dri-devel July 2011

dri-devel@lists.freedesktop.org

76 participants
443 discussions

[Bug 37227] New: z-buffer errors in doom3 outside the airlock
by bugzilla-daemon＠freedesktop.org 24 Jul '11

24 Jul '11

https://bugs.freedesktop.org/show_bug.cgi?id=37227 Summary: z-buffer errors in doom3 outside the airlock Product: Mesa Version: git Platform: Other OS/Version: All Status: NEW Severity: normal Priority: medium Component: Drivers/Gallium/r300 AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: aaalmosss(a)gmail.com Created an attachment (id=46742) --> (https://bugs.freedesktop.org/attachment.cgi?id=46742) z-error.png When the outer doors of the airlocks are open, the shadows inside the building are drawn on top of everything (except the HUD). Disabling shadows in the advanced settings makes them disappear. While being indoors, everything is rendered correctly AFAICT. With r300c every shadow was drawn on top of everything, but those covered the HUD as well. This might be a similar problem. -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 3

Re: [PATCH 8/8] rtc: class idr converted to ida and ida_simple_get used.
by Tejun Heo 23 Jul '11

23 Jul '11

On Fri, Jul 22, 2011 at 05:41:17PM +0100, Jonathan Cameron wrote: > This is the one use of an ida that doesn't retry on receiving -EAGAIN. > I'm assuming do so will cause no harm and may help on a rare occasion. > > Signed-off-by: Jonathan Cameron <jic23(a)cam.ac.uk> And all conversions seem correct to me. Please feel free to add Reviewed-by: Tejun Heo <tj(a)kernel.org> Thank you. -- tejun

1 0

[PATCH 1/8] ida: Simplified functions for id allocation.
by Jonathan Cameron 23 Jul '11

23 Jul '11

From: Rusty Russell <rusty(a)rustcorp.com.au> The current hyper-optimized functions are overkill if you simply want to allocate an id for a device. Create versions which use an internal lock. Thanks to Tejun for feedback. Signed-off-by: Rusty Russell <rusty(a)rustcorp.com.au> Acked-by: Tejun Heo <tj(a)kernel.org> Acked-by: Jonathan Cameron <jic23(a)cam.ac.uk> --- include/linux/idr.h | 4 +++ lib/idr.c | 67 +++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 71 insertions(+), 0 deletions(-) diff --git a/include/linux/idr.h b/include/linux/idr.h index 13a801f..255491c 100644 --- a/include/linux/idr.h +++ b/include/linux/idr.h @@ -146,6 +146,10 @@ void ida_remove(struct ida *ida, int id); void ida_destroy(struct ida *ida); void ida_init(struct ida *ida); +int ida_simple_get(struct ida *ida, unsigned int start, unsigned int end, + gfp_t gfp_mask); +void ida_simple_remove(struct ida *ida, unsigned int id); + void __init idr_init_cache(void); #endif /* __IDR_H__ */ diff --git a/lib/idr.c b/lib/idr.c index e15502e..db040ce 100644 --- a/lib/idr.c +++ b/lib/idr.c @@ -34,8 +34,10 @@ #include <linux/err.h> #include <linux/string.h> #include <linux/idr.h> +#include <linux/spinlock.h> static struct kmem_cache *idr_layer_cache; +static DEFINE_SPINLOCK(simple_ida_lock); static struct idr_layer *get_from_free_list(struct idr *idp) { @@ -926,6 +928,71 @@ void ida_destroy(struct ida *ida) EXPORT_SYMBOL(ida_destroy); /** + * ida_simple_get - get a new id. + * @ida: the (initialized) ida. + * @start: the minimum id (inclusive, < 0x8000000) + * @end: the maximum id (exclusive, < 0x8000000 or 0) + * @gfp_mask: memory allocation flags + * + * Allocates an id in the range start <= id < end, or returns -ENOSPC. + * On memory allocation failure, returns -ENOMEM. + * + * Use ida_simple_remove() to get rid of an id. + */ +int ida_simple_get(struct ida *ida, unsigned int start, unsigned int end, + gfp_t gfp_mask) +{ + int ret, id; + unsigned int max; + + BUG_ON((int)start < 0); + BUG_ON((int)end < 0); + + if (end == 0) + max = 0x80000000; + else { + BUG_ON(end < start); + max = end - 1; + } + +again: + if (!ida_pre_get(ida, gfp_mask)) + return -ENOMEM; + + spin_lock(&simple_ida_lock); + ret = ida_get_new_above(ida, start, &id); + if (!ret) { + if (id > max) { + ida_remove(ida, id); + ret = -ENOSPC; + } else { + ret = id; + } + } + spin_unlock(&simple_ida_lock); + + if (unlikely(ret == -EAGAIN)) + goto again; + + return ret; +} +EXPORT_SYMBOL(ida_simple_get); + +/** + * ida_simple_remove - remove an allocated id. + * @ida: the (initialized) ida. + * @id: the id returned by ida_simple_get. + */ +void ida_simple_remove(struct ida *ida, unsigned int id) +{ + BUG_ON((int)id < 0); + spin_lock(&simple_ida_lock); + ida_remove(ida, id); + spin_unlock(&simple_ida_lock); +} +EXPORT_SYMBOL(ida_simple_remove); + +/** * ida_init - initialize ida handle * @ida: ida handle * -- 1.7.3.4

2 1

Re: [osd-dev] [PATCH 6/8] [SCSI] osduld: use ida_simple_get to handle id.
by Boaz Harrosh 22 Jul '11

22 Jul '11

On 07/22/2011 09:41 AM, Jonathan Cameron wrote: > This does involve additional use of the spin lock in idr.c. > Is this an issue? > Actually it looks like a bug fix. I had a TODO: to add one. > Also, some error mangling was needed to keep the interface > the same. Does this matter or can we return -ENOSPC instead > of -EBUSY? > Na. -ENOSPC is just fine. All the osd Application just check for "any error" not a specific one. > Signed-off-by: Jonathan Cameron <jic23(a)cam.ac.uk> Ack-by: Boaz Harrosh <bharrosh(a)panasas.com> > --- > drivers/scsi/osd/osd_uld.c | 22 ++++++++-------------- > 1 files changed, 8 insertions(+), 14 deletions(-) > > diff --git a/drivers/scsi/osd/osd_uld.c b/drivers/scsi/osd/osd_uld.c > index b31a8e3..fa849bd 100644 > --- a/drivers/scsi/osd/osd_uld.c > +++ b/drivers/scsi/osd/osd_uld.c > @@ -387,7 +387,7 @@ static void __remove(struct device *dev) > > if (oud->disk) > put_disk(oud->disk); > - ida_remove(&osd_minor_ida, oud->minor); > + ida_simple_remove(&osd_minor_ida, oud->minor); > > kfree(oud); > } > @@ -403,18 +403,12 @@ static int osd_probe(struct device *dev) > if (scsi_device->type != TYPE_OSD) > return -ENODEV; > > - do { > - if (!ida_pre_get(&osd_minor_ida, GFP_KERNEL)) > - return -ENODEV; > - > - error = ida_get_new(&osd_minor_ida, &minor); > - } while (error == -EAGAIN); > - > - if (error) > - return error; > - if (minor >= SCSI_OSD_MAX_MINOR) { > - error = -EBUSY; > - goto err_retract_minor; > + minor = ida_simple_get(&osd_minor_ida, 0, > + SCSI_OSD_MAX_MINOR, GFP_KERNEL); > + if (minor < 0) { > + if (minor == -ENOSPC) > + return -EBUSY; Just drop the translation is fine as well > + return minor; > } > > error = -ENOMEM; > @@ -491,7 +485,7 @@ err_free_osd: > dev_set_drvdata(dev, NULL); > kfree(oud); > err_retract_minor: > - ida_remove(&osd_minor_ida, minor); > + ida_simple_remove(&osd_minor_ida, minor); > return error; > } >

1 0

[PATCH] drm/radeon/kms: use DIG*EncoderControl for DP link training on all generation
by j.glisse＠gmail.com 22 Jul '11

22 Jul '11

From: Jerome Glisse <jglisse(a)redhat.com> atombios functuion DIG*EncoderControl works more reliably (100% link training success vs 30% link training success on some monitor like HP ZR22w) for DP link training than DPEncoderService for DCE3 display block. Signed-off-by: Jerome Glisse <jglisse(a)redhat.com> --- drivers/gpu/drm/radeon/atombios_dp.c | 52 ++++++++++------------------------ 1 files changed, 15 insertions(+), 37 deletions(-) diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c index 8c0f9e3..9093e5e 100644 --- a/drivers/gpu/drm/radeon/atombios_dp.c +++ b/drivers/gpu/drm/radeon/atombios_dp.c @@ -645,32 +645,18 @@ static void radeon_dp_set_tp(struct radeon_dp_link_train_info *dp_info, int tp) { int rtp = 0; - /* set training pattern on the source */ - if (ASIC_IS_DCE4(dp_info->rdev)) { - switch (tp) { - case DP_TRAINING_PATTERN_1: - rtp = ATOM_ENCODER_CMD_DP_LINK_TRAINING_PATTERN1; - break; - case DP_TRAINING_PATTERN_2: - rtp = ATOM_ENCODER_CMD_DP_LINK_TRAINING_PATTERN2; - break; - case DP_TRAINING_PATTERN_3: - rtp = ATOM_ENCODER_CMD_DP_LINK_TRAINING_PATTERN3; - break; - } - atombios_dig_encoder_setup(dp_info->encoder, rtp, 0); - } else { - switch (tp) { - case DP_TRAINING_PATTERN_1: - rtp = 0; - break; - case DP_TRAINING_PATTERN_2: - rtp = 1; - break; - } - radeon_dp_encoder_service(dp_info->rdev, ATOM_DP_ACTION_TRAINING_PATTERN_SEL, - dp_info->dp_clock, dp_info->enc_id, rtp); + switch (tp) { + case DP_TRAINING_PATTERN_1: + rtp = ATOM_ENCODER_CMD_DP_LINK_TRAINING_PATTERN1; + break; + case DP_TRAINING_PATTERN_2: + rtp = ATOM_ENCODER_CMD_DP_LINK_TRAINING_PATTERN2; + break; + case DP_TRAINING_PATTERN_3: + rtp = ATOM_ENCODER_CMD_DP_LINK_TRAINING_PATTERN3; + break; } + atombios_dig_encoder_setup(dp_info->encoder, rtp, 0); /* enable training pattern on the sink */ radeon_write_dpcd_reg(dp_info->radeon_connector, DP_TRAINING_PATTERN_SET, tp); @@ -706,12 +692,8 @@ static int radeon_dp_link_train_init(struct radeon_dp_link_train_info *dp_info) radeon_write_dpcd_reg(dp_info->radeon_connector, DP_LINK_BW_SET, tmp); /* start training on the source */ - if (ASIC_IS_DCE4(dp_info->rdev)) - atombios_dig_encoder_setup(dp_info->encoder, - ATOM_ENCODER_CMD_DP_LINK_TRAINING_START, 0); - else - radeon_dp_encoder_service(dp_info->rdev, ATOM_DP_ACTION_TRAINING_START, - dp_info->dp_clock, dp_info->enc_id, 0); + atombios_dig_encoder_setup(dp_info->encoder, + ATOM_ENCODER_CMD_DP_LINK_TRAINING_START, 0); /* disable the training pattern on the sink */ radeon_write_dpcd_reg(dp_info->radeon_connector, @@ -731,12 +713,8 @@ static int radeon_dp_link_train_finish(struct radeon_dp_link_train_info *dp_info DP_TRAINING_PATTERN_DISABLE); /* disable the training pattern on the source */ - if (ASIC_IS_DCE4(dp_info->rdev)) - atombios_dig_encoder_setup(dp_info->encoder, - ATOM_ENCODER_CMD_DP_LINK_TRAINING_COMPLETE, 0); - else - radeon_dp_encoder_service(dp_info->rdev, ATOM_DP_ACTION_TRAINING_COMPLETE, - dp_info->dp_clock, dp_info->enc_id, 0); + atombios_dig_encoder_setup(dp_info->encoder, + ATOM_ENCODER_CMD_DP_LINK_TRAINING_COMPLETE, 0); return 0; } -- 1.7.1

2 1

[PATCH 0/8] RFC: Introduce ida_simple interfaces and use them.
by Jonathan Cameron 22 Jul '11

22 Jul '11

Hi All, Firstly sorry for the huge cc list, but this series does get about. This series came out of two separate lkml threads: https://lkml.org/lkml/2011/7/13/64 http://thread.gmane.org/gmane.linux.kernel/1148513/focus=74236 Basically Rusty and I both got annoyed with yet more instances of the same cut and past ida allocation code for the very simple case of wanting a unique id for some device. Tejun helpfully joined the two threads up. Anyhow, Rusty proposed more or less the first patch here (with a name change requested by Tejun applied.) The others patches are some of the more obvious looking cases for applying this code. My original reason was for IIO but those patches will have to wait a little while for some prior changes to merge. For reference, it saves about 40 lines there so I'm happy ;) There are a number of other cases in tree that can be cleaned up in a second series once these are sorted. There are a couple of cases in here where I've carefully mangled the error codes to keep consistent with the existing code. Please can people who know those subsystems well decide on whether the mangling is necessary. All comments welcome. I expect I've broken at least one driver doing this so please take a close look. Rusty, Tejun. I kept your sign of an ack for the first patch. Could you quickly verify that's fine? Thanks, Jonathan Jonathan Cameron (7): hwmon: convert idr to ida and use ida_simple interface. hwmon: ibmaem: convert idr to ida and use ida_simple_get [SCSI] use ida_simple_get and ida_simple remove in place of boilerplate code. drm/vmwgfx: use ida_simple_get for id allocation. [SCSI] osduld: use ida_simple_get to handle id. w1: ds2760 and ds2780, use ida for id and ida_simple_get to get it. rtc: class idr converted to ida and ida_simple_get used. Rusty Russell (1): ida: Simplified functions for id allocation. drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c | 34 +++---------- drivers/hwmon/hwmon.c | 32 +++--------- drivers/hwmon/ibmaem.c | 47 +++-------------- drivers/rtc/class.c | 32 +++--------- drivers/scsi/osd/osd_uld.c | 22 +++----- drivers/scsi/sd.c | 32 ++++-------- drivers/w1/slaves/w1_ds2760.c | 48 ++--------------- drivers/w1/slaves/w1_ds2780.c | 48 ++--------------- include/linux/idr.h | 4 ++ lib/idr.c | 67 +++++++++++++++++++++++++ 10 files changed, 134 insertions(+), 232 deletions(-) -- 1.7.3.4

1 0

[PATCH 5/8] drm/vmwgfx: use ida_simple_get for id allocation.
by Jonathan Cameron 22 Jul '11

22 Jul '11

Some messing with error codes to return 0 on out id's and match current situation. Is this necessary? Looks a touch 'interesting'. Signed-off-by: Jonathan Cameron <jic23(a)cam.ac.uk> --- drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c | 34 ++++++------------------- 1 files changed, 8 insertions(+), 26 deletions(-) diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c index ac6e0d1..d51577b 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c @@ -37,7 +37,6 @@ #include <linux/kernel.h> struct vmwgfx_gmrid_man { - spinlock_t lock; struct ida gmr_ida; uint32_t max_gmr_ids; }; @@ -49,34 +48,20 @@ static int vmw_gmrid_man_get_node(struct ttm_mem_type_manager *man, { struct vmwgfx_gmrid_man *gman = (struct vmwgfx_gmrid_man *)man->priv; - int ret; int id; mem->mm_node = NULL; - do { - if (unlikely(ida_pre_get(&gman->gmr_ida, GFP_KERNEL) == 0)) - return -ENOMEM; - - spin_lock(&gman->lock); - ret = ida_get_new(&gman->gmr_ida, &id); - - if (unlikely(ret == 0 && id >= gman->max_gmr_ids)) { - ida_remove(&gman->gmr_ida, id); - spin_unlock(&gman->lock); + id = ida_simple_get(&gman->gmr_ida, 0, gman->max_gmr_ids, GFP_KERNEL); + if (id < 0) { + if (id == -ENOSPC) return 0; - } - - spin_unlock(&gman->lock); - - } while (ret == -EAGAIN); - - if (likely(ret == 0)) { - mem->mm_node = gman; - mem->start = id; + return id; } + mem->mm_node = gman; + mem->start = id; - return ret; + return 0; } static void vmw_gmrid_man_put_node(struct ttm_mem_type_manager *man, @@ -86,9 +71,7 @@ static void vmw_gmrid_man_put_node(struct ttm_mem_type_manager *man, (struct vmwgfx_gmrid_man *)man->priv; if (mem->mm_node) { - spin_lock(&gman->lock); - ida_remove(&gman->gmr_ida, mem->start); - spin_unlock(&gman->lock); + ida_simple_remove(&gman->gmr_ida, mem->start); mem->mm_node = NULL; } } @@ -102,7 +85,6 @@ static int vmw_gmrid_man_init(struct ttm_mem_type_manager *man, if (unlikely(gman == NULL)) return -ENOMEM; - spin_lock_init(&gman->lock); ida_init(&gman->gmr_ida); gman->max_gmr_ids = p_size; man->priv = (void *) gman; -- 1.7.3.4

1 0

Re: Erroneous package power limit notification since kernel 2.6.39
by Florian Mickler 22 Jul '11

22 Jul '11

On Wed, 25 May 2011 15:40:24 +0000 (UTC) Olaf Freyer <aaron667(a)gmx.net> wrote: > Hi, > > I tried to git-bisect to find the root cause of this issue, but feel somewhat > lost after spending two evening without any usefull results. > > I started with v2.6.38 as last known good and v2.6.39-rc1 as first known bad > version. My presumption was that bisection would now just happen between those > two versions - but somehow I ended up building 2.6.38.rc3,+ 2.6.38.rc7+, > 2.6.38.rc2+ and 2.6.38.rc1+ during this process. Versions that are beyond the > starting range I intended to use. Or maybe I just haven't understood the whole > process yet. This is to be expected. It is because you are asked to test something which was based on a version before 2.6.38 but which got merged into Linus' tree after 2.6.38 and so it really ended up in Linus tree between 2.6.38 and 2.6.39 but has still the before-2.6.38-tag. I hope you did trust git-bisect and tested those versions? > > Both of my bisection attempts ended near the same set of drm/i915 changes that > resulted in non-bootable kernels. Considering I'm using a Intel(R) Core(TM) > i7-2720QM with some Intel Sandybridge Chipset graphics those might sound > somewhat plausible to someone knowing the internals, but don't help me at all. There are 2 untested commits left. But they don't seem to be relevant. Guessing on those gives me: The first bad commit could be any of: b0b544cd37c060e261afb2cf486296983fcb56da f67a559daaa0e2ba616bfe9438f202bc57bc8c72 18b2190ca5bd3f19717421b1591c79c9b0372428 6f06ce184c765fd8d50669a8d12fdd566c920859 0fc932b8ec36116bb759105ce910b0475e63112a 311bd68e024f9006db66cbadc3bd9f62fd663f4b 040484af3a4efa65786b6e107fbe74747679e17c ccab5c82759e2ace74b2e84f82d1e0eedd932571 aa9b500ddf1a6318e7cf8b1754696edddae86db9 d9b6cb568bc6eca8db88357bf8bbb92d42a91b1e 92f2584a083986c05fc811bbdf380c3fa7c12296 9a4114ffa7b6f5f4635e3745a8dc051d15d4596a 633f2ea26665d37bb3c8ae30799aa14988622653 63d7bbe9ded4146e3f78e5742b119fa1fdb52665 417ae1476de3ae9689a374d70565f41b3474641e ea0760cfc00b9e534423fdaf630d1c8ce7a5ede0 b24e71798871089da1a4ab049db2800afc1aac0c fe4402931e43e81a4129eba41d05cf8907603af5 65993d64a31844ad444694efb2d159eb9c883e49 c0c06bd244179f754d68684fd87674585a153e40 01fe9dbde19a1a27b8ee63e2d964562962e1eb78 a37f2f87edc1b6e5932becf6e51535d36b690f2a bdd92c9ad287e03a2ec52f5a89c470cd5caae1c2 I'd guess ccab5c82759e2ace74b2e84f82d1e0eedd932571 could be the cause. Can you check if the appended revert of that commit makes things disappear? > > I'm not yet 100% sure, but I think the package power limit notification events I > see in /var/log/messages might coincide with the moment of my xorg startup. (But > those are just wall clock time estimations.) I will try some bootups without X > later this evening and report back if those package power limit notification > events also trigger without X running. > > The "result" of my last git bisection attempt was: > > git bisect start > # good: [521cb40b0c44418a4fd36dc633f575813d59a43d] Linux 2.6.38 > git bisect good 521cb40b0c44418a4fd36dc633f575813d59a43d > # bad: [0ce790e7d736cedc563e1fb4e998babf5a4dbc3d] Linux 2.6.39-rc1 > git bisect bad 0ce790e7d736cedc563e1fb4e998babf5a4dbc3d > # bad: [179198373cf374f0ef793f1023c1cdd83b53674d] Merge branch 'nfs-for-2.6.39' > of git://git.linux-nfs.org/projects/trondmy/nfs-2.6 > git bisect bad 179198373cf374f0ef793f1023c1cdd83b53674d > # good: [6445ced8670f37cfc2c5e24a9de9b413dbfc788d] Merge branch 'staging-next' > of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging-2.6 > git bisect good 6445ced8670f37cfc2c5e24a9de9b413dbfc788d > # good: [d72751ede1b9bf993d7bd3377305c8e9e36a3cc4] Merge branch 'for-davem' of > ssh://master.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6 > git bisect good d72751ede1b9bf993d7bd3377305c8e9e36a3cc4 > # good: [16d8775700f1815076f879719ce14b33f50a3171] Merge branch 'for-linus' of > master.kernel.org:/home/rmk/linux-2.6-arm > git bisect good 16d8775700f1815076f879719ce14b33f50a3171 > # bad: [fd34b0dee4d237ce9332cc62b03adebfe4fa9f9d] Merge > git://git.kernel.org/pub/scm/linux/kernel/git/lethal/fbdev-2.6 > git bisect bad fd34b0dee4d237ce9332cc62b03adebfe4fa9f9d > # good: [bd35fe5a7930bf83ed56422ea4e4b6471ee6f739] drm/nouveau: fix > __nouveau_fence_wait performance > git bisect good bd35fe5a7930bf83ed56422ea4e4b6471ee6f739 > # bad: [942b0e95c34f1ba432d08e1c0288ed032d32c3b2] drm/radeon/kms: fix typo in > atom overscan setup > git bisect bad 942b0e95c34f1ba432d08e1c0288ed032d32c3b2 > # bad: [308977ac03117706c342099a40919b3da2667cce] drm/i915: Use DEBUG_KMS for > the self-refresh watermarks > git bisect bad 308977ac03117706c342099a40919b3da2667cce > # skip: [417ae1476de3ae9689a374d70565f41b3474641e] drm/i915: Include TLB miss > latency in g4x watermark computations > git bisect skip 417ae1476de3ae9689a374d70565f41b3474641e > # skip: [b0b544cd37c060e261afb2cf486296983fcb56da] drm/i915: Use PM QoS to > prevent C-State starvation of gen3 GPU > git bisect skip b0b544cd37c060e261afb2cf486296983fcb56da > # skip: [fe4402931e43e81a4129eba41d05cf8907603af5] Merge branch > 'drm-intel-fixes' into drm-intel-next > git bisect skip fe4402931e43e81a4129eba41d05cf8907603af5 > # bad: [987a709e1589cf10e250e04ce9df910b735d4f60] drm/i915: remove now > unnecessary delays in eDP panel power sequencing > git bisect bad 987a709e1589cf10e250e04ce9df910b735d4f60 > # skip: [18b2190ca5bd3f19717421b1591c79c9b0372428] drm/i915: allow 945 to > control self refresh (CxSR) automatically > git bisect skip 18b2190ca5bd3f19717421b1591c79c9b0372428 > # skip: [6f06ce184c765fd8d50669a8d12fdd566c920859] drm/i915: set phase sync > pointer override enable before setting phase sync pointer > git bisect skip 6f06ce184c765fd8d50669a8d12fdd566c920859 > # skip: [b24e71798871089da1a4ab049db2800afc1aac0c] drm/i915: add pipe/plane > enable/disable functions > git bisect skip b24e71798871089da1a4ab049db2800afc1aac0c > # skip: [f67a559daaa0e2ba616bfe9438f202bc57bc8c72] drm/i915: skip FDI & PCH > enabling for DP_A > git bisect skip f67a559daaa0e2ba616bfe9438f202bc57bc8c72 > # skip: [ea0760cfc00b9e534423fdaf630d1c8ce7a5ede0] drm/i915: add panel lock > assertion function > git bisect skip ea0760cfc00b9e534423fdaf630d1c8ce7a5ede0 > # skip: [ccab5c82759e2ace74b2e84f82d1e0eedd932571] drm/i915: tune Sandy Bridge > DRPS constants > git bisect skip ccab5c82759e2ace74b2e84f82d1e0eedd932571 > # skip: [a37f2f87edc1b6e5932becf6e51535d36b690f2a] drm/i915: Remove unused code: > i915_enable_interrupt() > git bisect skip a37f2f87edc1b6e5932becf6e51535d36b690f2a > # skip: [aa9b500ddf1a6318e7cf8b1754696edddae86db9] drm/i915: Honour LVDS sync > polarity from EDID > git bisect skip aa9b500ddf1a6318e7cf8b1754696edddae86db9 > # skip: [c0c06bd244179f754d68684fd87674585a153e40] drm/i915/ringbuffer: Kill an > annoyingly frequent debug message > git bisect skip c0c06bd244179f754d68684fd87674585a153e40 > # skip: [311bd68e024f9006db66cbadc3bd9f62fd663f4b] drm/i915: Trivial sparse fixes > git bisect skip 311bd68e024f9006db66cbadc3bd9f62fd663f4b > # skip: [63d7bbe9ded4146e3f78e5742b119fa1fdb52665] drm/i915: add PLL > enable/disable functions > git bisect skip 63d7bbe9ded4146e3f78e5742b119fa1fdb52665 > # skip: [0fc932b8ec36116bb759105ce910b0475e63112a] drm/i915: factor out FDI > disable and add FDI assertions > git bisect skip 0fc932b8ec36116bb759105ce910b0475e63112a > # bad: [bdd92c9ad287e03a2ec52f5a89c470cd5caae1c2] Merge branch 'drm-intel-fixes' > into drm-intel-next > git bisect bad bdd92c9ad287e03a2ec52f5a89c470cd5caae1c2 > # skip: [040484af3a4efa65786b6e107fbe74747679e17c] drm/i915: add transcoder > enable/disable functions > git bisect skip 040484af3a4efa65786b6e107fbe74747679e17c > # skip: [633f2ea26665d37bb3c8ae30799aa14988622653] drm/i915: Disable SSC for > outputs other than LVDS or DP > git bisect skip 633f2ea26665d37bb3c8ae30799aa14988622653 > # skip: [d9b6cb568bc6eca8db88357bf8bbb92d42a91b1e] drm/i915: assert panel is > unlocked before writing transcoder timing regs > git bisect skip d9b6cb568bc6eca8db88357bf8bbb92d42a91b1e > # skip: [9a4114ffa7b6f5f4635e3745a8dc051d15d4596a] drm/i915/bios: Change default > clock source on PineView to use SSC > git bisect skip 9a4114ffa7b6f5f4635e3745a8dc051d15d4596a > # skip: [92f2584a083986c05fc811bbdf380c3fa7c12296] drm/i915: add PCH DPLL > enable/disable functions > git bisect skip 92f2584a083986c05fc811bbdf380c3fa7c12296 > # skip: [65993d64a31844ad444694efb2d159eb9c883e49] drm/i915: don't enable plane, > pipe and PLL prematurely > git bisect skip 65993d64a31844ad444694efb2d159eb9c883e49 > # skip: [01fe9dbde19a1a27b8ee63e2d964562962e1eb78] drm/i915: Use ACPI OpRegion > to determine lid status > git bisect skip 01fe9dbde19a1a27b8ee63e2d964562962e1eb78 > > Regards > Olaf Freyer > ------>8---------->8--------------->8-------------->8------------ Revert "drm/i915: tune Sandy Bridge DRPS constants" This reverts parts of commit ccab5c82759e2ace74b2e84f82d1e0eedd932571. It does not touch debugfs.c Conflicts: drivers/gpu/drm/i915/i915_debugfs.c diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 5d5def7..a245742 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -3378,28 +3378,15 @@ #define GEN6_RP_DOWN_TIMEOUT 0xA010 #define GEN6_RP_INTERRUPT_LIMITS 0xA014 #define GEN6_RPSTAT1 0xA01C -#define GEN6_CAGF_SHIFT 8 -#define GEN6_CAGF_MASK (0x7f << GEN6_CAGF_SHIFT) #define GEN6_RP_CONTROL 0xA024 #define GEN6_RP_MEDIA_TURBO (1<<11) #define GEN6_RP_USE_NORMAL_FREQ (1<<9) #define GEN6_RP_MEDIA_IS_GFX (1<<8) #define GEN6_RP_ENABLE (1<<7) -#define GEN6_RP_UP_IDLE_MIN (0x1<<3) -#define GEN6_RP_UP_BUSY_AVG (0x2<<3) -#define GEN6_RP_UP_BUSY_CONT (0x4<<3) -#define GEN6_RP_DOWN_IDLE_CONT (0x1<<0) +#define GEN6_RP_UP_BUSY_MAX (0x2<<3) +#define GEN6_RP_DOWN_BUSY_MIN (0x2<<0) #define GEN6_RP_UP_THRESHOLD 0xA02C #define GEN6_RP_DOWN_THRESHOLD 0xA030 -#define GEN6_RP_CUR_UP_EI 0xA050 -#define GEN6_CURICONT_MASK 0xffffff -#define GEN6_RP_CUR_UP 0xA054 -#define GEN6_CURBSYTAVG_MASK 0xffffff -#define GEN6_RP_PREV_UP 0xA058 -#define GEN6_RP_CUR_DOWN_EI 0xA05C -#define GEN6_CURIAVG_MASK 0xffffff -#define GEN6_RP_CUR_DOWN 0xA060 -#define GEN6_RP_PREV_DOWN 0xA064 #define GEN6_RP_UP_EI 0xA068 #define GEN6_RP_DOWN_EI 0xA06C #define GEN6_RP_IDLE_HYSTERSIS 0xA070 diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index aa43e7b..be9890e 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -7101,18 +7101,18 @@ void gen6_enable_rps(struct drm_i915_private *dev_priv) I915_WRITE(GEN6_RP_INTERRUPT_LIMITS, 18 << 24 | 6 << 16); - I915_WRITE(GEN6_RP_UP_THRESHOLD, 10000); - I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 1000000); + I915_WRITE(GEN6_RP_UP_THRESHOLD, 90000); + I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 100000); I915_WRITE(GEN6_RP_UP_EI, 100000); - I915_WRITE(GEN6_RP_DOWN_EI, 5000000); + I915_WRITE(GEN6_RP_DOWN_EI, 300000); I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10); I915_WRITE(GEN6_RP_CONTROL, GEN6_RP_MEDIA_TURBO | GEN6_RP_USE_NORMAL_FREQ | GEN6_RP_MEDIA_IS_GFX | GEN6_RP_ENABLE | - GEN6_RP_UP_BUSY_AVG | - GEN6_RP_DOWN_IDLE_CONT); + GEN6_RP_UP_BUSY_MAX | + GEN6_RP_DOWN_BUSY_MIN); if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0, 500))

3 8

[PATCH] drm/i915: Fix typo in DRM_I915_OVERLAY_PUT_IMAGE ioctl define
by Ole Henrik Jahren 22 Jul '11

22 Jul '11

Because of a typo, calling ioctl with DRM_IOCTL_I915_OVERLAY_PUT_IMAGE is broken if the macro is used directly. When using libdrm the bug is not hit, since libdrm handles the ioctl encoding internally. The typo also leads to the .cmd and .cmd_drv fields of the drm_ioctl structure for DRM_I915_OVERLAY_PUT_IMAGE having inconsistent content. Signed-off-by: Ole Henrik Jahren <olehenja(a)alumni.ntnu.no> Acked-by: Daniel Vetter <daniel.vetter(a)ffwll.ch> --- include/drm/i915_drm.h | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/include/drm/i915_drm.h b/include/drm/i915_drm.h index c4d6dbf..28c0d11 100644 --- a/include/drm/i915_drm.h +++ b/include/drm/i915_drm.h @@ -237,7 +237,7 @@ typedef struct _drm_i915_sarea { #define DRM_IOCTL_I915_GEM_GET_APERTURE DRM_IOR (DRM_COMMAND_BASE + DRM_I915_GEM_GET_APERTURE, struct drm_i915_gem_get_aperture) #define DRM_IOCTL_I915_GET_PIPE_FROM_CRTC_ID DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GET_PIPE_FROM_CRTC_ID, struct drm_i915_get_pipe_from_crtc_id) #define DRM_IOCTL_I915_GEM_MADVISE DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_MADVISE, struct drm_i915_gem_madvise) -#define DRM_IOCTL_I915_OVERLAY_PUT_IMAGE DRM_IOW(DRM_COMMAND_BASE + DRM_IOCTL_I915_OVERLAY_ATTRS, struct drm_intel_overlay_put_image) +#define DRM_IOCTL_I915_OVERLAY_PUT_IMAGE DRM_IOW(DRM_COMMAND_BASE + DRM_I915_OVERLAY_PUT_IMAGE, struct drm_intel_overlay_put_image) #define DRM_IOCTL_I915_OVERLAY_ATTRS DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_OVERLAY_ATTRS, struct drm_intel_overlay_attrs) /* Allow drivers to submit batchbuffers directly to hardware, relying -- 1.7.6

1 0

[RFC] Updated DRM plane handling patches
by Jesse Barnes 21 Jul '11

21 Jul '11

This patchset updates the previous one, incorporating the feedback I received: 1) uses the v4l fourcc codes to communicate pixel format 2) adds a new addfb ioctl that takes a format 3) adds working SNB support for the new code Comments welcome. I'll be pushing intel-gpu-tools testdisplay support for this shortly if people want to play around with it. Next step is to incorporate it into X and Wayland to see if the API is rich enough for our needs. Thanks, Jesse

7 19

← Newer
1
...
33
34
35
36
37
38
39
...
45
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel July 2011