dri-devel September 2010

dri-devel@lists.freedesktop.org

71 participants
367 discussions

[Bug 30025] New: r600 kms, massave slowdown with latest drm-radeon-testing
by bugzilla-daemon＠freedesktop.org 05 Sep '10

05 Sep '10

https://bugs.freedesktop.org/show_bug.cgi?id=30025 Summary: r600 kms, massave slowdown with latest drm-radeon-testing Product: DRI Version: XOrg CVS Platform: x86-64 (AMD64) OS/Version: Linux (All) Status: NEW Severity: normal Priority: medium Component: DRM/Radeon AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: chainsawbike(a)gmail.com ive bisected to this commit: http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=commitdiff;… odd thing i noted: if i run compiz with vblank_mode=0 and run glxgears it reports a frame-rate of about 1500 fps ( being the same as i get when running the previous revision kernel) but its still looks as slow as molasses if i unset vblank_mode then run it it reports about 5 fps but looks the same as before -- 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

[Bug 28402] Kernel 2.6.34 freezes randomly
by bugzilla-daemon＠freedesktop.org 05 Sep '10

05 Sep '10

https://bugs.freedesktop.org/show_bug.cgi?id=28402 --- Comment #36 from Lukas Schneiderbauer <lukas.schneiderbauer(a)gmail.com> 2010-09-05 05:11:36 PDT --- (In reply to comment #34) > If you could now confirm 44ca7478d46aaad488d916f7262253e000ee60f9 as not > causing the freeze we have finally isolated the exact commit that is causing > the freezes. Confirmed. 44ca7478d46aaad488d916f7262253e000ee60f9 is in use now for ~ 3 days and no freeze has occured. That would support your assumption. And I think, it's a very likely one. -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 0

[Bug 30022] New: Crash while launching supertuxkart (svn), "Too many temporaries"
by bugzilla-daemon＠freedesktop.org 05 Sep '10

05 Sep '10

https://bugs.freedesktop.org/show_bug.cgi?id=30022 Summary: Crash while launching supertuxkart (svn), "Too many temporaries" Product: Mesa Version: git Platform: x86 (IA32) OS/Version: Linux (All) Status: NEW Severity: normal Priority: medium Component: Drivers/DRI/r300 AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: gsr.bugs(a)infernal-iceberg.com Launching supertuxkart (svn 5892) generates the following when run with RADEON_DEBUG=vp. ---8<--- r300: Initial vertex program VERT DCL IN[0] DCL IN[1] DCL OUT[0], POSITION DCL OUT[1], COLOR 0: MOV OUT[0], IN[0] 1: MOV OUT[1], IN[1] 2: END Vertex Program: before compilation # Radeon Compiler Program 0: MOV temp[0], input[0]; 1: MOV output[1], input[1]; 2: MOV output[0], temp[0]; 3: MOV output[2], temp[0]; Vertex Program: after 'transform loops' # Radeon Compiler Program 0: MOV temp[0], input[0]; 1: MOV output[1], input[1]; 2: MOV output[0], temp[0]; 3: MOV output[2], temp[0]; Vertex Program: after 'emulate branches' # Radeon Compiler Program 0: MOV temp[0], input[0]; 1: MOV output[1], input[1]; 2: MOV output[0], temp[0]; 3: MOV output[2], temp[0]; Vertex Program: after 'emulate negative addressing' # Radeon Compiler Program 0: MOV temp[0], input[0]; 1: MOV output[1], input[1]; 2: MOV output[0], temp[0]; 3: MOV output[2], temp[0]; Vertex Program: after 'native rewrite' # Radeon Compiler Program 0: MOV temp[0], input[0]; 1: MOV output[1], input[1]; 2: MOV output[0], temp[0]; 3: MOV output[2], temp[0]; Vertex Program: after 'emulate modifiers' # Radeon Compiler Program 0: MOV temp[0], input[0]; 1: MOV output[1], input[1]; 2: MOV output[0], temp[0]; 3: MOV output[2], temp[0]; Vertex Program: after 'deadcode' # Radeon Compiler Program 0: MOV temp[0], input[0]; 1: MOV output[1], input[1]; 2: MOV output[0], temp[0]; 3: MOV output[2], temp[0]; Vertex Program: after 'dataflow optimize' # Radeon Compiler Program 0: MOV output[1], input[1]; 1: MOV output[0], input[0]; 2: MOV output[2], input[0]; Vertex Program: after 'source conflict resolve' # Radeon Compiler Program 0: MOV output[1], input[1]; 1: MOV output[0], input[0]; 2: MOV output[2], input[0]; Vertex Program: after 'dataflow swizzles' # Radeon Compiler Program 0: MOV output[1], input[1]; 1: MOV output[0], input[0]; 2: MOV output[2], input[0]; Vertex Program: after 'register allocation' # Radeon Compiler Program 0: MOV output[1], input[1]; 1: MOV output[0], input[0]; 2: MOV output[2], input[0]; Vertex Program: after 'dead constants' # Radeon Compiler Program 0: MOV output[1], input[1]; 1: MOV output[0], input[0]; 2: MOV output[2], input[0]; num=140903841 max=32 r300compiler error: Too many temporaries. r300 VP: Compiler error: Too many temporaries. Using a dummy shader instead. r300: Initial vertex program VERT DCL OUT[0], POSITION IMM FLT32 { 0.0000, 1.0000, 0.0000, 0.0000} 0: MOV OUT[0], IMM[0].xxxy 1: END Vertex Program: before compilation # Radeon Compiler Program 0: MOV temp[1], temp[0].0001; 1: MOV output[0], temp[1]; 2: MOV output[1], temp[1]; Vertex Program: after 'transform loops' # Radeon Compiler Program 0: MOV temp[1], temp[0].0001; 1: MOV output[0], temp[1]; 2: MOV output[1], temp[1]; Vertex Program: after 'emulate branches' # Radeon Compiler Program 0: MOV temp[1], temp[0].0001; 1: MOV output[0], temp[1]; 2: MOV output[1], temp[1]; Vertex Program: after 'emulate negative addressing' # Radeon Compiler Program 0: MOV temp[1], temp[0].0001; 1: MOV output[0], temp[1]; 2: MOV output[1], temp[1]; Vertex Program: after 'native rewrite' # Radeon Compiler Program 0: MOV temp[1], temp[0].0001; 1: MOV output[0], temp[1]; 2: MOV output[1], temp[1]; Vertex Program: after 'emulate modifiers' # Radeon Compiler Program 0: MOV temp[1], temp[0].0001; 1: MOV output[0], temp[1]; 2: MOV output[1], temp[1]; Vertex Program: after 'deadcode' # Radeon Compiler Program 0: MOV temp[1], temp[0].0001; 1: MOV output[0], temp[1]; 2: MOV output[1], temp[1]; Vertex Program: after 'dataflow optimize' # Radeon Compiler Program 0: MOV output[0], temp[0].0001; 1: MOV output[1], temp[0].0001; Vertex Program: after 'source conflict resolve' # Radeon Compiler Program 0: MOV output[0], temp[0].0001; 1: MOV output[1], temp[0].0001; Vertex Program: after 'dataflow swizzles' # Radeon Compiler Program 0: MOV output[0], temp[0].0001; 1: MOV output[1], temp[0].0001; Vertex Program: after 'register allocation' # Radeon Compiler Program 0: MOV output[0], temp[0].0001; 1: MOV output[1], temp[0].0001; Vertex Program: after 'dead constants' # Radeon Compiler Program 0: MOV output[0], temp[0].0001; 1: MOV output[1], temp[0].0001; num=140903841 max=32 r300compiler error: Too many temporaries. r300 VP: Compiler error: Too many temporaries. Using a dummy shader instead. r300 VP: Cannot compile the dummy shader! Giving up... Program received signal SIGABRT, Aborted. 0xb7fe2424 in __kernel_vsyscall () (gdb) bt full #0 0xb7fe2424 in __kernel_vsyscall () No symbol table info available. #1 0xb7cbc751 in *__GI_raise (sig=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64 resultvar = <value optimized out> pid = -1210236940 selftid = 24065 #2 0xb7cbfb82 in *__GI_abort () at abort.c:92 act = {__sigaction_handler = {sa_handler = 0xb71047a6, sa_sigaction = 0xb71047a6}, sa_mask = {__val = {1, 1, 3068832739, 3221218148, 3071297470, 1, 1, 3068900000, 0, 3071297488, 1, 1, 3068848457, 0, 3071297508, 1, 1, 3068904887, 140920680, 3071297523, 0, 1, 3221218136, 55, 3071293612, 3084185523, 3084730356, 3083833012, 2, 3071293612, 55, 3084731744}}, sa_flags = -1, sa_restorer = 0xffffffff} sigs = {__val = {32, 0 <repeats 31 times>}} #3 0xb6ea7a01 in r300_translate_vertex_shader () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #4 0xb6ea77d8 in r300_dummy_vertex_shader () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #5 0xb6ea7a21 in r300_translate_vertex_shader () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #6 0xb6ea47e1 in r300_create_vs_state () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #7 0xb70b54a3 in ureg_create_shader () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #8 0xb70e4420 in ureg_create_shader_and_destroy () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #9 0xb70e496e in util_make_vertex_passthrough_shader () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #10 0xb70b8e32 in util_blitter_create () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #11 0xb6e95015 in r300_create_context () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #12 0xb6f9210f in create_context () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #13 0xb6f92209 in st_api_create_context () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #14 0xb6e7bc66 in dri_create_context () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #15 0xb6e7ba50 in dri2_create_context () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #16 0xb6e775d1 in dri2CreateNewContextForAPI () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #17 0xb6e77620 in dri2CreateNewContext () from /tmp/gallium/lib/gallium//r300_dri.so No symbol table info available. #18 0xb7fa42cd in ?? () from /usr/lib/libGL.so.1 No symbol table info available. #19 0xb7f7ce3b in ?? () from /usr/lib/libGL.so.1 No symbol table info available. #20 0xb7f7d2de in glXCreateContext () from /usr/lib/libGL.so.1 No symbol table info available. #21 0x081c01d8 in irr::CIrrDeviceLinux::createWindow() () No symbol table info available. #22 0x081c615c in irr::CIrrDeviceLinux::CIrrDeviceLinux(irr::SIrrlichtCreationParameters const&) () No symbol table info available. #23 0x081be07a in createDeviceEx () No symbol table info available. #24 0x081be180 in createDevice () No symbol table info available. #25 0x080854b7 in IrrDriver::initDevice (this=0x861af00) at graphics/irr_driver.cpp:164 bits = 32 type = irr::video::EDT_OPENGL driver_type = 0 firstTime = false numDrivers = 5 material2D = <value optimized out> #26 0x0808688a in IrrDriver (this=0x861af00) at graphics/irr_driver.cpp:57 No locals. #27 0x08050be4 in initRest () at main.cpp:527 No locals. #28 0x08051f63 in main (argc=1, argv=0xbffff234) at main.cpp:611 materials_file = {static npos = 4294967295, _M_dataplus = {<std::allocator<char>> = {<__gnu_cxx::new_allocator<char>> = {<No data fields>}, <No data fields>}, _M_p = 0xbffff148 "X\361\377\277\244\332\004\b\260\020\377\267\230\245]\b\210\361\377\277\t\255S\b\004Cݷ\364?ݷ\220\361\377\277\364?ݷ"}} device = <value optimized out> driver = 0x85de520 --->8--- git bisect suggests: ---8<--- d2f4ceaa47695f9dc203c28b39d332eea8a115b2 is the first bad commit commit d2f4ceaa47695f9dc203c28b39d332eea8a115b2 Author: Marek Olšák <maraeo(a)gmail.com> Date: Wed Sep 1 08:10:32 2010 +0200 r300/compiler: compute the final number of temporaries during translation And not during the register allocation, which may be skipped for debugging purposes. Also the predicate register is now added to the number of temps. :040000 040000 4c1e1bbcb6034c5f25621048602e3b11a918ee88 e89a3bc7b5ab99cc566e78e692a166e887212e57 M src --->8--- -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 7

[PATCH] drm/radeon/kms/r6xx+: use new style fencing (v3)
by Alex Deucher 04 Sep '10

04 Sep '10

On r6xx+ a newer fence mechanism was implemented to replace the old wait_until plus scratch regs setup. A single EOP event will flush the destination caches, write a fence value, and generate an interrupt. This is the recommended fence mechanism on r6xx+ asics. This requires my previous writeback patch. v2: fix typo that enabled event fence checking on all asics rather than just r6xx+. v3: properly enable EOP interrupts Should fix: https://bugs.freedesktop.org/show_bug.cgi?id=29972 Signed-off-by: Alex Deucher <alexdeucher(a)gmail.com> --- drivers/gpu/drm/radeon/evergreen.c | 2 + drivers/gpu/drm/radeon/r600.c | 42 ++++++++++++++++++++----------- drivers/gpu/drm/radeon/r600d.h | 21 ++++++++++++++++ drivers/gpu/drm/radeon/radeon.h | 2 + drivers/gpu/drm/radeon/radeon_device.c | 8 +++++- drivers/gpu/drm/radeon/radeon_fence.c | 6 ++++- 6 files changed, 64 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index e439f9e..0354596 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -1583,6 +1583,7 @@ int evergreen_irq_set(struct radeon_device *rdev) if (rdev->irq.sw_int) { DRM_DEBUG("evergreen_irq_set: sw int\n"); cp_int_cntl |= RB_INT_ENABLE; + cp_int_cntl |= TIME_STAMP_INT_ENABLE; } if (rdev->irq.crtc_vblank_int[0]) { DRM_DEBUG("evergreen_irq_set: vblank 0\n"); @@ -2001,6 +2002,7 @@ restart_ih: break; case 181: /* CP EOP event */ DRM_DEBUG("IH: CP EOP\n"); + radeon_fence_process(rdev); break; case 233: /* GUI IDLE */ DRM_DEBUG("IH: CP EOP\n"); diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index dc953e3..41ca173 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -2279,21 +2279,31 @@ int r600_ring_test(struct radeon_device *rdev) void r600_fence_ring_emit(struct radeon_device *rdev, struct radeon_fence *fence) { - /* Also consider EVENT_WRITE_EOP. it handles the interrupts + timestamps + events */ - - radeon_ring_write(rdev, PACKET3(PACKET3_EVENT_WRITE, 0)); - radeon_ring_write(rdev, CACHE_FLUSH_AND_INV_EVENT); - /* wait for 3D idle clean */ - radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONFIG_REG, 1)); - radeon_ring_write(rdev, (WAIT_UNTIL - PACKET3_SET_CONFIG_REG_OFFSET) >> 2); - radeon_ring_write(rdev, WAIT_3D_IDLE_bit | WAIT_3D_IDLECLEAN_bit); - /* Emit fence sequence & fire IRQ */ - radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONFIG_REG, 1)); - radeon_ring_write(rdev, ((rdev->fence_drv.scratch_reg - PACKET3_SET_CONFIG_REG_OFFSET) >> 2)); - radeon_ring_write(rdev, fence->seq); - /* CP_INTERRUPT packet 3 no longer exists, use packet 0 */ - radeon_ring_write(rdev, PACKET0(CP_INT_STATUS, 0)); - radeon_ring_write(rdev, RB_INT_STAT); + if (rdev->wb.use_event) { + u64 addr = rdev->wb.gpu_addr + R600_WB_EVENT_OFFSET + + (u64)(rdev->fence_drv.scratch_reg - rdev->scratch.reg_base); + /* EVENT_WRITE_EOP - flush caches, send int */ + radeon_ring_write(rdev, PACKET3(PACKET3_EVENT_WRITE_EOP, 4)); + radeon_ring_write(rdev, EVENT_TYPE(CACHE_FLUSH_AND_INV_EVENT_TS) | EVENT_INDEX(5)); + radeon_ring_write(rdev, addr & 0xffffffff); + radeon_ring_write(rdev, (upper_32_bits(addr) & 0xff) | DATA_SEL(1) | INT_SEL(2)); + radeon_ring_write(rdev, fence->seq); + radeon_ring_write(rdev, 0); + } else { + radeon_ring_write(rdev, PACKET3(PACKET3_EVENT_WRITE, 0)); + radeon_ring_write(rdev, EVENT_TYPE(CACHE_FLUSH_AND_INV_EVENT) | EVENT_INDEX(0)); + /* wait for 3D idle clean */ + radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + radeon_ring_write(rdev, (WAIT_UNTIL - PACKET3_SET_CONFIG_REG_OFFSET) >> 2); + radeon_ring_write(rdev, WAIT_3D_IDLE_bit | WAIT_3D_IDLECLEAN_bit); + /* Emit fence sequence & fire IRQ */ + radeon_ring_write(rdev, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + radeon_ring_write(rdev, ((rdev->fence_drv.scratch_reg - PACKET3_SET_CONFIG_REG_OFFSET) >> 2)); + radeon_ring_write(rdev, fence->seq); + /* CP_INTERRUPT packet 3 no longer exists, use packet 0 */ + radeon_ring_write(rdev, PACKET0(CP_INT_STATUS, 0)); + radeon_ring_write(rdev, RB_INT_STAT); + } } int r600_copy_blit(struct radeon_device *rdev, @@ -3012,6 +3022,7 @@ int r600_irq_set(struct radeon_device *rdev) if (rdev->irq.sw_int) { DRM_DEBUG("r600_irq_set: sw int\n"); cp_int_cntl |= RB_INT_ENABLE; + cp_int_cntl |= TIME_STAMP_INT_ENABLE; } if (rdev->irq.crtc_vblank_int[0]) { DRM_DEBUG("r600_irq_set: vblank 0\n"); @@ -3377,6 +3388,7 @@ restart_ih: break; case 181: /* CP EOP event */ DRM_DEBUG("IH: CP EOP\n"); + radeon_fence_process(rdev); break; case 233: /* GUI IDLE */ DRM_DEBUG("IH: CP EOP\n"); diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h index 858a192..966a793 100644 --- a/drivers/gpu/drm/radeon/r600d.h +++ b/drivers/gpu/drm/radeon/r600d.h @@ -474,6 +474,7 @@ #define VGT_VERTEX_REUSE_BLOCK_CNTL 0x28C58 #define VTX_REUSE_DEPTH_MASK 0x000000FF #define VGT_EVENT_INITIATOR 0x28a90 +# define CACHE_FLUSH_AND_INV_EVENT_TS (0x14 << 0) # define CACHE_FLUSH_AND_INV_EVENT (0x16 << 0) #define VM_CONTEXT0_CNTL 0x1410 @@ -775,7 +776,27 @@ #define PACKET3_ME_INITIALIZE_DEVICE_ID(x) ((x) << 16) #define PACKET3_COND_WRITE 0x45 #define PACKET3_EVENT_WRITE 0x46 +#define EVENT_TYPE(x) ((x) << 0) +#define EVENT_INDEX(x) ((x) << 8) + /* 0 - any non-TS event + * 1 - ZPASS_DONE + * 2 - SAMPLE_PIPELINESTAT + * 3 - SAMPLE_STREAMOUTSTAT* + * 4 - *S_PARTIAL_FLUSH + * 5 - TS events + */ #define PACKET3_EVENT_WRITE_EOP 0x47 +#define DATA_SEL(x) ((x) << 29) + /* 0 - discard + * 1 - send low 32bit data + * 2 - send 64bit data + * 3 - send 64bit counter value + */ +#define INT_SEL(x) ((x) << 24) + /* 0 - none + * 1 - interrupt only (DATA_SEL = 0) + * 2 - interrupt when data write is confirmed + */ #define PACKET3_ONE_REG_WRITE 0x57 #define PACKET3_SET_CONFIG_REG 0x68 #define PACKET3_SET_CONFIG_REG_OFFSET 0x00008000 diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 4e10938..2bfae56 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -595,11 +595,13 @@ struct radeon_wb { volatile uint32_t *wb; uint64_t gpu_addr; bool enabled; + bool use_event; }; #define RADEON_WB_SCRATCH_OFFSET 0 #define RADEON_WB_CP_RPTR_OFFSET 1024 #define R600_WB_IH_WPTR_OFFSET 2048 +#define R600_WB_EVENT_OFFSET 3072 /** * struct radeon_pm - power management datas diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index cfc162d..8adfedf 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -208,6 +208,8 @@ int radeon_wb_init(struct radeon_device *rdev) return r; } + /* disable event_write fences */ + rdev->wb.use_event = false; /* disabled via module param */ if (radeon_no_wb == 1) rdev->wb.enabled = false; @@ -215,8 +217,12 @@ int radeon_wb_init(struct radeon_device *rdev) /* often unreliable on AGP */ if (rdev->flags & RADEON_IS_AGP) { rdev->wb.enabled = false; - } else + } else { rdev->wb.enabled = true; + /* event_write fences are only available on r600+ */ + if (rdev->family >= CHIP_R600) + rdev->wb.use_event = true; + } } dev_info(rdev->dev, "WB %sabled\n", rdev->wb.enabled ? "en" : "dis"); diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 698a7ed..216392d 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -73,7 +73,11 @@ static bool radeon_fence_poll_locked(struct radeon_device *rdev) unsigned long cjiffies; if (rdev->wb.enabled) { - u32 scratch_index = RADEON_WB_SCRATCH_OFFSET + rdev->fence_drv.scratch_reg - rdev->scratch.reg_base; + u32 scratch_index; + if (rdev->wb.use_event) + scratch_index = R600_WB_EVENT_OFFSET + rdev->fence_drv.scratch_reg - rdev->scratch.reg_base; + else + scratch_index = RADEON_WB_SCRATCH_OFFSET + rdev->fence_drv.scratch_reg - rdev->scratch.reg_base; seq = rdev->wb.wb[scratch_index/4]; } else seq = RREG32(rdev->fence_drv.scratch_reg); -- 1.7.1.1

1 0

[Bug 29754] New: [r300g, glsl] Heroes of Newerth lockup GPU
by bugzilla-daemon＠freedesktop.org 03 Sep '10

03 Sep '10

https://bugs.freedesktop.org/show_bug.cgi?id=29754 Summary: [r300g, glsl] Heroes of Newerth lockup GPU Product: Mesa Version: git Platform: x86 (IA32) OS/Version: Linux (All) Status: NEW Severity: major Priority: medium Component: Drivers/DRI/r300 AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: drakkk(a)centrum.cz Created an attachment (id=38099) --> (https://bugs.freedesktop.org/attachment.cgi?id=38099) dmesg With latest mesa git when starting HoN it lockup my GPU. This behavior was introduced with glsl2 merge. radeon: The kernel rejected CS, see dmesg for more information. this is printed into terminal many times, no more helpful info, dmesg attached. To reproduce this just download the game and run it. It isn't free but it doesn't matter since you won't get as far as the login screen. BTW I'm not sure if the component should be mesa core or Drivers/DRI/r300, feel free to change it if it is wrong. My system: Kernel: 2.6.35 mesa: c907b947130c884de09e48e1ecbeecc9afc9f75b GPU: rv530 -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 4

[Bug 28966] New: [r300g] Dynamic branching 3 demo does not run
by bugzilla-daemon＠freedesktop.org 03 Sep '10

03 Sep '10

https://bugs.freedesktop.org/show_bug.cgi?id=28966 Summary: [r300g] Dynamic branching 3 demo does not run Product: Mesa Version: git Platform: Other URL: http://www.humus.name/index.php?page=3D&ID=67 OS/Version: All Status: NEW Severity: normal Priority: medium Component: Drivers/DRI/r300 AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: sa(a)whiz.se I'm filing this just in case it's of interest for those hacking shaders on r300. Humus DynamicBranching3 demo does not run: Note: 'for (i ... )' body is too large/complex to unroll Fragment program using varying vars not written by vertex shader For reference, this demo does work with the i965 driver. http://www.humus.name/index.php?page=3D&ID=67 System environment: -- system architecture: 32-bit -- Linux distribution: Debian unstable -- GPU: RV570 -- Model: Asus EAX1950Pro 256MB -- Display connector: DVI -- xf86-video-ati: 801e83227a59a29eea425ea612083bbf2b536c30 -- xserver: 1.8.1.901 -- mesa: 61a26cdfdc9c75a83c0d362c973d5436fe077be4 -- drm: 6ea2bda5f5ec8f27359760ce580fdad3df0464df -- kernel: 2.6.35-rc3 -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 9

[Bug 29913] New: [regression glsl r300g] Heroes of Newerth slowdown after glsl2 merge
by bugzilla-daemon＠freedesktop.org 03 Sep '10

03 Sep '10

https://bugs.freedesktop.org/show_bug.cgi?id=29913 Summary: [regression glsl r300g] Heroes of Newerth slowdown after glsl2 merge Product: Mesa Version: git Platform: Other URL: http://www.heroesofnewerth.com/download.php OS/Version: All Status: NEW Severity: normal Priority: medium Component: Drivers/DRI/r300 AssignedTo: dri-devel(a)lists.freedesktop.org ReportedBy: sa(a)whiz.se The game Heroes of Newerth (reproducible in the menu with the client) is slow as molasses after the glsl2 merge. Both the game itself, and normal procedures in X while the game is running (e.g. moving windows around) is noticeably laggy. CPU usage is still low, so I guess something is keeping the GPU busy? This bug have not been bisected, only confirmed by trying the commit immediately before the merge (15a3b42e) which is fine, and the commit with the glsl2 merge (6c03c57) where the problem is apparent. This is most likely the same slowdown I noticed in the game Savage 2 (bug 28517) but I'm filing this separately to be sure. System environment: -- system architecture: 32-bit -- Linux distribution: Debian unstable -- GPU: RV570 -- Model: Asus EAX1950Pro 256MB -- Display connector: DVI -- xf86-video-ati: fd686668289258ffaf6b81057545e50612aac6a8 -- xserver: 1.8.99.904 (1.9.0 RC 5) -- mesa: 99f3c9caa39fbe9dfa7561c919202395720e9472 -- drm: b61e81a191d3a5c269c5f7c40199aebc9ebc034c -- kernel: 2.6.35 -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 1

[Bug 26514] improper ENOMEM errors with Radeon 6.12.4
by bugzilla-daemon＠freedesktop.org 03 Sep '10

03 Sep '10

https://bugs.freedesktop.org/show_bug.cgi?id=26514 --- Comment #5 from Kenneth Koehler <kenneth.koehler(a)uc.edu> 2010-09-03 13:25:02 PDT --- Sorry - previous comment neglected to mention upgrade xf86-video-ati to 6.12.7 -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 0

[PATCH] drm/radeon/kms: fix tv-out on avivo asics
by Alex Deucher 03 Sep '10

03 Sep '10

digital underscan support regressed tv-out. fixes: https://bugs.freedesktop.org/show_bug.cgi?id=29985 Signed-off-by: Alex Deucher <alexdeucher(a)gmail.com> --- drivers/gpu/drm/radeon/atombios_crtc.c | 26 +++++++++++++++++++++++--- 1 files changed, 23 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index 71b3150..464a81a 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -332,6 +332,11 @@ static void atombios_crtc_set_timing(struct drm_crtc *crtc, args.usV_SyncWidth = cpu_to_le16(mode->crtc_vsync_end - mode->crtc_vsync_start); + args.ucOverscanRight = radeon_crtc->h_border; + args.ucOverscanLeft = radeon_crtc->h_border; + args.ucOverscanBottom = radeon_crtc->v_border; + args.ucOverscanTop = radeon_crtc->v_border; + if (mode->flags & DRM_MODE_FLAG_NVSYNC) misc |= ATOM_VSYNC_POLARITY; if (mode->flags & DRM_MODE_FLAG_NHSYNC) @@ -1211,8 +1216,18 @@ int atombios_crtc_mode_set(struct drm_crtc *crtc, struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc); struct drm_device *dev = crtc->dev; struct radeon_device *rdev = dev->dev_private; + struct drm_encoder *encoder; + bool is_tvcv = false; - /* TODO color tiling */ + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + /* find tv std */ + if (encoder->crtc == crtc) { + struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); + if (radeon_encoder->active_device & + (ATOM_DEVICE_TV_SUPPORT | ATOM_DEVICE_CV_SUPPORT)) + is_tvcv = true; + } + } atombios_disable_ss(crtc); /* always set DCPLL */ @@ -1221,9 +1236,14 @@ int atombios_crtc_mode_set(struct drm_crtc *crtc, atombios_crtc_set_pll(crtc, adjusted_mode); atombios_enable_ss(crtc); - if (ASIC_IS_AVIVO(rdev)) + if (ASIC_IS_DCE4(rdev)) atombios_set_crtc_dtd_timing(crtc, adjusted_mode); - else { + else if (ASIC_IS_AVIVO(rdev)) { + if (is_tvcv) + atombios_crtc_set_timing(crtc, adjusted_mode); + else + atombios_set_crtc_dtd_timing(crtc, adjusted_mode); + } else { atombios_crtc_set_timing(crtc, adjusted_mode); if (radeon_crtc->crtc_id == 0) atombios_set_crtc_dtd_timing(crtc, adjusted_mode); -- 1.7.1.1

1 0

[Bug 26514] improper ENOMEM errors with Radeon 6.12.4
by bugzilla-daemon＠freedesktop.org 03 Sep '10

03 Sep '10

https://bugs.freedesktop.org/show_bug.cgi?id=26514 --- Comment #4 from Kenneth Koehler <kenneth.koehler(a)uc.edu> 2010-09-03 12:48:50 PDT --- Back after long delay; sorry! Upgraded libdrm to 2.4.15 Mesa to 7.7.1 xorg-server to 1.7.7 No change in symptoms. I don't believe I can upgrade further without rebuilding entire system due to dependencies (ie., Mesa 7.8.2 requires libdrm 2.4.19 which requires atomic ops not supported by my compiler). Is there any expectation that this is fixable within xorg 7.5? -- Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug.

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel September 2010