dri-devel August 2010

dri-devel@lists.freedesktop.org

82 participants
335 discussions

[PATCH] drm/radeon/kms: allow FG_ALPHA_VALUE on r5xx
by Marek Olšák 11 Aug '10

11 Aug '10

This is a CS checker fix. I need this for FP16 alpha-test. Signed-off-by: Marek Olšák <maraeo(a)gmail.com> --- drivers/gpu/drm/radeon/reg_srcs/rv515 | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/drivers/gpu/drm/radeon/reg_srcs/rv515 b/drivers/gpu/drm/radeon/reg_srcs/rv515 index 8293855..b3f9f1d 100644 --- a/drivers/gpu/drm/radeon/reg_srcs/rv515 +++ b/drivers/gpu/drm/radeon/reg_srcs/rv515 @@ -316,6 +316,7 @@ rv515 0x6d40 0x4BD0 FG_FOG_COLOR_B 0x4BD4 FG_ALPHA_FUNC 0x4BD8 FG_DEPTH_SRC +0x4BE0 FG_ALPHA_VALUE 0x4C00 US_ALU_CONST_R_0 0x4C04 US_ALU_CONST_G_0 0x4C08 US_ALU_CONST_B_0 -- 1.7.0.4

1 0

[PATCH] drm/radeon/kms: another r6xx/r7xx CS checker fix
by Alex Deucher 11 Aug '10

11 Aug '10

add default case for buffer formats Signed-off-by: Alex Deucher <alexdeucher(a)gmail.com> Cc: Andre Maasikas <amaasikas(a)gmail.com> --- drivers/gpu/drm/radeon/r600_cs.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/drivers/gpu/drm/radeon/r600_cs.c b/drivers/gpu/drm/radeon/r600_cs.c index 52b5252..8e165b2 100644 --- a/drivers/gpu/drm/radeon/r600_cs.c +++ b/drivers/gpu/drm/radeon/r600_cs.c @@ -133,6 +133,7 @@ static inline int r600_bpe_from_format(u32 *bpe, u32 format) case V_038004_FMT_GB_GR: case V_038004_FMT_BG_RG: case V_038004_COLOR_INVALID: + default: *bpe = 16; return -EINVAL; } -- 1.7.1.1

1 0

Re: [DRM] BUG: sleeping function called from invalid context, drm_lastclose
by Luca Tettamanti 11 Aug '10

11 Aug '10

On Wed, Aug 11, 2010 at 10:50 AM, Dave Airlie <airlied(a)gmail.com> wrote: > On Wed, Aug 11, 2010 at 6:48 PM, Luca Tettamanti <kronos.it(a)gmail.com> wrote: >> Hi Arnd, >> this commit: >> >> commit 58374713c9dfb4d231f8c56cac089f6fbdedc2ec >> Author: Arnd Bergmann <arnd(a)arndb.de> >> Date: Sat Jul 10 23:51:39 2010 +0200 >> >> drm: kill BKL from common code >> >> >> moved the call to (inside drm_release) drm_lastclose inside dev->count_lock >> spinlock. >> drm_lastclose however takes dev->struct_mutex (now inside an atomic >> context): > > I have a patch from Chris Wilson that I need to push to fix this, > basically reducing the spin lock coverage, > and relying on the global mutex to handle the open race. Ok, thank you :) Luca

1 0

[PATCH] drm/i915: set LVDS connector status to unknown if the lid is closed
by Andrei Paskevich 11 Aug '10

11 Aug '10

Currently, the LVDS connector on i915-equipped laptops is always reported as connected, regardless of the lid state and presence of external monitors. This was done because of too many BIOSes reporting lid status in an unreliable way. However, in a quite usual setup of a laptop attached to an external monitor, the only way to obtain the monitor's resolution on boot is to disable the LVDS via the boot command line, which is unfortunate as one have to reboot to get the internal screen back. The following patch implements the same scheme as in the Nouveau driver: if the lid is reported as closed, the LVDS connector status is set to unknown, and the internal screen will be turned on only if there are no other connected monitors. The ACPI lid status is not checked if the "i915.ignorelid" parameter is set to zero (to account for the buggy BIOSes), which is the default value, so that the current behavior does not change. The patch applies to the 2.6.35 kernel. Best regards, Andrei diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 423dc90..5206136 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -45,6 +45,9 @@ module_param_named(fbpercrtc, i915_fbpercrtc, int, 0400); unsigned int i915_powersave = 1; module_param_named(powersave, i915_powersave, int, 0400); +unsigned int i915_ignorelid = 1; +module_param_named(ignorelid, i915_ignorelid, int, 0400); + unsigned int i915_lvds_downclock = 0; module_param_named(lvds_downclock, i915_lvds_downclock, int, 0400); diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 2e1744d..a03d344 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -815,6 +815,7 @@ extern struct drm_ioctl_desc i915_ioctls[]; extern int i915_max_ioctl; extern unsigned int i915_fbpercrtc; extern unsigned int i915_powersave; +extern unsigned int i915_ignorelid; extern unsigned int i915_lvds_downclock; extern int i915_suspend(struct drm_device *dev, pm_message_t state); diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 0eab8df..85a09c9 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -551,11 +551,8 @@ static enum drm_connector_status intel_lvds_detect(struct drm_connector *connect struct drm_device *dev = connector->dev; enum drm_connector_status status = connector_status_connected; - /* ACPI lid methods were generally unreliable in this generation, so - * don't even bother. - */ - if (IS_GEN2(dev) || IS_GEN3(dev)) - return connector_status_connected; + if (!i915_ignorelid && !acpi_lid_open()) + status = connector_status_unknown; return status; }

3 4

[PATCH] drm: Release spinlock for calling lastclose() after 58474713
by Chris Wilson 11 Aug '10

11 Aug '10

When removing of the BKL the locking around lastclose() was rearranged and resulted in the holding of the open_count spinlock over the call into drm_lastclose(). The drivers were not ready for this path to be atomic - it may indeed involve long waits to release old objects and cleanup the GPU - and so we ended up scheduling whilst atomic. [ 54.625598] BUG: scheduling while atomic: X/3546/0x00000002 [ 54.625600] Modules linked in: sco bridge stp llc input_polldev rfcomm bnep l2cap crc16 sch_sfq ipv6 md_mod acpi_cpufreq mperf cryptd aes_x86_64 aes_generic xts gf128mul dm_crypt dm_mod btusb bluetooth usbhid hid zaurus cdc_ether usbnet mii cdc_wdm cdc_acm uvcvideo videodev v4l1_compat v4l2_compat_ioctl32 snd_hda_codec_conexant arc4 pcmcia ecb snd_hda_intel joydev sdhci_pci sdhci snd_hda_codec tpm_tis firewire_ohci mmc_core e1000e uhci_hcd thinkpad_acpi nvram yenta_socket pcmcia_rsrc pcmcia_core tpm wmi sr_mod firewire_core iwlagn ehci_hcd snd_hwdep snd_pcm usbcore tpm_bios thermal led_class snd_timer iwlcore snd soundcore ac snd_page_alloc pcspkr psmouse serio_raw battery sg mac80211 evdev cfg80211 i2c_i801 iTCO_wdt iTCO_vendor_support cdrom processor crc_itu_t rfkill xfs exportfs sd_mod crc_t10dif ahci libahci libata scsi_mod [last unloaded: scsi_wait_scan] [ 54.625663] Pid: 3546, comm: X Not tainted 2.6.35-04771-g1787985 #301 [ 54.625665] Call Trace: [ 54.625671] [<ffffffff8102d599>] __schedule_bug+0x57/0x5c [ 54.625675] [<ffffffff81384141>] schedule+0xe5/0x832 [ 54.625679] [<ffffffff81163e77>] ? put_dec+0x20/0x3c [ 54.625682] [<ffffffff81384dd4>] schedule_timeout+0x275/0x29f [ 54.625686] [<ffffffff810455e1>] ? process_timeout+0x0/0xb [ 54.625688] [<ffffffff81384e17>] schedule_timeout_uninterruptible+0x19/0x1b [ 54.625691] [<ffffffff81045893>] msleep+0x16/0x1d [ 54.625695] [<ffffffff812a2e53>] i9xx_crtc_dpms+0x273/0x2ae [ 54.625698] [<ffffffff812a18be>] intel_crtc_dpms+0x28/0xe7 [ 54.625702] [<ffffffff811ec0fa>] drm_helper_disable_unused_functions+0xf0/0x118 [ 54.625705] [<ffffffff811ecde3>] drm_crtc_helper_set_config+0x644/0x7c8 [ 54.625708] [<ffffffff811f12dd>] ? drm_copy_field+0x40/0x50 [ 54.625711] [<ffffffff811ebca2>] drm_fb_helper_force_kernel_mode+0x3e/0x85 [ 54.625713] [<ffffffff811ebcf2>] drm_fb_helper_restore+0x9/0x24 [ 54.625717] [<ffffffff81290a41>] i915_driver_lastclose+0x2b/0x5c [ 54.625720] [<ffffffff811f14a7>] drm_lastclose+0x44/0x2ad [ 54.625722] [<ffffffff811f1ed2>] drm_release+0x5c6/0x609 [ 54.625726] [<ffffffff810d1275>] fput+0x109/0x1c7 [ 54.625728] [<ffffffff810ce5e4>] filp_close+0x61/0x6b [ 54.625731] [<ffffffff810ce680>] sys_close+0x92/0xd4 [ 54.625734] [<ffffffff81002a2b>] system_call_fastpath+0x16/0x1b Signed-off-by: Chris Wilson <chris(a)chris-wilson.co.uk> --- drivers/gpu/drm/drm_fops.c | 9 ++++----- 1 files changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c index 2ca8df8..6367961 100644 --- a/drivers/gpu/drm/drm_fops.c +++ b/drivers/gpu/drm/drm_fops.c @@ -572,16 +572,15 @@ int drm_release(struct inode *inode, struct file *filp) atomic_inc(&dev->counts[_DRM_STAT_CLOSES]); spin_lock(&dev->count_lock); if (!--dev->open_count) { + spin_unlock(&dev->count_lock); + if (atomic_read(&dev->ioctl_count)) { DRM_ERROR("Device busy: %d\n", atomic_read(&dev->ioctl_count)); retcode = -EBUSY; - goto out; - } - retcode = drm_lastclose(dev); + } else + retcode = drm_lastclose(dev); } -out: - spin_unlock(&dev->count_lock); mutex_unlock(&drm_global_mutex); return retcode; -- 1.7.1

3 5

[PATCH] i915: Remove redundant NULL check upon kfree()
by Davidlohr Bueso 11 Aug '10

11 Aug '10

i915: Remove redundant NULL check upon kfree(). Signed-off-by: Davidlohr Bueso <dave(a)gnu.org> --- drivers/gpu/drm/i915/intel_dvo.c | 3 +-- 1 files changed, 1 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c index 227feca..b3955bc 100644 --- a/drivers/gpu/drm/i915/intel_dvo.c +++ b/drivers/gpu/drm/i915/intel_dvo.c @@ -283,8 +283,7 @@ static void intel_dvo_enc_destroy(struct drm_encoder *encoder) if (dvo) { if (dvo->dev_ops->destroy) dvo->dev_ops->destroy(dvo); - if (dvo->panel_fixed_mode) - kfree(dvo->panel_fixed_mode); + kfree(dvo->panel_fixed_mode); } if (intel_encoder->i2c_bus) intel_i2c_destroy(intel_encoder->i2c_bus); -- 1.7.0.4

1 0

[PATCH] drm/radeon/kms: avoid corner case issue with unmappable vram V2
by glisse＠freedesktop.org 10 Aug '10

10 Aug '10

From: Jerome Glisse <jglisse(a)redhat.com> We should not allocate any object into unmappable vram if we have no means to access them which on all GPU means having the CP running and on newer GPU having the blit utility working. This patch limit the vram allocation to visible vram until we have acceleration up and running. Note that it's more than unlikely that we run into any issue related to that as when acceleration is not woring userspace should allocate any object in vram beside front buffer which should fit in visible vram. V2 use real_vram_size as mc_vram_size could be bigger than the actual amount of vram Signed-off-by: Jerome Glisse <jglisse(a)redhat.com> --- drivers/gpu/drm/radeon/evergreen.c | 1 + drivers/gpu/drm/radeon/r100.c | 3 +++ drivers/gpu/drm/radeon/r600.c | 2 ++ drivers/gpu/drm/radeon/r600_blit_kms.c | 2 ++ drivers/gpu/drm/radeon/radeon.h | 1 + drivers/gpu/drm/radeon/radeon_object.c | 2 +- drivers/gpu/drm/radeon/rs600.c | 1 + drivers/gpu/drm/radeon/rs690.c | 1 + drivers/gpu/drm/radeon/rv770.c | 1 + 9 files changed, 13 insertions(+), 1 deletions(-) diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 957d506..4243344 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -1359,6 +1359,7 @@ int evergreen_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE) * 1024 * 1024; rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE) * 1024 * 1024; rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r600_vram_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index e817a0b..566c1e5 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -1030,6 +1030,7 @@ int r100_cp_init(struct radeon_device *rdev, unsigned ring_size) return r; } rdev->cp.ready = true; + rdev->mc.active_vram_size = rdev->mc.real_vram_size; return 0; } @@ -1047,6 +1048,7 @@ void r100_cp_fini(struct radeon_device *rdev) void r100_cp_disable(struct radeon_device *rdev) { /* Disable ring */ + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; rdev->cp.ready = false; WREG32(RADEON_CP_CSQ_MODE, 0); WREG32(RADEON_CP_CSQ_CNTL, 0); @@ -2306,6 +2308,7 @@ void r100_vram_init_sizes(struct radeon_device *rdev) /* FIXME we don't use the second aperture yet when we could use it */ if (rdev->mc.visible_vram_size > rdev->mc.aper_size) rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; config_aper_size = RREG32(RADEON_CONFIG_APER_SIZE); if (rdev->flags & RADEON_IS_IGP) { uint32_t tom; diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index d0ebae9..acc71f2 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -1248,6 +1248,7 @@ int r600_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r600_vram_gtt_location(rdev, &rdev->mc); if (rdev->flags & RADEON_IS_IGP) { @@ -1917,6 +1918,7 @@ void r600_pciep_wreg(struct radeon_device *rdev, u32 reg, u32 v) */ void r600_cp_stop(struct radeon_device *rdev) { + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1)); } diff --git a/drivers/gpu/drm/radeon/r600_blit_kms.c b/drivers/gpu/drm/radeon/r600_blit_kms.c index d13622a..968ff3a 100644 --- a/drivers/gpu/drm/radeon/r600_blit_kms.c +++ b/drivers/gpu/drm/radeon/r600_blit_kms.c @@ -507,6 +507,7 @@ int r600_blit_init(struct radeon_device *rdev) memcpy(ptr + rdev->r600_blit.ps_offset, r6xx_ps, r6xx_ps_size * 4); radeon_bo_kunmap(rdev->r600_blit.shader_obj); radeon_bo_unreserve(rdev->r600_blit.shader_obj); + rdev->mc.active_vram_size = rdev->mc.real_vram_size; return 0; } @@ -514,6 +515,7 @@ void r600_blit_fini(struct radeon_device *rdev) { int r; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; if (rdev->r600_blit.shader_obj == NULL) return; /* If we can't reserve the bo, unref should be enough to destroy diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 3cd1c47..feeac20 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -344,6 +344,7 @@ struct radeon_mc { * about vram size near mc fb location */ u64 mc_vram_size; u64 visible_vram_size; + u64 active_vram_size; u64 gtt_size; u64 gtt_start; u64 gtt_end; diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 0afd1e6..b3b5306 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -69,7 +69,7 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain) u32 c = 0; rbo->placement.fpfn = 0; - rbo->placement.lpfn = 0; + rbo->placement.lpfn = rbo->rdev->mc.active_vram_size >> PAGE_SHIFT; rbo->placement.placement = rbo->placements; rbo->placement.busy_placement = rbo->placements; if (domain & RADEON_GEM_DOMAIN_VRAM) diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c index cc05b23..51d5f7b 100644 --- a/drivers/gpu/drm/radeon/rs600.c +++ b/drivers/gpu/drm/radeon/rs600.c @@ -693,6 +693,7 @@ void rs600_mc_init(struct radeon_device *rdev) rdev->mc.real_vram_size = RREG32(RADEON_CONFIG_MEMSIZE); rdev->mc.mc_vram_size = rdev->mc.real_vram_size; rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); base = RREG32_MC(R_000004_MC_FB_LOCATION); base = G_000004_MC_FB_START(base) << 16; diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c index 3e3f757..4dc2a87 100644 --- a/drivers/gpu/drm/radeon/rs690.c +++ b/drivers/gpu/drm/radeon/rs690.c @@ -157,6 +157,7 @@ void rs690_mc_init(struct radeon_device *rdev) rdev->mc.aper_base = pci_resource_start(rdev->pdev, 0); rdev->mc.aper_size = pci_resource_len(rdev->pdev, 0); rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; base = RREG32_MC(R_000100_MCCFG_FB_LOCATION); base = G_000100_MC_FB_START(base) << 16; rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index f1c79681..3e6293e 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -944,6 +944,7 @@ int rv770_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r600_vram_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); -- 1.7.2

2 1

[PATCH] drm/radeon/kms: avoid corner case issue with unmappable vram
by glisse＠freedesktop.org 10 Aug '10

10 Aug '10

From: Jerome Glisse <jglisse(a)redhat.com> We should not allocate any object into unmappable vram if we have no means to access them which on all GPU means having the CP running and on newer GPU having the blit utility working. This patch limit the vram allocation to visible vram until we have acceleration up and running. Note that it's more than unlikely that we run into any issue related to that as when acceleration is not woring userspace should allocate any object in vram beside front buffer which should fit in visible vram. Signed-off-by: Jerome Glisse <jglisse(a)redhat.com> --- drivers/gpu/drm/radeon/evergreen.c | 1 + drivers/gpu/drm/radeon/r100.c | 3 +++ drivers/gpu/drm/radeon/r600.c | 2 ++ drivers/gpu/drm/radeon/r600_blit_kms.c | 2 ++ drivers/gpu/drm/radeon/radeon.h | 1 + drivers/gpu/drm/radeon/radeon_object.c | 2 +- drivers/gpu/drm/radeon/rs600.c | 1 + drivers/gpu/drm/radeon/rs690.c | 1 + drivers/gpu/drm/radeon/rv770.c | 1 + 9 files changed, 13 insertions(+), 1 deletions(-) diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 957d506..4243344 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -1359,6 +1359,7 @@ int evergreen_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE) * 1024 * 1024; rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE) * 1024 * 1024; rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r600_vram_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c index e817a0b..4be71ba 100644 --- a/drivers/gpu/drm/radeon/r100.c +++ b/drivers/gpu/drm/radeon/r100.c @@ -1030,6 +1030,7 @@ int r100_cp_init(struct radeon_device *rdev, unsigned ring_size) return r; } rdev->cp.ready = true; + rdev->mc.active_vram_size = rdev->mc.mc_vram_size; return 0; } @@ -1047,6 +1048,7 @@ void r100_cp_fini(struct radeon_device *rdev) void r100_cp_disable(struct radeon_device *rdev) { /* Disable ring */ + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; rdev->cp.ready = false; WREG32(RADEON_CP_CSQ_MODE, 0); WREG32(RADEON_CP_CSQ_CNTL, 0); @@ -2306,6 +2308,7 @@ void r100_vram_init_sizes(struct radeon_device *rdev) /* FIXME we don't use the second aperture yet when we could use it */ if (rdev->mc.visible_vram_size > rdev->mc.aper_size) rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; config_aper_size = RREG32(RADEON_CONFIG_APER_SIZE); if (rdev->flags & RADEON_IS_IGP) { uint32_t tom; diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index d0ebae9..acc71f2 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -1248,6 +1248,7 @@ int r600_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r600_vram_gtt_location(rdev, &rdev->mc); if (rdev->flags & RADEON_IS_IGP) { @@ -1917,6 +1918,7 @@ void r600_pciep_wreg(struct radeon_device *rdev, u32 reg, u32 v) */ void r600_cp_stop(struct radeon_device *rdev) { + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1)); } diff --git a/drivers/gpu/drm/radeon/r600_blit_kms.c b/drivers/gpu/drm/radeon/r600_blit_kms.c index d13622a..c7caa6e 100644 --- a/drivers/gpu/drm/radeon/r600_blit_kms.c +++ b/drivers/gpu/drm/radeon/r600_blit_kms.c @@ -507,6 +507,7 @@ int r600_blit_init(struct radeon_device *rdev) memcpy(ptr + rdev->r600_blit.ps_offset, r6xx_ps, r6xx_ps_size * 4); radeon_bo_kunmap(rdev->r600_blit.shader_obj); radeon_bo_unreserve(rdev->r600_blit.shader_obj); + rdev->mc.active_vram_size = rdev->mc.mc_vram_size; return 0; } @@ -514,6 +515,7 @@ void r600_blit_fini(struct radeon_device *rdev) { int r; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; if (rdev->r600_blit.shader_obj == NULL) return; /* If we can't reserve the bo, unref should be enough to destroy diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 3cd1c47..feeac20 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -344,6 +344,7 @@ struct radeon_mc { * about vram size near mc fb location */ u64 mc_vram_size; u64 visible_vram_size; + u64 active_vram_size; u64 gtt_size; u64 gtt_start; u64 gtt_end; diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 0afd1e6..b3b5306 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -69,7 +69,7 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain) u32 c = 0; rbo->placement.fpfn = 0; - rbo->placement.lpfn = 0; + rbo->placement.lpfn = rbo->rdev->mc.active_vram_size >> PAGE_SHIFT; rbo->placement.placement = rbo->placements; rbo->placement.busy_placement = rbo->placements; if (domain & RADEON_GEM_DOMAIN_VRAM) diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c index cc05b23..51d5f7b 100644 --- a/drivers/gpu/drm/radeon/rs600.c +++ b/drivers/gpu/drm/radeon/rs600.c @@ -693,6 +693,7 @@ void rs600_mc_init(struct radeon_device *rdev) rdev->mc.real_vram_size = RREG32(RADEON_CONFIG_MEMSIZE); rdev->mc.mc_vram_size = rdev->mc.real_vram_size; rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); base = RREG32_MC(R_000004_MC_FB_LOCATION); base = G_000004_MC_FB_START(base) << 16; diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c index 3e3f757..4dc2a87 100644 --- a/drivers/gpu/drm/radeon/rs690.c +++ b/drivers/gpu/drm/radeon/rs690.c @@ -157,6 +157,7 @@ void rs690_mc_init(struct radeon_device *rdev) rdev->mc.aper_base = pci_resource_start(rdev->pdev, 0); rdev->mc.aper_size = pci_resource_len(rdev->pdev, 0); rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; base = RREG32_MC(R_000100_MCCFG_FB_LOCATION); base = G_000100_MC_FB_START(base) << 16; rdev->mc.igp_sideport_enabled = radeon_atombios_sideport_present(rdev); diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index f1c79681..3e6293e 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -944,6 +944,7 @@ int rv770_mc_init(struct radeon_device *rdev) rdev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE); rdev->mc.visible_vram_size = rdev->mc.aper_size; + rdev->mc.active_vram_size = rdev->mc.visible_vram_size; r600_vram_gtt_location(rdev, &rdev->mc); radeon_update_bandwidth_info(rdev); -- 1.7.2

1 0

[PATCH] drm/radeon/kms: use tracked values for sclk and mclk
by Alex Deucher 10 Aug '10

10 Aug '10

Rather than calling get_memory_clock and get_engine_clock, used the tracked values from the pm code. Calling the tables adds additional latency in the modesetting and pm paths. Signed-off-by: Alex Deucher <alexdeucher(a)gmail.com> --- drivers/gpu/drm/radeon/radeon_clocks.c | 8 ++++++++ drivers/gpu/drm/radeon/radeon_device.c | 28 +++++++++------------------- drivers/gpu/drm/radeon/radeon_i2c.c | 2 +- 3 files changed, 18 insertions(+), 20 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_clocks.c b/drivers/gpu/drm/radeon/radeon_clocks.c index f64936c..5e916df 100644 --- a/drivers/gpu/drm/radeon/radeon_clocks.c +++ b/drivers/gpu/drm/radeon/radeon_clocks.c @@ -246,6 +246,14 @@ void radeon_get_clock_info(struct drm_device *dev) mpll->max_feedback_div = 0xff; mpll->best_vco = 0; + if (!rdev->clock.default_sclk) + rdev->clock.default_sclk = radeon_get_engine_clock(rdev); + if ((!rdev->clock.default_mclk) && rdev->asic->get_memory_clock) + rdev->clock.default_mclk = radeon_get_memory_clock(rdev); + + rdev->pm.current_sclk = rdev->clock.default_sclk; + rdev->pm.current_mclk = rdev->clock.default_mclk; + } /* 10 khz */ diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 39a2b97..a1f3361 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -363,30 +363,20 @@ bool radeon_card_posted(struct radeon_device *rdev) void radeon_update_bandwidth_info(struct radeon_device *rdev) { fixed20_12 a; - u32 sclk, mclk; + u32 sclk = rdev->pm.current_sclk; + u32 mclk = rdev->pm.current_mclk; - if (rdev->flags & RADEON_IS_IGP) { - sclk = radeon_get_engine_clock(rdev); - mclk = rdev->clock.default_mclk; - - a.full = dfixed_const(100); - rdev->pm.sclk.full = dfixed_const(sclk); - rdev->pm.sclk.full = dfixed_div(rdev->pm.sclk, a); - rdev->pm.mclk.full = dfixed_const(mclk); - rdev->pm.mclk.full = dfixed_div(rdev->pm.mclk, a); + /* sclk/mclk in Mhz */ + a.full = dfixed_const(100); + rdev->pm.sclk.full = dfixed_const(sclk); + rdev->pm.sclk.full = dfixed_div(rdev->pm.sclk, a); + rdev->pm.mclk.full = dfixed_const(mclk); + rdev->pm.mclk.full = dfixed_div(rdev->pm.mclk, a); + if (rdev->flags & RADEON_IS_IGP) { a.full = dfixed_const(16); /* core_bandwidth = sclk(Mhz) * 16 */ rdev->pm.core_bandwidth.full = dfixed_div(rdev->pm.sclk, a); - } else { - sclk = radeon_get_engine_clock(rdev); - mclk = radeon_get_memory_clock(rdev); - - a.full = dfixed_const(100); - rdev->pm.sclk.full = dfixed_const(sclk); - rdev->pm.sclk.full = dfixed_div(rdev->pm.sclk, a); - rdev->pm.mclk.full = dfixed_const(mclk); - rdev->pm.mclk.full = dfixed_div(rdev->pm.mclk, a); } } diff --git a/drivers/gpu/drm/radeon/radeon_i2c.c b/drivers/gpu/drm/radeon/radeon_i2c.c index bfd2ce5..ea15e62 100644 --- a/drivers/gpu/drm/radeon/radeon_i2c.c +++ b/drivers/gpu/drm/radeon/radeon_i2c.c @@ -206,7 +206,7 @@ static void post_xfer(struct i2c_adapter *i2c_adap) static u32 radeon_get_i2c_prescale(struct radeon_device *rdev) { - u32 sclk = radeon_get_engine_clock(rdev); + u32 sclk = rdev->pm.current_sclk; u32 prescale = 0; u32 nm; u8 n, m, loop; -- 1.7.1.1

1 0

[git pull] drm tree some more bits
by Dave Airlie 10 Aug '10

10 Aug '10

The main reason for this, is Ben got a nvidia fermi card and added modesetting support just before the window opened, there is no accel support, but having kms support for these GPUs is a good start. Its all the GTX4xx cards. The other bulk is because ajax moved a load of modes to their own header file because they were making things ugly, Alex added support for a new modesetting object that we are seeing used on radeons in the field, with this they can bring up a monitor in a few configurations that aren't possible before. Also there is an API change to the hyperz info ioctl we just added that we'd really like to make before -rc1 along with an oops fixer in that function. Also myself and BenH discussed fixing pci domains a long time ago, and we forgot to finish the job, so his patch to make drm work on non-0 domains without breaking userspace is in here finally. The following changes since commit 96576a9e1a0cdb8a43d3af5846be0948f52b4460: agp: intel-agp: do not use PCI resources before pci_enable_device() (2010-08-05 12:28:25 +1000) are available in the git repository at: ssh://master.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6.git drm-core-next Adam Jackson (5): drm: Remove unused fields from drm_display_info drm/edid: Add detailed block walk for CEA extensions drm/edid: Add detailed block walk for VTB extensions drm/edid: Rewrite mode parse to use the generic detailed block walk drm/edid: Split mode lists out to their own header for readability Alex Deucher (7): drm/radeon/kms: clean up r6xx/r7xx blit init drm/radeon/kms: r600 CS parser fixes drm/radeon/kms: unify i2c handling drm/radeon/kms: rework combios i2c handling drm/radeon/kms: add support for router objects drm/radeon/kms: add additional quirk for Acer rv620 laptop drm/radeon/kms: mark 3D power states as performance Ben Skeggs (11): drm/nv50: use custom i2c algo for dp auxch drm/nouveau: remove unused ttm bo list drm/nvc0: starting point for GF100 support, everything stubbed drm/nvc0: allow INIT_GPIO drm/nvc0: implement memory detection drm/nvc0: rudimentary instmem support drm/nvc0: fix evo dma object so we display something drm/nvc0: implement crtc pll setting drm/nvc0: backup bar3 channel on suspend drm/nouveau: reduce severity of some "error" messages drm/nvc0: fix typo in PRAMIN flush Benjamin Herrenschmidt (1): drm: Fix support for PCI domains Chris Wilson (3): drm: Propagate error code from fb_create() drm: Only set DPMS once on the CRTC not after every encoder. drm: Use ENOENT consistently for the error return for an unmatched handle. Dan Carpenter (2): drm/nouveau: unwind on load errors radeon: handle errors in radeon_hwmon_init() Dave Airlie (1): Merge remote branch 'nouveau/for-airlied' of /ssd/git/drm-nouveau-next into drm-core-next Dr. David Alan Gilbert (1): drm/radeon/kms: add missing copy from user Francisco Jerez (13): drm/nouveau: Fix AGP reset when AGP FW is already enabled on init. drm/nouveau: Don't pass misaligned offsets to io_mapping_map_atomic_wc(). drm/nouveau: Fix the INIT_CONFIGURE_PREINIT BIOS opcode. drm/nouveau: Ack the context switch interrupt before switching contexts. drm/nv10: Fix up switching of NV10TCL_DMA_VTXBUF. drm/nv17-nv4x: Attempt to init some external TMDS transmitters. drm/nouveau: No need to set slave TV encoder configs explicitly. drm/nv30: Workaround dual TMDS brain damage. drm/nouveau: Init dcb->or on cards that have no usable DCB table. drm/i2c/ch7006: Don't use POWER_LEVEL_FULL_POWER_OFF on early chip versions. drm/nv04: Fix up SGRAM density detection. drm/nv30: Fix PFB init for nv31. drm/nouveau: Fix DCB TMDS config parsing. James Simmons (1): drm: expand gamma_set Marcin Kościelnicki (2): drm/nouveau: implement init table op 0x57, INIT_LTIME drm/nouveau: implement init table opcodex 0x5e and 0x9a Marek Olšák (1): drm/radeon: add a way to revoke hyper-z access drivers/gpu/drm/drm_crtc.c | 6 +- drivers/gpu/drm/drm_crtc_helper.c | 10 +- drivers/gpu/drm/drm_edid.c | 841 +++++++++------------------ drivers/gpu/drm/drm_edid_modes.h | 380 ++++++++++++ drivers/gpu/drm/drm_gem.c | 2 +- drivers/gpu/drm/drm_ioctl.c | 1 + drivers/gpu/drm/i2c/ch7006_drv.c | 1 + drivers/gpu/drm/i2c/ch7006_mode.c | 5 +- drivers/gpu/drm/i2c/ch7006_priv.h | 1 + drivers/gpu/drm/i915/i915_gem.c | 26 +- drivers/gpu/drm/i915/i915_gem_tiling.c | 4 +- drivers/gpu/drm/i915/intel_display.c | 15 +- drivers/gpu/drm/i915/intel_fb.c | 4 +- drivers/gpu/drm/nouveau/Makefile | 8 +- drivers/gpu/drm/nouveau/nouveau_bios.c | 194 ++++++- drivers/gpu/drm/nouveau/nouveau_bios.h | 1 + drivers/gpu/drm/nouveau/nouveau_bo.c | 6 - drivers/gpu/drm/nouveau/nouveau_connector.c | 19 +- drivers/gpu/drm/nouveau/nouveau_display.c | 6 +- drivers/gpu/drm/nouveau/nouveau_dp.c | 87 ++-- drivers/gpu/drm/nouveau/nouveau_drv.h | 45 ++- drivers/gpu/drm/nouveau/nouveau_encoder.h | 6 + drivers/gpu/drm/nouveau/nouveau_fbcon.c | 2 + drivers/gpu/drm/nouveau/nouveau_gem.c | 8 +- drivers/gpu/drm/nouveau/nouveau_hw.c | 11 +- drivers/gpu/drm/nouveau/nouveau_i2c.c | 41 +- drivers/gpu/drm/nouveau/nouveau_i2c.h | 8 +- drivers/gpu/drm/nouveau/nouveau_irq.c | 6 +- drivers/gpu/drm/nouveau/nouveau_mem.c | 15 +- drivers/gpu/drm/nouveau/nouveau_reg.h | 18 +- drivers/gpu/drm/nouveau/nouveau_state.c | 91 +++- drivers/gpu/drm/nouveau/nv04_crtc.c | 15 +- drivers/gpu/drm/nouveau/nv04_dfp.c | 73 +++ drivers/gpu/drm/nouveau/nv04_tv.c | 10 +- drivers/gpu/drm/nouveau/nv10_graph.c | 175 ++++-- drivers/gpu/drm/nouveau/nv30_fb.c | 24 +- drivers/gpu/drm/nouveau/nv50_crtc.c | 34 +- drivers/gpu/drm/nouveau/nv50_display.c | 5 +- drivers/gpu/drm/nouveau/nvc0_fb.c | 38 ++ drivers/gpu/drm/nouveau/nvc0_fifo.c | 96 +++ drivers/gpu/drm/nouveau/nvc0_graph.c | 75 +++ drivers/gpu/drm/nouveau/nvc0_instmem.c | 232 ++++++++ drivers/gpu/drm/radeon/r600.c | 14 +- drivers/gpu/drm/radeon/r600_blit_kms.c | 16 +- drivers/gpu/drm/radeon/r600_cs.c | 27 +- drivers/gpu/drm/radeon/radeon.h | 2 + drivers/gpu/drm/radeon/radeon_atombios.c | 180 +++++- drivers/gpu/drm/radeon/radeon_combios.c | 388 ++++++------- drivers/gpu/drm/radeon/radeon_connectors.c | 45 +- drivers/gpu/drm/radeon/radeon_cs.c | 2 +- drivers/gpu/drm/radeon/radeon_cursor.c | 2 +- drivers/gpu/drm/radeon/radeon_display.c | 34 +- drivers/gpu/drm/radeon/radeon_gem.c | 12 +- drivers/gpu/drm/radeon/radeon_i2c.c | 82 +++ drivers/gpu/drm/radeon/radeon_kms.c | 28 +- drivers/gpu/drm/radeon/radeon_mode.h | 24 + drivers/gpu/drm/radeon/radeon_pm.c | 24 +- drivers/gpu/drm/radeon/rv770.c | 14 +- drivers/gpu/drm/vmwgfx/vmwgfx_kms.c | 8 +- drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c | 2 +- include/drm/drmP.h | 22 +- include/drm/drm_core.h | 2 +- include/drm/drm_crtc.h | 37 +-- include/drm/drm_edid.h | 6 + 64 files changed, 2431 insertions(+), 1185 deletions(-) create mode 100644 drivers/gpu/drm/drm_edid_modes.h create mode 100644 drivers/gpu/drm/nouveau/nvc0_fb.c create mode 100644 drivers/gpu/drm/nouveau/nvc0_fifo.c create mode 100644 drivers/gpu/drm/nouveau/nvc0_graph.c create mode 100644 drivers/gpu/drm/nouveau/nvc0_instmem.c

1 1

← Newer
1
...
22
23
24
25
26
27
28
...
34
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel August 2010