From: Archit Taneja <archit@ti.com> To: tomi.valkeinen@ti.com Cc: linux-fbdev@vger.kernel.org, linux-omap@vger.kernel.org, sumit.semwal@ti.com, rob@ti.com, Archit Taneja <archit@ti.com> Subject: [RFC 04/17] OMAPDSS: DPI: Add locking for DPI interface Date: Wed, 01 Aug 2012 10:43:15 +0000 [thread overview] Message-ID: <1343817088-29645-5-git-send-email-archit@ti.com> (raw) In-Reply-To: <1343817088-29645-1-git-send-email-archit@ti.com> The DPI interface driver currently relies on the panel driver to ensure calls like omapdss_dpi_display_enable() and omapdss_dpi_display_disable() are executed sequentially. Also, currently, there is no way to protect the DPI driver data. All DPI panel drivers don't ensure this, and in general, a DPI panel driver should use it's lock to that ensure it's own driver data and omap_dss_device states are taken care of, and not worry about the DPI interface. Add mutex locking in the DPI enable/disable/set_timings ops. Signed-off-by: Archit Taneja <archit@ti.com> --- drivers/video/omap2/dss/dpi.c | 26 ++++++++++++++++++++++++-- 1 file changed, 24 insertions(+), 2 deletions(-) diff --git a/drivers/video/omap2/dss/dpi.c b/drivers/video/omap2/dss/dpi.c index c20fe23..add47fe 100644 --- a/drivers/video/omap2/dss/dpi.c +++ b/drivers/video/omap2/dss/dpi.c @@ -39,6 +39,8 @@ static struct { struct regulator *vdds_dsi_reg; struct platform_device *dsidev; + struct mutex lock; + struct dss_lcd_mgr_config mgr_config; } dpi; @@ -184,14 +186,18 @@ int omapdss_dpi_display_enable(struct omap_dss_device *dssdev) { int r; + mutex_lock(&dpi.lock); + if (cpu_is_omap34xx() && !dpi.vdds_dsi_reg) { DSSERR("no VDSS_DSI regulator\n"); - return -ENODEV; + r = -ENODEV; + goto err_no_reg; } if (dssdev->manager = NULL) { DSSERR("failed to enable display: no manager\n"); - return -ENODEV; + r = -ENODEV; + goto err_no_mgr; } if (dpi_use_dsi_pll(dssdev)) { @@ -238,6 +244,8 @@ int omapdss_dpi_display_enable(struct omap_dss_device *dssdev) if (r) goto err_mgr_enable; + mutex_unlock(&dpi.lock); + return 0; err_mgr_enable: @@ -255,12 +263,17 @@ err_get_dispc: err_reg_enable: omap_dss_stop_device(dssdev); err_start_dev: +err_no_mgr: +err_no_reg: + mutex_unlock(&dpi.lock); return r; } EXPORT_SYMBOL(omapdss_dpi_display_enable); void omapdss_dpi_display_disable(struct omap_dss_device *dssdev) { + mutex_lock(&dpi.lock); + dss_mgr_disable(dssdev->manager); if (dpi_use_dsi_pll(dssdev)) { @@ -275,6 +288,8 @@ void omapdss_dpi_display_disable(struct omap_dss_device *dssdev) regulator_disable(dpi.vdds_dsi_reg); omap_dss_stop_device(dssdev); + + mutex_unlock(&dpi.lock); } EXPORT_SYMBOL(omapdss_dpi_display_disable); @@ -284,6 +299,9 @@ void dpi_set_timings(struct omap_dss_device *dssdev, int r; DSSDBG("dpi_set_timings\n"); + + mutex_lock(&dpi.lock); + dssdev->panel.timings = *timings; if (dssdev->state = OMAP_DSS_DISPLAY_ACTIVE) { r = dispc_runtime_get(); @@ -296,6 +314,8 @@ void dpi_set_timings(struct omap_dss_device *dssdev, } else { dss_mgr_set_timings(dssdev->manager, timings); } + + mutex_unlock(&dpi.lock); } EXPORT_SYMBOL(dpi_set_timings); @@ -392,6 +412,8 @@ static void __init dpi_probe_pdata(struct platform_device *pdev) static int __init omap_dpi_probe(struct platform_device *pdev) { + mutex_init(&dpi.lock); + dpi_probe_pdata(pdev); return 0; -- 1.7.9.5
WARNING: multiple messages have this Message-ID (diff)
From: Archit Taneja <archit@ti.com> To: tomi.valkeinen@ti.com Cc: linux-fbdev@vger.kernel.org, linux-omap@vger.kernel.org, sumit.semwal@ti.com, rob@ti.com, Archit Taneja <archit@ti.com> Subject: [RFC 04/17] OMAPDSS: DPI: Add locking for DPI interface Date: Wed, 1 Aug 2012 16:01:15 +0530 [thread overview] Message-ID: <1343817088-29645-5-git-send-email-archit@ti.com> (raw) In-Reply-To: <1343817088-29645-1-git-send-email-archit@ti.com> The DPI interface driver currently relies on the panel driver to ensure calls like omapdss_dpi_display_enable() and omapdss_dpi_display_disable() are executed sequentially. Also, currently, there is no way to protect the DPI driver data. All DPI panel drivers don't ensure this, and in general, a DPI panel driver should use it's lock to that ensure it's own driver data and omap_dss_device states are taken care of, and not worry about the DPI interface. Add mutex locking in the DPI enable/disable/set_timings ops. Signed-off-by: Archit Taneja <archit@ti.com> --- drivers/video/omap2/dss/dpi.c | 26 ++++++++++++++++++++++++-- 1 file changed, 24 insertions(+), 2 deletions(-) diff --git a/drivers/video/omap2/dss/dpi.c b/drivers/video/omap2/dss/dpi.c index c20fe23..add47fe 100644 --- a/drivers/video/omap2/dss/dpi.c +++ b/drivers/video/omap2/dss/dpi.c @@ -39,6 +39,8 @@ static struct { struct regulator *vdds_dsi_reg; struct platform_device *dsidev; + struct mutex lock; + struct dss_lcd_mgr_config mgr_config; } dpi; @@ -184,14 +186,18 @@ int omapdss_dpi_display_enable(struct omap_dss_device *dssdev) { int r; + mutex_lock(&dpi.lock); + if (cpu_is_omap34xx() && !dpi.vdds_dsi_reg) { DSSERR("no VDSS_DSI regulator\n"); - return -ENODEV; + r = -ENODEV; + goto err_no_reg; } if (dssdev->manager == NULL) { DSSERR("failed to enable display: no manager\n"); - return -ENODEV; + r = -ENODEV; + goto err_no_mgr; } if (dpi_use_dsi_pll(dssdev)) { @@ -238,6 +244,8 @@ int omapdss_dpi_display_enable(struct omap_dss_device *dssdev) if (r) goto err_mgr_enable; + mutex_unlock(&dpi.lock); + return 0; err_mgr_enable: @@ -255,12 +263,17 @@ err_get_dispc: err_reg_enable: omap_dss_stop_device(dssdev); err_start_dev: +err_no_mgr: +err_no_reg: + mutex_unlock(&dpi.lock); return r; } EXPORT_SYMBOL(omapdss_dpi_display_enable); void omapdss_dpi_display_disable(struct omap_dss_device *dssdev) { + mutex_lock(&dpi.lock); + dss_mgr_disable(dssdev->manager); if (dpi_use_dsi_pll(dssdev)) { @@ -275,6 +288,8 @@ void omapdss_dpi_display_disable(struct omap_dss_device *dssdev) regulator_disable(dpi.vdds_dsi_reg); omap_dss_stop_device(dssdev); + + mutex_unlock(&dpi.lock); } EXPORT_SYMBOL(omapdss_dpi_display_disable); @@ -284,6 +299,9 @@ void dpi_set_timings(struct omap_dss_device *dssdev, int r; DSSDBG("dpi_set_timings\n"); + + mutex_lock(&dpi.lock); + dssdev->panel.timings = *timings; if (dssdev->state == OMAP_DSS_DISPLAY_ACTIVE) { r = dispc_runtime_get(); @@ -296,6 +314,8 @@ void dpi_set_timings(struct omap_dss_device *dssdev, } else { dss_mgr_set_timings(dssdev->manager, timings); } + + mutex_unlock(&dpi.lock); } EXPORT_SYMBOL(dpi_set_timings); @@ -392,6 +412,8 @@ static void __init dpi_probe_pdata(struct platform_device *pdev) static int __init omap_dpi_probe(struct platform_device *pdev) { + mutex_init(&dpi.lock); + dpi_probe_pdata(pdev); return 0; -- 1.7.9.5
next prev parent reply other threads:[~2012-08-01 10:43 UTC|newest] Thread overview: 122+ messages / expand[flat|nested] mbox.gz Atom feed top 2012-08-01 10:31 [RFC 00/17] OMAPDSS: Change way of passing timings from panel driver to interface Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 01/17] OMAPDSS: APPLY: Constify timings argument in dss_mgr_set_timings Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 02/17] OMAPDSS: DPI: Remove omap_dss_device arguments in dpi_set_dsi_clk/dpi_set_dispc_clk Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 03/17] OMAPDSS: HDMI: Remove omap_dss_device argument from hdmi_compute_pll Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` Archit Taneja [this message] 2012-08-01 10:43 ` [RFC 04/17] OMAPDSS: DPI: Add locking for DPI interface Archit Taneja 2012-08-01 10:31 ` [RFC 05/17] OMAPDSS: DPI: Maintain our own timings field in driver data Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 06/17] OMAPDSS: DPI displays: Take care of panel timings in the driver itself Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 07/17] OMAPDSS: Displays: Add locking in generic DPI panel driver Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 08/17] OMAPDSS: DSI: Maintain own copy of timings in driver data Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-07 14:07 ` Tomi Valkeinen 2012-08-07 14:07 ` Tomi Valkeinen 2012-08-08 5:57 ` Archit Taneja 2012-08-08 6:09 ` Archit Taneja 2012-08-08 6:15 ` Tomi Valkeinen 2012-08-08 6:15 ` Tomi Valkeinen 2012-08-08 6:29 ` Archit Taneja 2012-08-08 6:41 ` Archit Taneja 2012-08-08 7:10 ` Tomi Valkeinen 2012-08-08 7:10 ` Tomi Valkeinen 2012-08-08 7:54 ` Archit Taneja 2012-08-08 8:06 ` Archit Taneja 2012-08-01 10:31 ` [RFC 09/17] OMAPDSS: HDMI: Use our own omap_video_timings field when setting interface timings Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 10/17] OMAPDSS: HDMI: Add a get_timing function for HDMI interface Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 11/17] OMAPDSS: HDMI: Add locking for hdmi interface get/set timing functions Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 12/17] OMAPDSS: SDI: Create a separate function for timing/clock configurations Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 13/17] OMAPDSS: SDI: Create a function to set timings Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-07 14:20 ` Tomi Valkeinen 2012-08-07 14:20 ` Tomi Valkeinen 2012-08-08 5:58 ` Archit Taneja 2012-08-08 6:10 ` Archit Taneja 2012-08-01 10:31 ` [RFC 14/17] OMAPDSS: SDI: Maintain our own timings field in driver data Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 15/17] OMAPDSS: VENC: Split VENC into interface and panel driver Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 16/17] OMAPDSS: VENC: Maintain our own timings field in driver data Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:31 ` [RFC 17/17] OMAPDSS: VENC: Add a get_timing function for VENC interface Archit Taneja 2012-08-01 10:43 ` Archit Taneja 2012-08-01 10:35 ` [RFC 00/17] OMAPDSS: Change way of passing timings from panel driver to interface Archit Taneja 2012-08-01 10:47 ` Archit Taneja 2012-08-07 14:32 ` Tomi Valkeinen 2012-08-07 14:32 ` Tomi Valkeinen 2012-08-08 6:05 ` Archit Taneja 2012-08-08 6:17 ` Archit Taneja 2012-08-08 6:25 ` Tomi Valkeinen 2012-08-08 6:25 ` Tomi Valkeinen 2012-08-08 6:47 ` Archit Taneja 2012-08-08 6:59 ` Archit Taneja 2012-08-08 7:27 ` Tomi Valkeinen 2012-08-08 7:27 ` Tomi Valkeinen 2012-08-08 7:59 ` Archit Taneja 2012-08-08 8:11 ` Archit Taneja 2012-08-08 8:13 ` Tomi Valkeinen 2012-08-08 8:13 ` Tomi Valkeinen 2012-08-08 8:38 ` Archit Taneja 2012-08-08 8:50 ` Archit Taneja 2012-08-08 8:48 ` Tomi Valkeinen 2012-08-08 8:48 ` Tomi Valkeinen 2012-08-08 9:24 ` Archit Taneja 2012-08-08 9:36 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 00/13] " Archit Taneja 2012-08-09 11:55 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 01/13] OMAPDSS: DPI: Maintain our own timings field in driver data Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 02/13] OMAPDSS: DPI displays: Take care of panel timings in the driver itself Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 03/13] OMAPDSS: DSI: Maintain own copy of timings in driver data Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 04/13] OMAPDSS: DSI: Add function to set panel size for command mode panels Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 05/13] OMAPDSS: DSI: Update manager timings on a manual update Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 06/13] OMAPDSS: HDMI: Use our own omap_video_timings field when setting interface timings Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 07/13] OMAPDSS: HDMI: Add a get_timing function for HDMI interface Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-14 13:02 ` Tomi Valkeinen 2012-08-14 13:02 ` Tomi Valkeinen 2012-08-14 13:15 ` Archit Taneja 2012-08-14 13:27 ` Archit Taneja 2012-08-14 14:10 ` Tomi Valkeinen 2012-08-14 14:10 ` Tomi Valkeinen 2012-08-14 17:16 ` Archit Taneja 2012-08-14 17:28 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 08/13] OMAPDSS: HDMI: Add locking for hdmi interface get/set timing functions Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 09/13] OMAPDSS: SDI: Create a function to set timings Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-14 13:44 ` Tomi Valkeinen 2012-08-14 13:44 ` Tomi Valkeinen 2012-08-14 16:56 ` Archit Taneja 2012-08-14 17:08 ` Archit Taneja 2012-08-14 17:33 ` Tomi Valkeinen 2012-08-14 17:33 ` Tomi Valkeinen 2012-08-14 19:08 ` Archit Taneja 2012-08-14 19:20 ` Archit Taneja 2012-08-15 6:43 ` Tomi Valkeinen 2012-08-15 6:43 ` Tomi Valkeinen 2012-08-14 19:26 ` Rob Clark 2012-08-14 19:26 ` Rob Clark 2012-08-09 11:49 ` [PATCH v2 10/13] OMAPDSS: SDI: Maintain our own timings field in driver data Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 11/13] OMAPDSS: VENC: Split VENC into interface and panel driver Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 12/13] OMAPDSS: VENC: Maintain our own timings field in driver data Archit Taneja 2012-08-09 11:56 ` Archit Taneja 2012-08-09 11:49 ` [PATCH v2 13/13] OMAPDSS: VENC: Add a get_timing function for VENC interface Archit Taneja 2012-08-09 11:56 ` Archit Taneja
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=1343817088-29645-5-git-send-email-archit@ti.com \ --to=archit@ti.com \ --cc=linux-fbdev@vger.kernel.org \ --cc=linux-omap@vger.kernel.org \ --cc=rob@ti.com \ --cc=sumit.semwal@ti.com \ --cc=tomi.valkeinen@ti.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.