 @@ tda998x_encoder_init(struct i2c_client *client, priv->vip_cntrl_2 = video; } + /* install the optional HDMI connect IRQ */ + priv->irq = irq_of_parse_and_map(np, 0); + if (priv->irq < 0) + priv->irq = 0; + if (priv->irq != 0) + tda_irq_init(priv); + return 0; fail: diff --git a/include/drm/i2c/tda998x.h b/include/drm/i2c/tda998x.h index d469b07..8dc2982 100644 --- a/include/drm/i2c/tda998x.h +++ b/include/drm/i2c/tda998x.h @@ -25,6 +25,8 @@ struct tda998x_encoder_params { } audio_format; unsigned audio_sample_rate; + + int irq; }; #endif
This patch adds the optional treatment of the tda998x IRQ. The interrupt function is used to know the display connection status without polling and to speedup reading the EDID. The interrupt number may be defined either in the DT or at encoder set config time for non-DT boards. Signed-off-by: Jean-Francois Moine <moinejf@free.fr> --- v3 - remarks from Russell King - move the setting of the wq_edid_wait flag before asking the device to read the EDID Thanks about this bug fix, but I don't see the other problem: there is no reason for the read_edid_block function to be called more than once... - use '0' as 'no irq' - remove the useless delay after irq init --- drivers/gpu/drm/i2c/tda998x_drv.c | 126 +++++++++++++++++++++++++++++++++----- include/drm/i2c/tda998x.h | 2 + 2 files changed, 114 insertions(+), 14 deletions(-)