dri-devel October 2013

dri-devel@lists.freedesktop.org

171 participants
926 discussions

[patch] drm/radeon/dpm: cleanup a type issue with rv6xx_clocks_per_unit()
by Dan Carpenter 02 Oct '13

02 Oct '13

The rv6xx_clocks_per_unit() function pretends it can set flags in a u64 bitfield but really because "1" is an int it doesn't work for more than 32 bits. The only caller truncates the high bits away anyway. I've just changed it to be a u32. Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> diff --git a/drivers/gpu/drm/radeon/rv6xx_dpm.c b/drivers/gpu/drm/radeon/rv6xx_dpm.c index 0e8b7d9..a4b80f23 100644 --- a/drivers/gpu/drm/radeon/rv6xx_dpm.c +++ b/drivers/gpu/drm/radeon/… [View More]

2 1

[patch] drm/radeon: forever loop on error in radeon_do_test_moves()
by Dan Carpenter 02 Oct '13

02 Oct '13

The error path does this: for (--i; i >= 0; --i) { which is a forever loop because "i" is unsigned. Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c index f4d6bce..12e8099 100644 --- a/drivers/gpu/drm/radeon/radeon_test.c +++ b/drivers/gpu/drm/radeon/radeon_test.c @@ -36,8 +36,8 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag) struct radeon_bo *vram_obj = … [View More]

2 1

[PATCH] drm/radeon: signal all fences after lockup to avoid endless waiting in GEM_WAIT
by Marek Olšák 02 Oct '13

02 Oct '13

From: Marek Olšák <marek.olsak(a)amd.com> After a lockup, fences are not signalled sometimes, causing the GEM_WAIT_IDLE ioctl to never return, which sometimes results in an X server freeze. This fixes only one of many deadlocks which can occur during a lockup. Signed-off-by: Marek Olšák <marek.olsak(a)amd.com> --- drivers/gpu/drm/radeon/radeon_device.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/… [View More]

1 0

[Bug 70035] GPU Lockup on AMD RS880 HD4200
by bugzilla-daemon＠freedesktop.org 02 Oct '13

02 Oct '13

1 0

[PATCH] drm: Prevent use of uninitialised values whilst loading edid firmware
by Chris Wilson 02 Oct '13

02 Oct '13

CC drivers/gpu/drm/drm_edid_load.o drivers/gpu/drm/drm_edid_load.c: In function ‘drm_load_edid_firmware’: include/linux/err.h:39:17: warning: ‘edid’ may be used uninitialised in this function [-Wuninitialized] drivers/gpu/drm/drm_edid_load.c:141:22: note: ‘edid’ was declared here In the process, we can make the error handling more resilient. Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> --- drivers/gpu/drm/drm_edid_load.c | 75 +++++++++++++++++---------------------- … [View More]1 file changed, 32 insertions(+), 43 deletions(-) diff --git a/drivers/gpu/drm/drm_edid_load.c b/drivers/gpu/drm/drm_edid_load.c index 271b42b..4b57a4c 100644 --- a/drivers/gpu/drm/drm_edid_load.c +++ b/drivers/gpu/drm/drm_edid_load.c @@ -136,59 +136,51 @@ static u8 generic_edid[GENERIC_EDIDS][128] = { static u8 *edid_load(struct drm_connector *connector, const char *name, const char *connector_name) { - const struct firmware *fw; + const struct firmware *fw = NULL; struct platform_device *pdev; - u8 *fwdata = NULL, *edid, *new_edid; - int fwsize, expected; - int builtin = 0, err = 0; + u8 *fwdata, *edid; + int fwsize, expected, err, builtin; int i, valid_extensions = 0; bool print_bad_edid = !connector->bad_edid_counter || (drm_debug & DRM_UT_KMS); pdev = platform_device_register_simple(connector_name, -1, NULL, 0); - if (IS_ERR(pdev)) { - DRM_ERROR("Failed to register EDID firmware platform device " - "for connector \"%s\"\n", connector_name); - err = -EINVAL; - goto out; - } - - err = request_firmware(&fw, name, &pdev->dev); - platform_device_unregister(pdev); + if (!IS_ERR(pdev)) { + err = request_firmware(&fw, name, &pdev->dev); + platform_device_unregister(pdev); + } else + err = PTR_ERR(pdev); - if (err) { + if (err == 0) { + fwdata = (u8 *)fw->data; + fwsize = fw->size; + builtin = 0; + } else { i = 0; while (i < GENERIC_EDIDS && strcmp(name, generic_edid_name[i])) i++; - if (i < GENERIC_EDIDS) { - err = 0; - builtin = 1; - fwdata = generic_edid[i]; - fwsize = sizeof(generic_edid[i]); + if (i >= GENERIC_EDIDS) { + DRM_ERROR("Requesting EDID firmware \"%s\" failed (err=%d)\n", + name, err); + edid = ERR_PTR(err); + goto out; } - } - if (err) { - DRM_ERROR("Requesting EDID firmware \"%s\" failed (err=%d)\n", - name, err); - goto out; - } - - if (fwdata == NULL) { - fwdata = (u8 *) fw->data; - fwsize = fw->size; + fwdata = generic_edid[i]; + fwsize = sizeof(generic_edid[i]); + builtin = 1; } expected = (fwdata[0x7e] + 1) * EDID_LENGTH; if (expected != fwsize) { DRM_ERROR("Size of EDID firmware \"%s\" is invalid " "(expected %d, got %d)\n", name, expected, (int) fwsize); - err = -EINVAL; + edid = ERR_PTR(-EINVAL); goto relfw_out; } edid = kmemdup(fwdata, fwsize, GFP_KERNEL); if (edid == NULL) { - err = -ENOMEM; + edid = ERR_PTR(-ENOMEM); goto relfw_out; } @@ -197,7 +189,7 @@ static u8 *edid_load(struct drm_connector *connector, const char *name, DRM_ERROR("Base block of EDID firmware \"%s\" is invalid ", name); kfree(edid); - err = -EINVAL; + edid = ERR_PTR(-EINVAL); goto relfw_out; } @@ -210,19 +202,18 @@ static u8 *edid_load(struct drm_connector *connector, const char *name, } if (valid_extensions != edid[0x7e]) { + u8 *new_edid; + edid[EDID_LENGTH-1] += edid[0x7e] - valid_extensions; DRM_INFO("Found %d valid extensions instead of %d in EDID data " "\"%s\" for connector \"%s\"\n", valid_extensions, edid[0x7e], name, connector_name); edid[0x7e] = valid_extensions; + new_edid = krealloc(edid, (valid_extensions + 1) * EDID_LENGTH, - GFP_KERNEL); - if (new_edid == NULL) { - err = -ENOMEM; - kfree(edid); - goto relfw_out; - } - edid = new_edid; + GFP_KERNEL); + if (new_edid) + edid = new_edid; } DRM_INFO("Got %s EDID base block and %d extension%s from " @@ -231,12 +222,10 @@ static u8 *edid_load(struct drm_connector *connector, const char *name, name, connector_name); relfw_out: - release_firmware(fw); + if (fw) + release_firmware(fw); out: - if (err) - return ERR_PTR(err); - return edid; } -- 1.7.9.5 [View Less]

3 9

[PATCH] vgacon: prevent vgacon from reloading
by David Herrmann 02 Oct '13

02 Oct '13

If vgacon got unloaded for any reason, we can never be sure that vga registers are still accessible. fbdev/DRM/xycon might reconfigure graphics devices in a way that prevents vgacon from working again. Therefore, inhibit rebinding vgacon. Note that this _might_ break use-cases where users unbind vgacon and then rebind it without breaking VGA mode. However, this seems to be rarely useful and given that several GPUs cannot even restore VGA mode it seems safer to prevent it entirely. This also … [View More]adds a dependency to dummy_con as the VT layer does not allow failure during console-rebinding. Hence, we need a guarantee that dummy_con is available. To avoid recursive Kconfig dependencies, we move the vga-con/sgi-con conflict directly into VGA_CONSOLE and make DUMMY_CONSOLE always active. Signed-off-by: David Herrmann <dh.herrmann(a)gmail.com> --- drivers/video/console/Kconfig | 2 +- drivers/video/console/vgacon.c | 10 ++++++---- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/drivers/video/console/Kconfig b/drivers/video/console/Kconfig index 846caab..173f1e7 100644 --- a/drivers/video/console/Kconfig +++ b/drivers/video/console/Kconfig @@ -8,6 +8,7 @@ config VGA_CONSOLE bool "VGA text console" if EXPERT || !X86 depends on !4xx && !8xx && !SPARC && !M68K && !PARISC && !FRV && \ !SUPERH && !BLACKFIN && !AVR32 && !MN10300 && !CRIS && \ + SGI_NEWPORT_CONSOLE!=y && DUMMY_CONSOLE && \ (!ARM || ARCH_FOOTBRIDGE || ARCH_INTEGRATOR || ARCH_NETWINDER) default y help @@ -71,7 +72,6 @@ config SGI_NEWPORT_CONSOLE config DUMMY_CONSOLE bool - depends on VGA_CONSOLE!=y || SGI_NEWPORT_CONSOLE!=y default y config DUMMY_CONSOLE_COLUMNS diff --git a/drivers/video/console/vgacon.c b/drivers/video/console/vgacon.c index 9d8feac..d598af6 100644 --- a/drivers/video/console/vgacon.c +++ b/drivers/video/console/vgacon.c @@ -368,6 +368,7 @@ static int vgacon_scrolldelta(struct vc_data *c, int lines) static const char *vgacon_startup(void) { + static bool vgacon_inhibit; const char *display_desc = NULL; u16 saved1, saved2; volatile u16 *p; @@ -375,14 +376,15 @@ static const char *vgacon_startup(void) if (screen_info.orig_video_isVGA == VIDEO_TYPE_VLFB || screen_info.orig_video_isVGA == VIDEO_TYPE_EFI) { no_vga: -#ifdef CONFIG_DUMMY_CONSOLE conswitchp = &dummy_con; return conswitchp->con_startup(); -#else - return NULL; -#endif } + /* fbdev/DRM can disable VGA-mode so prevent reloading */ + if (vgacon_inhibit) + goto no_vga; + vgacon_inhibit = true; + /* boot_params.screen_info initialized? */ if ((screen_info.orig_video_mode == 0) && (screen_info.orig_video_lines == 0) && -- 1.8.4 [View Less]

1 1

[PATCHv3 0/4] gpu: host1x: Add runtime pm support
by Arto Merilainen 02 Oct '13

02 Oct '13

This series adds runtime pm support for host1x, gr2d and dc. It retains the current behaviour if CONFIG_PM_RUNTIME is not enabled. The gr2d clock is enabled when a new job is submitted and disabled when the work is done. Due to parent->child relations between host1x->gr2d, this scheme enables and disables host1x clock. For dc, the clocks are enabled in .probe and disabled in .remove via runtime pm instead of direct clock APIs. Mayuresh is unfortunately not available to continue with … [View More]

2 8

[Bug 69671] New: rv790 hdmi sound regression since fix audio dto calculation on DCE3+ (v3)
by bugzilla-daemon＠freedesktop.org 01 Oct '13

01 Oct '13

https://bugs.freedesktop.org/show_bug.cgi?id=69671 Priority: medium Bug ID: 69671 Assignee: dri-devel(a)lists.freedesktop.org Summary: rv790 hdmi sound regression since fix audio dto calculation on DCE3+ (v3) Severity: normal Classification: Unclassified OS: Linux (All) Reporter: adf.lists(a)gmail.com Hardware: x86-64 (AMD64) Status: NEW Version: XOrg CVS … [View More]

1 6

[PATCH 0/6] Nouveau on ARM fixes
by Lucas Stach 01 Oct '13

01 Oct '13

This is the first set of patches to make Nouveau work on Tegra. Those are only the obvious correctness fixes, a lot of optimization work remains to be done, but at least it's enough to get accel working and let the machine survive a piglit run. A new BO flag is introduced to allow userspace to hint the kernel about possible optimizations. Lucas Stach (6): drm/ttm: recognize ARM arch in ioprot handler drm/ttm: introduce dma cache sync helpers drm/nouveau: hook up cache sync functions … [View More]

6 31

[libdrm] Stereo 3D v3
by Damien Lespiau 01 Oct '13

01 Oct '13

v2 was: http://lists.freedesktop.org/archives/dri-devel/2013-September/045354.html Only one change, the stereo part for the flags is now an enum instead of being a bit field. The kernel series has been pushed to drm-intel, the libdrm patches should follow if there isn't any objection. -- Damien

1 4

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel October 2013