[meta-freescale] [meta-fsl-arm-extra][PATCH] UDOO board support, recipe for UDOO KernelUnico 2013 maintaned by UDOO manufacturer
Daiane Angolini
daiane.angolini at freescale.com
Mon Dec 16 04:13:37 PST 2013
On 16-12-2013 07:31, Luca Simone Ronga wrote:
> From: Luca Simone Ronga <luca.ronga at cnit.it>
>
> Dear Freescale community,
> here is the kernel recipe for the UDOO Board kernel branch maintained by SECO
> (KernelUnico 2013).
> I wish to thank John Weber and Otavio Salvador for their useful comments.
> L.
>
>
Could you, please, format your commit log and commit subject?
You can use other commits to get the format.
Please, let me know if you need help
I have not participated on discussion about these new recipes, but it
looks like so much patch added in SRC_URI. Was new git repository suggested?
Please, see more comments inline, in linux bb file
>
> Signed-off-by: Luca Simone Ronga <luca.simone.ronga at gmail.com>
> ---
> ...8-4.6.9p11.1-gpu-GPU-Kernel-driver-integr.patch | 1040 ++++++++++
> ...Fix-getrusage-related-build-failure-on-gl.patch | 43 +
> ...fix-memset-related-crashes-caused-by-rece.patch | 259 +++
> ...5-gpu-Add-global-value-for-minimum-3D-clo.patch | 62 +
> .../0003-ARM-7670-1-fix-the-memset-fix.patch | 87 +
> ...4-4-gpu-use-new-PU-power-on-off-interface.patch | 53 +
> ...8-1-GPU-Integrate-4.6.9p12-release-kernel.patch | 2006 ++++++++++++++++++++
> ...6-Fix-build-break-when-CONFIG_CLK_DEBUG-i.patch | 43 +
> ...5-GPU-Correct-suspend-resume-calling-afte.patch | 61 +
> ...0005-ENGR00271359-Add-Multi-touch-support.patch | 98 +
> .../0006-Add-support-for-DVI-monitors.patch | 227 +++
> ...0-gpu-Correct-section-mismatch-in-gpu-ker.patch | 60 +
> ...6-board-mx6q_sabresd-Register-SDHC3-first.patch | 38 +
> recipes-kernel/linux/linux-udoo-3.0.35/defconfig | 246 +++
> ...m-vivante-Add-00-sufix-in-returned-bus-Id.patch | 31 +
> ...-Rename-mxcfb_epdc_kernel.h-to-mxc_epdc.h.patch | 143 ++
> recipes-kernel/linux/linux-udoo.inc | 17 +
> recipes-kernel/linux/linux-udoo_3.0.35.bb | 26 +
> 18 files changed, 4540 insertions(+)
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0001-ENGR00255688-4.6.9p11.1-gpu-GPU-Kernel-driver-integr.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0001-perf-tools-Fix-getrusage-related-build-failure-on-gl.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0002-ARM-7668-1-fix-memset-related-crashes-caused-by-rece.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0002-ENGR00265465-gpu-Add-global-value-for-minimum-3D-clo.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0003-ARM-7670-1-fix-the-memset-fix.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0003-ENGR00261814-4-gpu-use-new-PU-power-on-off-interface.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0004-ENGR00264288-1-GPU-Integrate-4.6.9p12-release-kernel.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0004-ENGR00271136-Fix-build-break-when-CONFIG_CLK_DEBUG-i.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0005-ENGR00264275-GPU-Correct-suspend-resume-calling-afte.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0005-ENGR00271359-Add-Multi-touch-support.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0006-Add-support-for-DVI-monitors.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0006-ENGR00265130-gpu-Correct-section-mismatch-in-gpu-ker.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/0007-ARM-mach-mx6-board-mx6q_sabresd-Register-SDHC3-first.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/defconfig
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/drm-vivante-Add-00-sufix-in-returned-bus-Id.patch
> create mode 100644 recipes-kernel/linux/linux-udoo-3.0.35/epdc-Rename-mxcfb_epdc_kernel.h-to-mxc_epdc.h.patch
> create mode 100644 recipes-kernel/linux/linux-udoo.inc
> create mode 100644 recipes-kernel/linux/linux-udoo_3.0.35.bb
>
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0001-ENGR00255688-4.6.9p11.1-gpu-GPU-Kernel-driver-integr.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0001-ENGR00255688-4.6.9p11.1-gpu-GPU-Kernel-driver-integr.patch
> new file mode 100644
> index 0000000..9c7cd44
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0001-ENGR00255688-4.6.9p11.1-gpu-GPU-Kernel-driver-integr.patch
> @@ -0,0 +1,1040 @@
> +From 46e3a6de5adb9379f9d6eef2c038c2f18637d407 Mon Sep 17 00:00:00 2001
> +From: Loren Huang <b02279 at freescale.com>
> +Date: Mon, 25 Mar 2013 15:43:57 +0800
> +Subject: [PATCH 1/6] ENGR00255688 4.6.9p11.1 [gpu]GPU Kernel driver
> + integration
> +
> +4.6.9p11.1 GPU kernel driver integration
> +Cherry pick from imx_3.0.35
> +
> +Upstream-Status: Backport [3.5.7-1.0.0]
> +
> +Signed-off-by: Loren Huang <b02279 at freescale.com>
> +Acked-by: Lily Zhang
> +---
> + drivers/mxc/gpu-viv/Kbuild | 2 +-
> + .../arch/XAQ2/hal/kernel/gc_hal_kernel_context.c | 2 +-
> + .../arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c | 7 +-
> + drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.h | 2 +-
> + drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c | 53 ++++--
> + .../mxc/gpu-viv/hal/kernel/gc_hal_kernel_event.c | 5 +-
> + drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c | 178 ++++++++++++---------
> + .../hal/kernel/gc_hal_kernel_video_memory.c | 3 +-
> + .../gpu-viv/hal/kernel/inc/gc_hal_eglplatform.h | 13 +-
> + drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h | 25 +++
> + .../mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h | 35 ++++
> + .../mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h | 2 +-
> + .../hal/os/linux/kernel/gc_hal_kernel_driver.c | 2 +-
> + .../hal/os/linux/kernel/gc_hal_kernel_linux.h | 6 +
> + .../gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c | 82 +++++++++-
> + 15 files changed, 304 insertions(+), 113 deletions(-)
> +
> +diff --git a/drivers/mxc/gpu-viv/Kbuild b/drivers/mxc/gpu-viv/Kbuild
> +index 0b18a7b..93b1259 100644
> +--- a/drivers/mxc/gpu-viv/Kbuild
> ++++ b/drivers/mxc/gpu-viv/Kbuild
> +@@ -1,6 +1,6 @@
> + ##############################################################################
> + #
> +-# Copyright (C) 2005 - 2012 by Vivante Corp.
> ++# Copyright (C) 2005 - 2013 by Vivante Corp.
> + #
> + # This program is free software; you can redistribute it and/or modify
> + # it under the terms of the GNU General Public License as published by
> +diff --git a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_context.c b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_context.c
> +index 22e1f27..24003e7 100644
> +--- a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_context.c
> ++++ b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_context.c
> +@@ -471,7 +471,7 @@ _InitializeContextBuffer(
> + index += _SwitchPipe(Context, index, gcvPIPE_3D);
> +
> + /* Current context pointer. */
> +-#if gcdDEBUG
> ++#if gcdDEBUG
> + index += _State(Context, index, 0x03850 >> 2, 0x00000000, 1, gcvFALSE, gcvFALSE);
> + #endif
> +
> +diff --git a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> +index a87259e..3829999 100644
> +--- a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> ++++ b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> +@@ -232,7 +232,8 @@ _IdentifyHardware(
> + }
> +
> + /* Exception for GC1000, revision 5035 & GC800, revision 4612 */
> +- if (((Identity->chipModel == gcv1000) && (Identity->chipRevision == 0x5035))
> ++ if (((Identity->chipModel == gcv1000) && ((Identity->chipRevision == 0x5035)
> ++ || (Identity->chipRevision == 0x5036)))
> + || ((Identity->chipModel == gcv800) && (Identity->chipRevision == 0x4612)))
> + {
> + Identity->superTileMode = 1;
> +@@ -751,7 +752,7 @@ gckHARDWARE_Construct(
> + /* Initialize the fast clear. */
> + gcmkONERROR(gckHARDWARE_SetFastClear(hardware, -1, -1));
> +
> +-#if !gcdENABLE_128B_MERGE
> ++#if !gcdENABLE_128B_MERGE
> +
> + if (((((gctUINT32) (hardware->identity.chipMinorFeatures2)) >> (0 ? 21:21) & ((gctUINT32) ((((1 ? 21:21) - (0 ? 21:21) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 21:21) - (0 ? 21:21) + 1)))))) == (0x1 & ((gctUINT32) ((((1 ? 21:21) - (0 ? 21:21) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 21:21) - (0 ? 21:21) + 1))))))))
> + {
> +@@ -1027,7 +1028,7 @@ gckHARDWARE_InitializeHardware(
> + 0x00424,
> + baseAddress));
> +
> +-#if !VIVANTE_PROFILER
> ++#if !VIVANTE_PROFILER
> + {
> + gctUINT32 data;
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.h b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.h
> +index 1da80b7..5896e93 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.h
> +@@ -186,7 +186,7 @@ typedef struct _gcsDATABASE
> + gctUINT64 idle;
> +
> + /* Pointer to database. */
> +- gcsDATABASE_RECORD_PTR list;
> ++ gcsDATABASE_RECORD_PTR list[48];
> +
> + #if gcdSECURE_USER
> + /* Secure cache. */
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c
> +index 1fb18fb..bc5f083 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c
> +@@ -26,6 +26,9 @@
> + /*******************************************************************************
> + ***** Private fuctions ********************************************************/
> +
> ++#define _GetSlot(database, x) \
> ++ (gctUINT32)(((gcmPTR_TO_UINT64(x) >> 7) % gcmCOUNTOF(database->list)))
> ++
> + /*******************************************************************************
> + ** gckKERNEL_NewDatabase
> + **
> +@@ -56,6 +59,7 @@ gckKERNEL_NewDatabase(
> + gcsDATABASE_PTR database;
> + gctBOOL acquired = gcvFALSE;
> + gctSIZE_T slot;
> ++ gcsDATABASE_PTR existingDatabase;
> +
> + gcmkHEADER_ARG("Kernel=0x%x ProcessID=%d", Kernel, ProcessID);
> +
> +@@ -63,6 +67,21 @@ gckKERNEL_NewDatabase(
> + gcmkONERROR(gckOS_AcquireMutex(Kernel->os, Kernel->db->dbMutex, gcvINFINITE));
> + acquired = gcvTRUE;
> +
> ++ /* Compute the hash for the database. */
> ++ slot = ProcessID % gcmCOUNTOF(Kernel->db->db);
> ++
> ++ /* Walk the hash list. */
> ++ for (existingDatabase = Kernel->db->db[slot];
> ++ existingDatabase != gcvNULL;
> ++ existingDatabase = existingDatabase->next)
> ++ {
> ++ if (existingDatabase->processID == ProcessID)
> ++ {
> ++ /* One process can't be added twice. */
> ++ gcmkONERROR(gcvSTATUS_NOT_SUPPORTED);
> ++ }
> ++ }
> ++
> + if (Kernel->db->freeDatabase != gcvNULL)
> + {
> + /* Allocate a database from the free list. */
> +@@ -81,9 +100,6 @@ gckKERNEL_NewDatabase(
> + database = pointer;
> + }
> +
> +- /* Compute the hash for the database. */
> +- slot = ProcessID % gcmCOUNTOF(Kernel->db->db);
> +-
> + /* Insert the database into the hash. */
> + database->next = Kernel->db->db[slot];
> + Kernel->db->db[slot] = database;
> +@@ -350,6 +366,7 @@ static gceSTATUS
> + gckKERNEL_NewRecord(
> + IN gckKERNEL Kernel,
> + IN gcsDATABASE_PTR Database,
> ++ IN gctUINT32 Slot,
> + OUT gcsDATABASE_RECORD_PTR * Record
> + )
> + {
> +@@ -383,8 +400,8 @@ gckKERNEL_NewRecord(
> + }
> +
> + /* Insert the record in the database. */
> +- record->next = Database->list;
> +- Database->list = record;
> ++ record->next = Database->list[Slot];
> ++ Database->list[Slot] = record;
> +
> + /* Release the database mutex. */
> + gcmkONERROR(gckOS_ReleaseMutex(Kernel->os, Kernel->db->dbMutex));
> +@@ -449,6 +466,7 @@ gckKERNEL_DeleteRecord(
> + gceSTATUS status;
> + gctBOOL acquired = gcvFALSE;
> + gcsDATABASE_RECORD_PTR record, previous;
> ++ gctUINT32 slot = _GetSlot(Database, Data);
> +
> + gcmkHEADER_ARG("Kernel=0x%x Database=0x%x Type=%d Data=0x%x",
> + Kernel, Database, Type, Data);
> +@@ -458,8 +476,9 @@ gckKERNEL_DeleteRecord(
> + gckOS_AcquireMutex(Kernel->os, Kernel->db->dbMutex, gcvINFINITE));
> + acquired = gcvTRUE;
> +
> ++
> + /* Scan the database for this record. */
> +- for (record = Database->list, previous = gcvNULL;
> ++ for (record = Database->list[slot], previous = gcvNULL;
> + record != gcvNULL;
> + record = record->next
> + )
> +@@ -490,7 +509,7 @@ gckKERNEL_DeleteRecord(
> + /* Remove record from database. */
> + if (previous == gcvNULL)
> + {
> +- Database->list = record->next;
> ++ Database->list[slot] = record->next;
> + }
> + else
> + {
> +@@ -557,6 +576,7 @@ gckKERNEL_FindRecord(
> + gceSTATUS status;
> + gctBOOL acquired = gcvFALSE;
> + gcsDATABASE_RECORD_PTR record;
> ++ gctUINT32 slot = _GetSlot(Database, Data);
> +
> + gcmkHEADER_ARG("Kernel=0x%x Database=0x%x Type=%d Data=0x%x",
> + Kernel, Database, Type, Data);
> +@@ -567,7 +587,7 @@ gckKERNEL_FindRecord(
> + acquired = gcvTRUE;
> +
> + /* Scan the database for this record. */
> +- for (record = Database->list;
> ++ for (record = Database->list[slot];
> + record != gcvNULL;
> + record = record->next
> + )
> +@@ -642,6 +662,7 @@ gckKERNEL_CreateProcessDB(
> + {
> + gceSTATUS status;
> + gcsDATABASE_PTR database = gcvNULL;
> ++ gctUINT32 i;
> +
> + gcmkHEADER_ARG("Kernel=0x%x ProcessID=%d", Kernel, ProcessID);
> +
> +@@ -668,7 +689,11 @@ gckKERNEL_CreateProcessDB(
> + database->mapUserMemory.bytes = 0;
> + database->mapUserMemory.maxBytes = 0;
> + database->mapUserMemory.totalBytes = 0;
> +- database->list = gcvNULL;
> ++
> ++ for (i = 0; i < gcmCOUNTOF(database->list); i++)
> ++ {
> ++ database->list[i] = gcvNULL;
> ++ }
> +
> + #if gcdSECURE_USER
> + {
> +@@ -848,7 +873,7 @@ gckKERNEL_AddProcessDB(
> + gcmkONERROR(gckKERNEL_FindDatabase(Kernel, ProcessID, gcvFALSE, &database));
> +
> + /* Create a new record in the database. */
> +- gcmkONERROR(gckKERNEL_NewRecord(Kernel, database, &record));
> ++ gcmkONERROR(gckKERNEL_NewRecord(Kernel, database, _GetSlot(database, Pointer), &record));
> +
> + /* Initialize the record. */
> + record->kernel = Kernel;
> +@@ -1086,6 +1111,7 @@ gckKERNEL_DestroyProcessDB(
> + gctPHYS_ADDR physical;
> + gcuVIDMEM_NODE_PTR node;
> + gckKERNEL kernel = Kernel;
> ++ gctUINT32 i;
> +
> + gcmkHEADER_ARG("Kernel=0x%x ProcessID=%d", Kernel, ProcessID);
> +
> +@@ -1126,8 +1152,11 @@ gckKERNEL_DestroyProcessDB(
> + ProcessID);
> + }
> +
> ++ for(i = 0; i < gcmCOUNTOF(database->list); i++)
> ++ {
> ++
> + /* Walk all records. */
> +- for (record = database->list; record != gcvNULL; record = next)
> ++ for (record = database->list[i]; record != gcvNULL; record = next)
> + {
> + /* Next next record. */
> + next = record->next;
> +@@ -1293,6 +1322,8 @@ gckKERNEL_DestroyProcessDB(
> + gcvNULL));
> + }
> +
> ++ }
> ++
> + /* Delete the database. */
> + gcmkONERROR(gckKERNEL_DeleteDatabase(Kernel, database));
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_event.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_event.c
> +index f78d096..217f7f1 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_event.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_event.c
> +@@ -959,6 +959,8 @@ gckEVENT_AddList(
> + record->kernel = Event->kernel;
> + #endif
> +
> ++ gcmkONERROR(__RemoveRecordFromProcessDB(Event, record));
> ++
> + /* Acquire the mutex. */
> + gcmkONERROR(gckOS_AcquireMutex(Event->os, Event->eventListMutex, gcvINFINITE));
> + acquired = gcvTRUE;
> +@@ -1539,9 +1541,6 @@ gckEVENT_Submit(
> + gcmkONERROR(gckOS_ReleaseMutex(Event->os, Event->eventListMutex));
> + acquired = gcvFALSE;
> +
> +- gcmkONERROR(__RemoveRecordFromProcessDB(Event,
> +- Event->queues[id].head));
> +-
> + #if gcdNULL_DRIVER
> + /* Notify immediately on infinite hardware. */
> + gcmkONERROR(gckEVENT_Interrupt(Event, 1 << id));
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c
> +index 0c71e28..43c9297 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c
> +@@ -97,6 +97,43 @@ static gcsMirrorPageTable_PTR mirrorPageTable = gcvNULL;
> + static gctPOINTER mirrorPageTableMutex = gcvNULL;
> + #endif
> +
> ++static void
> ++_WritePageEntry(
> ++ IN gctUINT32_PTR PageEntry,
> ++ IN gctUINT32 EntryValue
> ++ )
> ++{
> ++ static gctUINT16 data = 0xff00;
> ++
> ++ if (*(gctUINT8 *)&data == 0xff)
> ++ {
> ++ *PageEntry = gcmSWAB32(EntryValue);
> ++ }
> ++ else
> ++ {
> ++ *PageEntry = EntryValue;
> ++ }
> ++}
> ++
> ++static gctUINT32
> ++_ReadPageEntry(
> ++ IN gctUINT32_PTR PageEntry
> ++ )
> ++{
> ++ static gctUINT16 data = 0xff00;
> ++ gctUINT32 entryValue;
> ++
> ++ if (*(gctUINT8 *)&data == 0xff)
> ++ {
> ++ entryValue = *PageEntry;
> ++ return gcmSWAB32(entryValue);
> ++ }
> ++ else
> ++ {
> ++ return *PageEntry;
> ++ }
> ++}
> ++
> + static gceSTATUS
> + _FillPageTable(
> + IN gctUINT32_PTR PageTable,
> +@@ -108,7 +145,7 @@ _FillPageTable(
> +
> + for (i = 0; i < PageCount; i++)
> + {
> +- PageTable[i] = EntryValue;
> ++ _WritePageEntry(PageTable + i, EntryValue);
> + }
> +
> + return gcvSTATUS_OK;
> +@@ -132,16 +169,16 @@ _Link(
> + gctUINT32_PTR pageTable = Mmu->pageTableLogical;
> +
> + /* Dispatch on node type. */
> +- switch (gcmENTRY_TYPE(pageTable[Index]))
> ++ switch (gcmENTRY_TYPE(_ReadPageEntry(&pageTable[Index])))
> + {
> + case gcvMMU_SINGLE:
> + /* Set single index. */
> +- pageTable[Index] = (Next << 8) | gcvMMU_SINGLE;
> ++ _WritePageEntry(&pageTable[Index], (Next << 8) | gcvMMU_SINGLE);
> + break;
> +
> + case gcvMMU_FREE:
> + /* Set index. */
> +- pageTable[Index + 1] = Next;
> ++ _WritePageEntry(&pageTable[Index + 1], Next);
> + break;
> +
> + default:
> +@@ -167,13 +204,13 @@ _AddFree(
> + if (Count == 1)
> + {
> + /* Initialize a single page node. */
> +- pageTable[Node] = (~((1U<<8)-1)) | gcvMMU_SINGLE;
> ++ _WritePageEntry(pageTable + Node, (~((1U<<8)-1)) | gcvMMU_SINGLE);
> + }
> + else
> + {
> + /* Initialize the node. */
> +- pageTable[Node + 0] = (Count << 8) | gcvMMU_FREE;
> +- pageTable[Node + 1] = ~0U;
> ++ _WritePageEntry(pageTable + Node + 0, (Count << 8) | gcvMMU_FREE);
> ++ _WritePageEntry(pageTable + Node + 1, ~0U);
> + }
> +
> + /* Append the node. */
> +@@ -196,7 +233,7 @@ _Collect(
> + for (i = 0; i < Mmu->pageTableEntries; ++i)
> + {
> + /* Dispatch based on type of page. */
> +- switch (gcmENTRY_TYPE(pageTable[i]))
> ++ switch (gcmENTRY_TYPE(_ReadPageEntry(&pageTable[i])))
> + {
> + case gcvMMU_USED:
> + /* Used page, so close any open node. */
> +@@ -229,10 +266,10 @@ _Collect(
> + }
> +
> + /* Advance the count. */
> +- count += pageTable[i] >> 8;
> ++ count += _ReadPageEntry(&pageTable[i]) >> 8;
> +
> + /* Advance the index into the page table. */
> +- i += (pageTable[i] >> 8) - 1;
> ++ i += (_ReadPageEntry(&pageTable[i]) >> 8) - 1;
> + break;
> +
> + default:
> +@@ -341,19 +378,20 @@ _FillFlatMapping(
> + gcmkONERROR(gcvSTATUS_NOT_ALIGNED);
> + }
> +
> +- *(Mmu->mtlbLogical + mStart)
> +- = stlb->physBase
> +- /* 64KB page size */
> +- | (1 << 2)
> +- /* Ignore exception */
> +- | (0 << 1)
> +- /* Present */
> +- | (1 << 0);
> ++ _WritePageEntry(Mmu->mtlbLogical + mStart,
> ++ stlb->physBase
> ++ /* 64KB page size */
> ++ | (1 << 2)
> ++ /* Ignore exception */
> ++ | (0 << 1)
> ++ /* Present */
> ++ | (1 << 0)
> ++ );
> + #if gcdMMU_TABLE_DUMP
> + gckOS_Print("%s(%d): insert MTLB[%d]: %08x\n",
> + __FUNCTION__, __LINE__,
> + mStart,
> +- *(Mmu->mtlbLogical + mStart));
> ++ _ReadPageEntry(Mmu->mtlbLogical + mStart));
> + #endif
> +
> + stlb->mtlbIndex = mStart;
> +@@ -368,12 +406,12 @@ _FillFlatMapping(
> + while (sStart <= last)
> + {
> + gcmkASSERT(!(start & gcdMMU_PAGE_64K_MASK));
> +- *(stlb->logical + sStart) = _SetPage(start);
> ++ _WritePageEntry(stlb->logical + sStart, _SetPage(start));
> + #if gcdMMU_TABLE_DUMP
> + gckOS_Print("%s(%d): insert STLB[%d]: %08x\n",
> + __FUNCTION__, __LINE__,
> + sStart,
> +- *(stlb->logical + sStart));
> ++ _ReadPageEntry(stlb->logical + sStart));
> + #endif
> + /* next page. */
> + start += gcdMMU_PAGE_64K_SIZE;
> +@@ -428,7 +466,7 @@ OnError:
> + if (pre->mtlbEntryNum != 0)
> + {
> + gcmkASSERT(pre->mtlbEntryNum == 1);
> +- *(Mmu->mtlbLogical + pre->mtlbIndex) = 0;
> ++ _WritePageEntry(Mmu->mtlbLogical + pre->mtlbIndex, 0);
> + }
> +
> + gcmkVERIFY_OK(gcmkOS_SAFE_FREE(Mmu->os, pre));
> +@@ -493,8 +531,8 @@ _SetupDynamicSpace(
> +
> + /* Initilization. */
> + pageTable = Mmu->pageTableLogical;
> +- pageTable[0] = (Mmu->pageTableEntries << 8) | gcvMMU_FREE;
> +- pageTable[1] = ~0U;
> ++ _WritePageEntry(pageTable, (Mmu->pageTableEntries << 8) | gcvMMU_FREE);
> ++ _WritePageEntry(pageTable + 1, ~0U);
> + Mmu->heapList = 0;
> + Mmu->freeNodes = gcvFALSE;
> +
> +@@ -509,18 +547,20 @@ _SetupDynamicSpace(
> + /* Map to Master TLB. */
> + for (; i < gcdMMU_MTLB_ENTRY_NUM; i++)
> + {
> +- Mmu->mtlbLogical[i] = physical
> +- /* 4KB page size */
> +- | (0 << 2)
> +- /* Ignore exception */
> +- | (0 << 1)
> +- /* Present */
> +- | (1 << 0);
> ++ _WritePageEntry(Mmu->mtlbLogical + i,
> ++ physical
> ++ /* 4KB page size */
> ++ | (0 << 2)
> ++ /* Ignore exception */
> ++ | (0 << 1)
> ++ /* Present */
> ++ | (1 << 0)
> ++ );
> + #if gcdMMU_TABLE_DUMP
> + gckOS_Print("%s(%d): insert MTLB[%d]: %08x\n",
> + __FUNCTION__, __LINE__,
> + i,
> +- *(Mmu->mtlbLogical + i));
> ++ _ReadPageEntry(Mmu->mtlbLogical + i));
> + #endif
> + physical += gcdMMU_STLB_4K_SIZE;
> + }
> +@@ -645,18 +685,11 @@ _Construct(
> + pageTable = mmu->pageTableLogical;
> +
> + #if gcdMMU_CLEAR_VALUE
> +- {
> +- gctUINT32 i;
> +-
> +- for (i = 0; i < mmu->pageTableEntries; ++i)
> +- {
> +- pageTable[i] = gcdMMU_CLEAR_VALUE;
> +- }
> +- }
> ++ _FillPageTable(pageTable, mmu->pageTableEntries, gcdMMU_CLEAR_VALUE);
> + #endif
> +
> +- pageTable[0] = (mmu->pageTableEntries << 8) | gcvMMU_FREE;
> +- pageTable[1] = ~0U;
> ++ _WritePageEntry(pageTable, (mmu->pageTableEntries << 8) | gcvMMU_FREE);
> ++ _WritePageEntry(pageTable + 1, ~0U);
> + mmu->heapList = 0;
> + mmu->freeNodes = gcvFALSE;
> +
> +@@ -797,7 +830,7 @@ _Destroy(
> + if (pre->mtlbEntryNum != 0)
> + {
> + gcmkASSERT(pre->mtlbEntryNum == 1);
> +- *(Mmu->mtlbLogical + pre->mtlbIndex) = 0;
> ++ _WritePageEntry(Mmu->mtlbLogical + pre->mtlbIndex, 0);
> + #if gcdMMU_TABLE_DUMP
> + gckOS_Print("%s(%d): clean MTLB[%d]\n",
> + __FUNCTION__, __LINE__,
> +@@ -1044,7 +1077,7 @@ _AllocatePages(
> + for (index = Mmu->heapList; !gotIt && (index < Mmu->pageTableEntries);)
> + {
> + /* Check the node type. */
> +- switch (gcmENTRY_TYPE(pageTable[index]))
> ++ switch (gcmENTRY_TYPE(_ReadPageEntry(&pageTable[index])))
> + {
> + case gcvMMU_SINGLE:
> + /* Single odes are valid if we only need 1 page. */
> +@@ -1056,13 +1089,13 @@ _AllocatePages(
> + {
> + /* Move to next node. */
> + previous = index;
> +- index = pageTable[index] >> 8;
> ++ index = _ReadPageEntry(&pageTable[index]) >> 8;
> + }
> + break;
> +
> + case gcvMMU_FREE:
> + /* Test if the node has enough space. */
> +- if (PageCount <= (pageTable[index] >> 8))
> ++ if (PageCount <= (_ReadPageEntry(&pageTable[index]) >> 8))
> + {
> + gotIt = gcvTRUE;
> + }
> +@@ -1070,7 +1103,7 @@ _AllocatePages(
> + {
> + /* Move to next node. */
> + previous = index;
> +- index = pageTable[index + 1];
> ++ index = _ReadPageEntry(&pageTable[index + 1]);
> + }
> + break;
> +
> +@@ -1099,36 +1132,36 @@ _AllocatePages(
> + }
> + }
> +
> +- switch (gcmENTRY_TYPE(pageTable[index]))
> ++ switch (gcmENTRY_TYPE(_ReadPageEntry(&pageTable[index])))
> + {
> + case gcvMMU_SINGLE:
> + /* Unlink single node from free list. */
> + gcmkONERROR(
> +- _Link(Mmu, previous, pageTable[index] >> 8));
> ++ _Link(Mmu, previous, _ReadPageEntry(&pageTable[index]) >> 8));
> + break;
> +
> + case gcvMMU_FREE:
> + /* Check how many pages will be left. */
> +- left = (pageTable[index] >> 8) - PageCount;
> ++ left = (_ReadPageEntry(&pageTable[index]) >> 8) - PageCount;
> + switch (left)
> + {
> + case 0:
> + /* The entire node is consumed, just unlink it. */
> + gcmkONERROR(
> +- _Link(Mmu, previous, pageTable[index + 1]));
> ++ _Link(Mmu, previous, _ReadPageEntry(&pageTable[index + 1])));
> + break;
> +
> + case 1:
> + /* One page will remain. Convert the node to a single node and
> + ** advance the index. */
> +- pageTable[index] = (pageTable[index + 1] << 8) | gcvMMU_SINGLE;
> ++ _WritePageEntry(&pageTable[index], (_ReadPageEntry(&pageTable[index + 1]) << 8) | gcvMMU_SINGLE);
> + index ++;
> + break;
> +
> + default:
> + /* Enough pages remain for a new node. However, we will just adjust
> + ** the size of the current node and advance the index. */
> +- pageTable[index] = (left << 8) | gcvMMU_FREE;
> ++ _WritePageEntry(&pageTable[index], (left << 8) | gcvMMU_FREE);
> + index += left;
> + break;
> + }
> +@@ -1232,35 +1265,32 @@ _FreePages(
> + #if gcdMMU_CLEAR_VALUE
> + if (Mmu->hardware->mmuVersion == 0)
> + {
> +- gctUINT32 i;
> +-
> +- for (i = 0; i < PageCount; ++i)
> +- {
> +- pageTable[i] = gcdMMU_CLEAR_VALUE;
> +- }
> ++ _FillPageTable(pageTable, PageCount, gcdMMU_CLEAR_VALUE);
> + }
> + #endif
> +
> + if (PageCount == 1)
> + {
> + /* Single page node. */
> +- pageTable[0] = (~((1U<<8)-1)) | gcvMMU_SINGLE
> ++ _WritePageEntry(pageTable,
> ++ (~((1U<<8)-1)) | gcvMMU_SINGLE
> + #if gcdUSE_MMU_EXCEPTION
> +- /* Enable exception */
> +- | (1 << 1)
> ++ /* Enable exception */
> ++ | 1 << 1
> + #endif
> +- ;
> ++ );
> + }
> + else
> + {
> + /* Mark the node as free. */
> +- pageTable[0] = (PageCount << 8) | gcvMMU_FREE
> ++ _WritePageEntry(pageTable,
> ++ (PageCount << 8) | gcvMMU_FREE
> + #if gcdUSE_MMU_EXCEPTION
> +- /* Enable exception */
> +- | (1 << 1)
> ++ /* Enable exception */
> ++ | 1 << 1
> + #endif
> +- ;
> +- pageTable[1] = ~0U;
> ++ );
> ++ _WritePageEntry(pageTable + 1, ~0U);
> +
> + #if gcdUSE_MMU_EXCEPTION
> + /* Enable exception */
> +@@ -1509,12 +1539,8 @@ gckMMU_SetPage(
> + data = _SetPage(PageAddress);
> + }
> +
> +- if (Mmu->hardware->bigEndian)
> +- {
> +- data = gcmSWAB32(data);
> +- }
> ++ _WritePageEntry(PageEntry, data);
> +
> +- *PageEntry = data;
> + #if gcdMIRROR_PAGETABLE
> + for (i = 0; i < mirrorPageTable->reference; i++)
> + {
> +@@ -1526,11 +1552,11 @@ gckMMU_SetPage(
> +
> + if (mmu->hardware->mmuVersion == 0)
> + {
> +- *pageEntry = PageAddress;
> ++ _WritePageEntry(pageEntry, PageAddress);
> + }
> + else
> + {
> +- *pageEntry = _SetPage(PageAddress);
> ++ _WritePageEntry(pageEntry, _SetPage(PageAddress));
> + }
> + }
> +
> +@@ -1734,7 +1760,7 @@ gckMMU_DumpPageTableEntry(
> + * gcdMMU_STLB_4K_ENTRY_NUM
> + + stlb;
> +
> +- gcmkPRINT(" Page table entry = 0x%08X", pageTable[index]);
> ++ gcmkPRINT(" Page table entry = 0x%08X", _ReadPageEntry(pageTable + index));
> + }
> +
> + gcmkFOOTER_NO();
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_video_memory.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_video_memory.c
> +index d49aa64..8a442a2 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_video_memory.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_video_memory.c
> +@@ -1027,7 +1027,8 @@ gckVIDMEM_AllocateLinear(
> + )
> + {
> + /* The left memory is for small memory.*/
> +- gcmkONERROR(gcvSTATUS_OUT_OF_MEMORY);
> ++ status = gcvSTATUS_OUT_OF_MEMORY;
> ++ goto OnError;
> + }
> + #endif
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_eglplatform.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_eglplatform.h
> +index 496276e..06eea79 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_eglplatform.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_eglplatform.h
> +@@ -227,7 +227,8 @@ gcoOS_GetDisplayInfoEx(
> + );
> +
> + gceSTATUS
> +-gcoOS_GetNextDisplayInfoEx(
> ++gcoOS_GetNextDisplayInfoExByIndex(
> ++ IN gctINT Index,
> + IN HALNativeDisplayType Display,
> + IN HALNativeWindowType Window,
> + IN gctUINT DisplayInfoSize,
> +@@ -274,15 +275,15 @@ gcoOS_SetDisplayVirtualEx(
> +
> + gceSTATUS
> + gcoOS_SetSwapInterval(
> +- IN HALNativeDisplayType Display,
> +- IN gctINT Interval
> ++ IN HALNativeDisplayType Display,
> ++ IN gctINT Interval
> + );
> +
> + gceSTATUS
> + gcoOS_GetSwapInterval(
> +- IN HALNativeDisplayType Display,
> +- IN gctINT_PTR Min,
> +- IN gctINT_PTR Max
> ++ IN HALNativeDisplayType Display,
> ++ IN gctINT_PTR Min,
> ++ IN gctINT_PTR Max
> + );
> +
> + gceSTATUS
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h
> +index d441d1d..249b61b 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h
> +@@ -1430,6 +1430,16 @@ typedef enum _gceTEXTURE_FACE
> + }
> + gceTEXTURE_FACE;
> +
> ++#if gcdFORCE_MIPMAP
> ++typedef enum
> ++{
> ++ gcvForceMipDisabled = 0,
> ++ gcvForceMipEnable = 1,
> ++ gcvForceMipGenerated = 2,
> ++ gcvForceMipNever = 3,
> ++}gceFORCE_MIPMAP;
> ++#endif
> ++
> + typedef struct _gcsTEXTURE
> + {
> + /* Addressing modes. */
> +@@ -1446,6 +1456,10 @@ typedef struct _gcsTEXTURE
> + gceTEXTURE_FILTER mipFilter;
> + gctUINT anisoFilter;
> + gctBOOL forceTopLevel;
> ++ gctBOOL autoMipmap;
> ++#if gcdFORCE_MIPMAP
> ++ gceFORCE_MIPMAP forceMipmap;
> ++#endif
> + /* Level of detail. */
> + gctFIXED_POINT lodBias;
> + gctFIXED_POINT lodMin;
> +@@ -1479,7 +1493,18 @@ gceSTATUS
> + gcoTEXTURE_Destroy(
> + IN gcoTEXTURE Texture
> + );
> ++#if gcdFORCE_MIPMAP
> ++gceSTATUS
> ++gcoTEXTURE_DestroyForceMipmap(
> ++ IN gcoTEXTURE Texture
> ++ );
> +
> ++gceSTATUS
> ++gcoTEXTURE_GetMipLevels(
> ++ IN gcoTEXTURE Texture,
> ++ OUT gctINT * levels
> ++ );
> ++#endif
> + /* Replace a mipmap in gcoTEXTURE object. */
> + gceSTATUS
> + gcoTEXTURE_ReplaceMipMap(
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h
> +index 86e9133..afe83d0 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h
> +@@ -114,6 +114,30 @@
> + #define COMMAND_PROCESSOR_VERSION 1
> +
> + /*
> ++ gcdDUMP_KEY
> ++
> ++ Set this to a string that appears in 'cat /proc/<pid>/cmdline'. E.g. 'camera'.
> ++ HAL will create dumps for the processes matching this key.
> ++*/
> ++#ifndef gcdDUMP_KEY
> ++# define gcdDUMP_KEY "process"
> ++#endif
> ++
> ++/*
> ++ gcdDUMP_PATH
> ++
> ++ The dump file location. Some processes cannot write to the sdcard.
> ++ Try apps' data dir, e.g. /data/data/com.android.launcher
> ++*/
> ++#ifndef gcdDUMP_PATH
> ++#if defined(ANDROID)
> ++# define gcdDUMP_PATH "/mnt/sdcard/"
> ++#else
> ++# define gcdDUMP_PATH "./"
> ++#endif
> ++#endif
> ++
> ++/*
> + gcdDUMP
> +
> + When set to 1, a dump of all states and memory uploads, as well as other
> +@@ -342,6 +366,17 @@
> + #endif
> +
> + /*
> ++ gcdUSER_HEAP_ALLOCATOR
> ++
> ++ Set to 1 to enable user mode heap allocator for fast memory allocation
> ++ and destroying. Otherwise, memory allocation/destroying in user mode
> ++ will be directly managed by system. Only for linux for now.
> ++*/
> ++#ifndef gcdUSER_HEAP_ALLOCATOR
> ++# define gcdUSER_HEAP_ALLOCATOR 1
> ++#endif
> ++
> ++/*
> + gcdHEAP_SIZE
> +
> + Set the allocation size for the internal heaps. Each time a heap is
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h
> +index 2881604..808fde0 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h
> +@@ -28,7 +28,7 @@
> +
> + #define gcvVERSION_PATCH 9
> +
> +-#define gcvVERSION_BUILD 1210
> ++#define gcvVERSION_BUILD 4651
> +
> + #define gcvVERSION_DATE __DATE__
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +index 4e3819c..2ed3d0e 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +@@ -663,7 +663,7 @@ static int drv_mmap(
> +
> + #if !gcdPAGED_MEMORY_CACHEABLE
> + vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
> +- vma->vm_flags |= VM_IO | VM_DONTCOPY | VM_DONTEXPAND;
> ++ vma->vm_flags |= gcdVM_FLAGS;
> + #endif
> + vma->vm_pgoff = 0;
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_linux.h b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_linux.h
> +index 9c0bcd5..3c148f6 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_linux.h
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_linux.h
> +@@ -73,6 +73,12 @@
> +
> + #define GetPageCount(size, offset) ((((size) + ((offset) & ~PAGE_CACHE_MASK)) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT)
> +
> ++#if LINUX_VERSION_CODE >= KERNEL_VERSION (3,7,0)
> ++#define gcdVM_FLAGS (VM_IO | VM_DONTCOPY | VM_DONTEXPAND | VM_DONTDUMP)
> ++#else
> ++#define gcdVM_FLAGS (VM_IO | VM_DONTCOPY | VM_DONTEXPAND | VM_RESERVED)
> ++#endif
> ++
> + static inline gctINT
> + GetOrder(
> + IN gctINT numPages
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> +index c07ded8..9c2bae6 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> +@@ -869,6 +869,60 @@ _UnmapUserLogical(
> + #endif
> + }
> +
> ++gceSTATUS
> ++_QueryProcessPageTable(
> ++ IN gctPOINTER Logical,
> ++ OUT gctUINT32 * Address
> ++ )
> ++{
> ++ spinlock_t *lock;
> ++ gctUINTPTR_T logical = (gctUINTPTR_T)Logical;
> ++ pgd_t *pgd;
> ++ pud_t *pud;
> ++ pmd_t *pmd;
> ++ pte_t *pte;
> ++
> ++ if (!current->mm)
> ++ {
> ++ return gcvSTATUS_NOT_FOUND;
> ++ }
> ++
> ++ pgd = pgd_offset(current->mm, logical);
> ++ if (pgd_none(*pgd) || pgd_bad(*pgd))
> ++ {
> ++ return gcvSTATUS_NOT_FOUND;
> ++ }
> ++
> ++ pud = pud_offset(pgd, logical);
> ++ if (pud_none(*pud) || pud_bad(*pud))
> ++ {
> ++ return gcvSTATUS_NOT_FOUND;
> ++ }
> ++
> ++ pmd = pmd_offset(pud, logical);
> ++ if (pmd_none(*pmd) || pmd_bad(*pmd))
> ++ {
> ++ return gcvSTATUS_NOT_FOUND;
> ++ }
> ++
> ++ pte = pte_offset_map_lock(current->mm, pmd, logical, &lock);
> ++ if (!pte)
> ++ {
> ++ return gcvSTATUS_NOT_FOUND;
> ++ }
> ++
> ++ if (!pte_present(*pte))
> ++ {
> ++ pte_unmap_unlock(pte, lock);
> ++ return gcvSTATUS_NOT_FOUND;
> ++ }
> ++
> ++ *Address = (pte_pfn(*pte) << PAGE_SHIFT) | (logical & ~PAGE_MASK);
> ++ pte_unmap_unlock(pte, lock);
> ++
> ++ return gcvSTATUS_OK;
> ++}
> ++
> + /*******************************************************************************
> + **
> + ** gckOS_Construct
> +@@ -1106,6 +1160,9 @@ _CreateKernelVirtualMapping(
> + numPages,
> + 0,
> + PAGE_KERNEL);
> ++
> ++ /* Trigger a page fault. */
> ++ memset(addr, 0, numPages * PAGE_SIZE);
> + }
> + #else
> + struct page ** pages;
> +@@ -1136,6 +1193,9 @@ _CreateKernelVirtualMapping(
> + /* ioremap() can't work on system memory since 2.6.38. */
> + addr = vmap(pages, numPages, 0, gcmkNONPAGED_MEMROY_PROT(PAGE_KERNEL));
> +
> ++ /* Trigger a page fault. */
> ++ memset(addr, 0, numPages * PAGE_SIZE);
> ++
> + if (free)
> + {
> + kfree(pages);
> +@@ -1540,7 +1600,7 @@ gckOS_MapMemory(
> + #else
> + #if !gcdPAGED_MEMORY_CACHEABLE
> + mdlMap->vma->vm_page_prot = gcmkPAGED_MEMROY_PROT(mdlMap->vma->vm_page_prot);
> +- mdlMap->vma->vm_flags |= VM_IO | VM_DONTCOPY | VM_DONTEXPAND | VM_RESERVED;
> ++ mdlMap->vma->vm_flags |= gcdVM_FLAGS;
> + # endif
> + mdlMap->vma->vm_pgoff = 0;
> +
> +@@ -1987,7 +2047,7 @@ gckOS_AllocateNonPagedMemory(
> + }
> + #else
> + mdlMap->vma->vm_page_prot = gcmkNONPAGED_MEMROY_PROT(mdlMap->vma->vm_page_prot);
> +- mdlMap->vma->vm_flags |= VM_IO | VM_DONTCOPY | VM_DONTEXPAND | VM_RESERVED;
> ++ mdlMap->vma->vm_flags |= gcdVM_FLAGS;
> + mdlMap->vma->vm_pgoff = 0;
> +
> + if (remap_pfn_range(mdlMap->vma,
> +@@ -2367,12 +2427,18 @@ gckOS_GetPhysicalAddress(
> + gcmkVERIFY_OBJECT(Os, gcvOBJ_OS);
> + gcmkVERIFY_ARGUMENT(Address != gcvNULL);
> +
> +- /* Get current process ID. */
> +- processID = _GetProcessID();
> ++ /* Query page table of current process first. */
> ++ status = _QueryProcessPageTable(Logical, Address);
> +
> +- /* Route through other function. */
> +- gcmkONERROR(
> +- gckOS_GetPhysicalAddressProcess(Os, Logical, processID, Address));
> ++ if (gcmIS_ERROR(status))
> ++ {
> ++ /* Get current process ID. */
> ++ processID = _GetProcessID();
> ++
> ++ /* Route through other function. */
> ++ gcmkONERROR(
> ++ gckOS_GetPhysicalAddressProcess(Os, Logical, processID, Address));
> ++ }
> +
> + /* Success. */
> + gcmkFOOTER_ARG("*Address=0x%08x", *Address);
> +@@ -4139,7 +4205,7 @@ gckOS_LockPages(
> + return gcvSTATUS_OUT_OF_RESOURCES;
> + }
> +
> +- mdlMap->vma->vm_flags |= VM_RESERVED;
> ++ mdlMap->vma->vm_flags |= gcdVM_FLAGS;
> + #if !gcdPAGED_MEMORY_CACHEABLE
> + if (Cacheable == gcvFALSE)
> + {
> +--
> +1.8.3.2
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0001-perf-tools-Fix-getrusage-related-build-failure-on-gl.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0001-perf-tools-Fix-getrusage-related-build-failure-on-gl.patch
> new file mode 100644
> index 0000000..abc02c9
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0001-perf-tools-Fix-getrusage-related-build-failure-on-gl.patch
> @@ -0,0 +1,43 @@
> +From 503daf4789dd23e4dc1e16c256de0c163fc2bf87 Mon Sep 17 00:00:00 2001
> +From: Markus Trippelsdorf <markus at trippelsdorf.de>
> +Date: Wed, 4 Apr 2012 10:45:27 +0200
> +Subject: [PATCH] perf tools: Fix getrusage() related build failure on glibc
> + trunk
> +MIME-Version: 1.0
> +Content-Type: text/plain; charset=UTF-8
> +Content-Transfer-Encoding: 8bit
> +Organization: O.S. Systems Software LTDA.
> +
> +On a system running glibc trunk perf doesn't build:
> +
> + CC builtin-sched.o
> +builtin-sched.c: In function ‘get_cpu_usage_nsec_parent’: builtin-sched.c:399:16: error: storage size of ‘ru’ isn’t known builtin-sched.c:403:2: error: implicit declaration of function ‘getrusage’ [-Werror=implicit-function-declaration]
> + [...]
> +
> +Fix it by including sys/resource.h.
> +
> +Upstream-Status: Pending
> +
> +Signed-off-by: Markus Trippelsdorf <markus at trippelsdorf.de>
> +Cc: Peter Zijlstra <a.p.zijlstra at chello.nl>
> +Link: http://lkml.kernel.org/r/20120404084527.GA294@x4
> +Signed-off-by: Ingo Molnar <mingo at kernel.org>
> +---
> + tools/perf/builtin-sched.c | 1 +
> + 1 file changed, 1 insertion(+)
> +
> +diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
> +index dcfe887..3632c2f 100644
> +--- a/tools/perf/builtin-sched.c
> ++++ b/tools/perf/builtin-sched.c
> +@@ -14,6 +14,7 @@
> + #include "util/debug.h"
> +
> + #include <sys/prctl.h>
> ++#include <sys/resource.h>
> +
> + #include <semaphore.h>
> + #include <pthread.h>
> +--
> +1.8.4.rc3
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0002-ARM-7668-1-fix-memset-related-crashes-caused-by-rece.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0002-ARM-7668-1-fix-memset-related-crashes-caused-by-rece.patch
> new file mode 100644
> index 0000000..4c31e74
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0002-ARM-7668-1-fix-memset-related-crashes-caused-by-rece.patch
> @@ -0,0 +1,259 @@
> +From 2235b85f1c76d98b5f1e160cbd0a61a84c15e125 Mon Sep 17 00:00:00 2001
> +From: Ivan Djelic <ivan.djelic at parrot.com>
> +Date: Wed, 6 Mar 2013 20:09:27 +0100
> +Subject: [PATCH] ARM: 7668/1: fix memset-related crashes caused by recent GCC
> + (4.7.2) optimizations
> +Organization: O.S. Systems Software LTDA.
> +
> +Recent GCC versions (e.g. GCC-4.7.2) perform optimizations based on
> +assumptions about the implementation of memset and similar functions.
> +The current ARM optimized memset code does not return the value of
> +its first argument, as is usually expected from standard implementations.
> +
> +For instance in the following function:
> +
> +void debug_mutex_lock_common(struct mutex *lock, struct mutex_waiter *waiter)
> +{
> + memset(waiter, MUTEX_DEBUG_INIT, sizeof(*waiter));
> + waiter->magic = waiter;
> + INIT_LIST_HEAD(&waiter->list);
> +}
> +
> +compiled as:
> +
> +800554d0 <debug_mutex_lock_common>:
> +800554d0: e92d4008 push {r3, lr}
> +800554d4: e1a00001 mov r0, r1
> +800554d8: e3a02010 mov r2, #16 ; 0x10
> +800554dc: e3a01011 mov r1, #17 ; 0x11
> +800554e0: eb04426e bl 80165ea0 <memset>
> +800554e4: e1a03000 mov r3, r0
> +800554e8: e583000c str r0, [r3, #12]
> +800554ec: e5830000 str r0, [r3]
> +800554f0: e5830004 str r0, [r3, #4]
> +800554f4: e8bd8008 pop {r3, pc}
> +
> +GCC assumes memset returns the value of pointer 'waiter' in register r0; causing
> +register/memory corruptions.
> +
> +This patch fixes the return value of the assembly version of memset.
> +It adds a 'mov' instruction and merges an additional load+store into
> +existing load/store instructions.
> +For ease of review, here is a breakdown of the patch into 4 simple steps:
> +
> +Step 1
> +======
> +Perform the following substitutions:
> +ip -> r8, then
> +r0 -> ip,
> +and insert 'mov ip, r0' as the first statement of the function.
> +At this point, we have a memset() implementation returning the proper result,
> +but corrupting r8 on some paths (the ones that were using ip).
> +
> +Step 2
> +======
> +Make sure r8 is saved and restored when (! CALGN(1)+0) == 1:
> +
> +save r8:
> +- str lr, [sp, #-4]!
> ++ stmfd sp!, {r8, lr}
> +
> +and restore r8 on both exit paths:
> +- ldmeqfd sp!, {pc} @ Now <64 bytes to go.
> ++ ldmeqfd sp!, {r8, pc} @ Now <64 bytes to go.
> +(...)
> + tst r2, #16
> + stmneia ip!, {r1, r3, r8, lr}
> +- ldr lr, [sp], #4
> ++ ldmfd sp!, {r8, lr}
> +
> +Step 3
> +======
> +Make sure r8 is saved and restored when (! CALGN(1)+0) == 0:
> +
> +save r8:
> +- stmfd sp!, {r4-r7, lr}
> ++ stmfd sp!, {r4-r8, lr}
> +
> +and restore r8 on both exit paths:
> + bgt 3b
> +- ldmeqfd sp!, {r4-r7, pc}
> ++ ldmeqfd sp!, {r4-r8, pc}
> +(...)
> + tst r2, #16
> + stmneia ip!, {r4-r7}
> +- ldmfd sp!, {r4-r7, lr}
> ++ ldmfd sp!, {r4-r8, lr}
> +
> +Step 4
> +======
> +Rewrite register list "r4-r7, r8" as "r4-r8".
> +
> +Upstream-Status: Pending
> +
> +Signed-off-by: Ivan Djelic <ivan.djelic at parrot.com>
> +Reviewed-by: Nicolas Pitre <nico at linaro.org>
> +Signed-off-by: Dirk Behme <dirk.behme at gmail.com>
> +Signed-off-by: Russell King <rmk+kernel at arm.linux.org.uk>
> +(cherry picked from commit 455bd4c430b0c0a361f38e8658a0d6cb469942b5)
> +---
> + arch/arm/lib/memset.S | 85 ++++++++++++++++++++++++++-------------------------
> + 1 file changed, 44 insertions(+), 41 deletions(-)
> +
> +diff --git a/arch/arm/lib/memset.S b/arch/arm/lib/memset.S
> +index 650d592..d912e73 100644
> +--- a/arch/arm/lib/memset.S
> ++++ b/arch/arm/lib/memset.S
> +@@ -19,9 +19,9 @@
> + 1: subs r2, r2, #4 @ 1 do we have enough
> + blt 5f @ 1 bytes to align with?
> + cmp r3, #2 @ 1
> +- strltb r1, [r0], #1 @ 1
> +- strleb r1, [r0], #1 @ 1
> +- strb r1, [r0], #1 @ 1
> ++ strltb r1, [ip], #1 @ 1
> ++ strleb r1, [ip], #1 @ 1
> ++ strb r1, [ip], #1 @ 1
> + add r2, r2, r3 @ 1 (r2 = r2 - (4 - r3))
> + /*
> + * The pointer is now aligned and the length is adjusted. Try doing the
> +@@ -29,10 +29,14 @@
> + */
> +
> + ENTRY(memset)
> +- ands r3, r0, #3 @ 1 unaligned?
> ++/*
> ++ * Preserve the contents of r0 for the return value.
> ++ */
> ++ mov ip, r0
> ++ ands r3, ip, #3 @ 1 unaligned?
> + bne 1b @ 1
> + /*
> +- * we know that the pointer in r0 is aligned to a word boundary.
> ++ * we know that the pointer in ip is aligned to a word boundary.
> + */
> + orr r1, r1, r1, lsl #8
> + orr r1, r1, r1, lsl #16
> +@@ -43,29 +47,28 @@ ENTRY(memset)
> + #if ! CALGN(1)+0
> +
> + /*
> +- * We need an extra register for this loop - save the return address and
> +- * use the LR
> ++ * We need 2 extra registers for this loop - use r8 and the LR
> + */
> +- str lr, [sp, #-4]!
> +- mov ip, r1
> ++ stmfd sp!, {r8, lr}
> ++ mov r8, r1
> + mov lr, r1
> +
> + 2: subs r2, r2, #64
> +- stmgeia r0!, {r1, r3, ip, lr} @ 64 bytes at a time.
> +- stmgeia r0!, {r1, r3, ip, lr}
> +- stmgeia r0!, {r1, r3, ip, lr}
> +- stmgeia r0!, {r1, r3, ip, lr}
> ++ stmgeia ip!, {r1, r3, r8, lr} @ 64 bytes at a time.
> ++ stmgeia ip!, {r1, r3, r8, lr}
> ++ stmgeia ip!, {r1, r3, r8, lr}
> ++ stmgeia ip!, {r1, r3, r8, lr}
> + bgt 2b
> +- ldmeqfd sp!, {pc} @ Now <64 bytes to go.
> ++ ldmeqfd sp!, {r8, pc} @ Now <64 bytes to go.
> + /*
> + * No need to correct the count; we're only testing bits from now on
> + */
> + tst r2, #32
> +- stmneia r0!, {r1, r3, ip, lr}
> +- stmneia r0!, {r1, r3, ip, lr}
> ++ stmneia ip!, {r1, r3, r8, lr}
> ++ stmneia ip!, {r1, r3, r8, lr}
> + tst r2, #16
> +- stmneia r0!, {r1, r3, ip, lr}
> +- ldr lr, [sp], #4
> ++ stmneia ip!, {r1, r3, r8, lr}
> ++ ldmfd sp!, {r8, lr}
> +
> + #else
> +
> +@@ -74,54 +77,54 @@ ENTRY(memset)
> + * whole cache lines at once.
> + */
> +
> +- stmfd sp!, {r4-r7, lr}
> ++ stmfd sp!, {r4-r8, lr}
> + mov r4, r1
> + mov r5, r1
> + mov r6, r1
> + mov r7, r1
> +- mov ip, r1
> ++ mov r8, r1
> + mov lr, r1
> +
> + cmp r2, #96
> +- tstgt r0, #31
> ++ tstgt ip, #31
> + ble 3f
> +
> +- and ip, r0, #31
> +- rsb ip, ip, #32
> +- sub r2, r2, ip
> +- movs ip, ip, lsl #(32 - 4)
> +- stmcsia r0!, {r4, r5, r6, r7}
> +- stmmiia r0!, {r4, r5}
> +- tst ip, #(1 << 30)
> +- mov ip, r1
> +- strne r1, [r0], #4
> ++ and r8, ip, #31
> ++ rsb r8, r8, #32
> ++ sub r2, r2, r8
> ++ movs r8, r8, lsl #(32 - 4)
> ++ stmcsia ip!, {r4, r5, r6, r7}
> ++ stmmiia ip!, {r4, r5}
> ++ tst r8, #(1 << 30)
> ++ mov r8, r1
> ++ strne r1, [ip], #4
> +
> + 3: subs r2, r2, #64
> +- stmgeia r0!, {r1, r3-r7, ip, lr}
> +- stmgeia r0!, {r1, r3-r7, ip, lr}
> ++ stmgeia ip!, {r1, r3-r8, lr}
> ++ stmgeia ip!, {r1, r3-r8, lr}
> + bgt 3b
> +- ldmeqfd sp!, {r4-r7, pc}
> ++ ldmeqfd sp!, {r4-r8, pc}
> +
> + tst r2, #32
> +- stmneia r0!, {r1, r3-r7, ip, lr}
> ++ stmneia ip!, {r1, r3-r8, lr}
> + tst r2, #16
> +- stmneia r0!, {r4-r7}
> +- ldmfd sp!, {r4-r7, lr}
> ++ stmneia ip!, {r4-r7}
> ++ ldmfd sp!, {r4-r8, lr}
> +
> + #endif
> +
> + 4: tst r2, #8
> +- stmneia r0!, {r1, r3}
> ++ stmneia ip!, {r1, r3}
> + tst r2, #4
> +- strne r1, [r0], #4
> ++ strne r1, [ip], #4
> + /*
> + * When we get here, we've got less than 4 bytes to zero. We
> + * may have an unaligned pointer as well.
> + */
> + 5: tst r2, #2
> +- strneb r1, [r0], #1
> +- strneb r1, [r0], #1
> ++ strneb r1, [ip], #1
> ++ strneb r1, [ip], #1
> + tst r2, #1
> +- strneb r1, [r0], #1
> ++ strneb r1, [ip], #1
> + mov pc, lr
> + ENDPROC(memset)
> +--
> +1.8.4.rc3
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0002-ENGR00265465-gpu-Add-global-value-for-minimum-3D-clo.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0002-ENGR00265465-gpu-Add-global-value-for-minimum-3D-clo.patch
> new file mode 100644
> index 0000000..5725ab7
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0002-ENGR00265465-gpu-Add-global-value-for-minimum-3D-clo.patch
> @@ -0,0 +1,62 @@
> +From 2df4dba8faa9a781a5a1c6c09d646d2b692c9a0c Mon Sep 17 00:00:00 2001
> +From: Loren Huang <b02279 at freescale.com>
> +Date: Tue, 4 Jun 2013 15:08:15 +0800
> +Subject: [PATCH 2/6] ENGR00265465 gpu:Add global value for minimum 3D clock
> + export
> +
> +Add global value gpu3DMinClock so that minimum 3D clock can be change by user.
> +When gpu min clock is too low, it may cause IPU starvation issue in certain case.
> +Use echo x > /sys/module/galcore/parameters/gpu3DMinClock to change it.
> +
> +Cherry-pick from 3.0.35 branch.
> +
> +Upstream-Status: Backport [3.5.7-1.0.0]
> +
> +Signed-off-by: Loren Huang <b02279 at freescale.com>
> +Acked-by: Lily Zhang
> +---
> + drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c | 6 +++++-
> + drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c | 3 +++
> + 2 files changed, 8 insertions(+), 1 deletion(-)
> +
> +diff --git a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> +index 3829999..ebd36fe 100644
> +--- a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> ++++ b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> +@@ -36,6 +36,7 @@ typedef struct _gcsiDEBUG_REGISTERS
> + }
> + gcsiDEBUG_REGISTERS;
> +
> ++extern int gpu3DMinClock;
> + /******************************************************************************\
> + ********************************* Support Code *********************************
> + \******************************************************************************/
> +@@ -4630,7 +4631,10 @@ gckHARDWARE_GetFscaleValue(
> + )
> + {
> + *FscaleValue = Hardware->powerOnFscaleVal;
> +- *MinFscaleValue = 1;
> ++ if ((gpu3DMinClock > 0) && (gpu3DMinClock <= 64) && (Hardware->core == gcvCORE_MAJOR))
> ++ *MinFscaleValue = gpu3DMinClock;
> ++ else
> ++ *MinFscaleValue = 1;
> + *MaxFscaleValue = 64;
> +
> + return gcvSTATUS_OK;
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +index 2ed3d0e..64cace1 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +@@ -146,6 +146,9 @@ module_param(logFileSize,uint, 0644);
> + static int showArgs = 0;
> + module_param(showArgs, int, 0644);
> +
> ++int gpu3DMinClock = 0;
> ++module_param(gpu3DMinClock, int, 0644);
> ++
> + #if ENABLE_GPU_CLOCK_BY_DRIVER
> + unsigned long coreClock = 156000000;
> + module_param(coreClock, ulong, 0644);
> +--
> +1.8.3.2
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0003-ARM-7670-1-fix-the-memset-fix.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0003-ARM-7670-1-fix-the-memset-fix.patch
> new file mode 100644
> index 0000000..b8d6f53
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0003-ARM-7670-1-fix-the-memset-fix.patch
> @@ -0,0 +1,87 @@
> +From 2ba23fa6c4128febaaf57fe184420a7111caa237 Mon Sep 17 00:00:00 2001
> +From: Nicolas Pitre <nicolas.pitre at linaro.org>
> +Date: Tue, 12 Mar 2013 13:00:42 +0100
> +Subject: [PATCH] ARM: 7670/1: fix the memset fix
> +Organization: O.S. Systems Software LTDA.
> +
> +Commit 455bd4c430b0 ("ARM: 7668/1: fix memset-related crashes caused by
> +recent GCC (4.7.2) optimizations") attempted to fix a compliance issue
> +with the memset return value. However the memset itself became broken
> +by that patch for misaligned pointers.
> +
> +This fixes the above by branching over the entry code from the
> +misaligned fixup code to avoid reloading the original pointer.
> +
> +Also, because the function entry alignment is wrong in the Thumb mode
> +compilation, that fixup code is moved to the end.
> +
> +While at it, the entry instructions are slightly reworked to help dual
> +issue pipelines.
> +
> +Upstream-Status: Pending
> +
> +Signed-off-by: Nicolas Pitre <nico at linaro.org>
> +Tested-by: Alexander Holler <holler at ahsoftware.de>
> +Signed-off-by: Russell King <rmk+kernel at arm.linux.org.uk>
> +(cherry picked from commit 418df63adac56841ef6b0f1fcf435bc64d4ed177)
> +---
> + arch/arm/lib/memset.S | 33 +++++++++++++--------------------
> + 1 file changed, 13 insertions(+), 20 deletions(-)
> +
> +diff --git a/arch/arm/lib/memset.S b/arch/arm/lib/memset.S
> +index d912e73..94b0650 100644
> +--- a/arch/arm/lib/memset.S
> ++++ b/arch/arm/lib/memset.S
> +@@ -14,31 +14,15 @@
> +
> + .text
> + .align 5
> +- .word 0
> +-
> +-1: subs r2, r2, #4 @ 1 do we have enough
> +- blt 5f @ 1 bytes to align with?
> +- cmp r3, #2 @ 1
> +- strltb r1, [ip], #1 @ 1
> +- strleb r1, [ip], #1 @ 1
> +- strb r1, [ip], #1 @ 1
> +- add r2, r2, r3 @ 1 (r2 = r2 - (4 - r3))
> +-/*
> +- * The pointer is now aligned and the length is adjusted. Try doing the
> +- * memset again.
> +- */
> +
> + ENTRY(memset)
> +-/*
> +- * Preserve the contents of r0 for the return value.
> +- */
> +- mov ip, r0
> +- ands r3, ip, #3 @ 1 unaligned?
> +- bne 1b @ 1
> ++ ands r3, r0, #3 @ 1 unaligned?
> ++ mov ip, r0 @ preserve r0 as return value
> ++ bne 6f @ 1
> + /*
> + * we know that the pointer in ip is aligned to a word boundary.
> + */
> +- orr r1, r1, r1, lsl #8
> ++1: orr r1, r1, r1, lsl #8
> + orr r1, r1, r1, lsl #16
> + mov r3, r1
> + cmp r2, #16
> +@@ -127,4 +111,13 @@ ENTRY(memset)
> + tst r2, #1
> + strneb r1, [ip], #1
> + mov pc, lr
> ++
> ++6: subs r2, r2, #4 @ 1 do we have enough
> ++ blt 5b @ 1 bytes to align with?
> ++ cmp r3, #2 @ 1
> ++ strltb r1, [ip], #1 @ 1
> ++ strleb r1, [ip], #1 @ 1
> ++ strb r1, [ip], #1 @ 1
> ++ add r2, r2, r3 @ 1 (r2 = r2 - (4 - r3))
> ++ b 1b
> + ENDPROC(memset)
> +--
> +1.8.4.rc3
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0003-ENGR00261814-4-gpu-use-new-PU-power-on-off-interface.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0003-ENGR00261814-4-gpu-use-new-PU-power-on-off-interface.patch
> new file mode 100644
> index 0000000..23a415d
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0003-ENGR00261814-4-gpu-use-new-PU-power-on-off-interface.patch
> @@ -0,0 +1,53 @@
> +From 1579de9397783ab5321c80f1e76661653ef38ccd Mon Sep 17 00:00:00 2001
> +From: Robin Gong <b38343 at freescale.com>
> +Date: Thu, 9 May 2013 11:45:55 +0800
> +Subject: [PATCH 3/6] ENGR00261814-4 gpu: use new PU power on/off interface
> +
> +use new PU power on/off interface in GPU driver
> +
> +Upstream-Status: Backport [3.5.7-1.0.0]
> +
> +Signed-off-by: Robin Gong <b38343 at freescale.com>
> +Acked-by: Lily Zhang
> +---
> + .../mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c | 16 +++++++++++++---
> + 1 file changed, 13 insertions(+), 3 deletions(-)
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> +index 9c2bae6..dfbc699 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> +@@ -6819,8 +6819,13 @@ gckOS_SetGPUPower(
> + }
> + if((Power == gcvTRUE) && (oldPowerState == gcvFALSE))
> + {
> +- if(!IS_ERR(Os->device->gpu_regulator))
> +- regulator_enable(Os->device->gpu_regulator);
> ++#if LINUX_VERSION_CODE < KERNEL_VERSION(3,5,0)
> ++ if(!IS_ERR(Os->device->gpu_regulator))
> ++ regulator_enable(Os->device->gpu_regulator);
> ++#else
> ++ imx_gpc_power_up_pu(true);
> ++#endif
> ++
> + #ifdef CONFIG_PM
> + pm_runtime_get_sync(Os->device->pmdev);
> + #endif
> +@@ -6930,8 +6935,13 @@ gckOS_SetGPUPower(
> + #ifdef CONFIG_PM
> + pm_runtime_put_sync(Os->device->pmdev);
> + #endif
> ++
> ++#if LINUX_VERSION_CODE < KERNEL_VERSION(3,5,0)
> + if(!IS_ERR(Os->device->gpu_regulator))
> +- regulator_disable(Os->device->gpu_regulator);
> ++ regulator_disable(Os->device->gpu_regulator);
> ++#else
> ++ imx_gpc_power_up_pu(false);
> ++#endif
> + }
> + /* TODO: Put your code here. */
> + gcmkFOOTER_NO();
> +--
> +1.8.3.2
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0004-ENGR00264288-1-GPU-Integrate-4.6.9p12-release-kernel.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0004-ENGR00264288-1-GPU-Integrate-4.6.9p12-release-kernel.patch
> new file mode 100644
> index 0000000..08ca88a
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0004-ENGR00264288-1-GPU-Integrate-4.6.9p12-release-kernel.patch
> @@ -0,0 +1,2006 @@
> +From c090a0238315094d245de2503b6f9a5bce0bda03 Mon Sep 17 00:00:00 2001
> +From: Loren Huang <b02279 at freescale.com>
> +Date: Mon, 27 May 2013 17:45:48 +0800
> +Subject: [PATCH 4/6] ENGR00264288-1 [GPU]Integrate 4.6.9p12 release kernel
> + part code
> +
> +Integrate 4.6.9p12 release kernel part code.
> +Cherry-pick from 3.0.35 branch.
> +
> +Upstream-Status: Backport [3.5.7-1.0.0]
> +
> +Signed-off-by: Loren Huang <b02279 at freescale.com>
> +Acked-by: Lily Zhang
> +---
> + .../GC350/hal/kernel/gc_hal_kernel_hardware_vg.c | 63 +++-
> + .../GC350/hal/kernel/gc_hal_kernel_hardware_vg.h | 2 +
> + .../arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c | 174 ++++++++---
> + .../arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.h | 2 +
> + drivers/mxc/gpu-viv/config | 1 -
> + drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.c | 329 +++++++--------------
> + .../mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c | 6 +-
> + .../gpu-viv/hal/kernel/gc_hal_kernel_command_vg.c | 14 +-
> + drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c | 6 +-
> + drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c | 119 +++++++-
> + .../hal/kernel/gc_hal_kernel_video_memory.c | 3 +
> + drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal.h | 6 +
> + drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_base.h | 34 +--
> + .../mxc/gpu-viv/hal/kernel/inc/gc_hal_compiler.h | 20 +-
> + drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_driver.h | 35 ++-
> + drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h | 62 +---
> + drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_enum.h | 7 -
> + .../mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h | 31 +-
> + .../mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h | 2 +-
> + drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_vg.h | 6 +
> + .../hal/os/linux/kernel/gc_hal_kernel_device.c | 13 +
> + .../hal/os/linux/kernel/gc_hal_kernel_device.h | 1 +
> + .../hal/os/linux/kernel/gc_hal_kernel_driver.c | 10 +-
> + .../gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c | 74 +++--
> + .../gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.h | 3 +
> + 25 files changed, 574 insertions(+), 449 deletions(-)
> +
> +diff --git a/drivers/mxc/gpu-viv/arch/GC350/hal/kernel/gc_hal_kernel_hardware_vg.c b/drivers/mxc/gpu-viv/arch/GC350/hal/kernel/gc_hal_kernel_hardware_vg.c
> +index 4a6010d..70c2cd6 100644
> +--- a/drivers/mxc/gpu-viv/arch/GC350/hal/kernel/gc_hal_kernel_hardware_vg.c
> ++++ b/drivers/mxc/gpu-viv/arch/GC350/hal/kernel/gc_hal_kernel_hardware_vg.c
> +@@ -217,7 +217,6 @@ _IdentifyHardware(
> + return status;
> + }
> +
> +-#if gcdPOWER_MANAGEMENT
> + static gctTHREADFUNCRESULT gctTHREADFUNCTYPE
> + _TimeIdleThread(
> + gctTHREADFUNCPARAMETER ThreadParameter
> +@@ -262,8 +261,6 @@ _TimeIdleThread(
> + }
> + return 0;
> + }
> +-#endif
> +-
> +
> + /******************************************************************************\
> + ****************************** gckVGHARDWARE API code *****************************
> +@@ -309,6 +306,7 @@ gckVGHARDWARE_Construct(
> + do
> + {
> + gcmkERR_BREAK(gckOS_SetGPUPower(Os, gcvCORE_VG, gcvTRUE, gcvTRUE));
> ++
> + status = _ResetGPU(Os);
> +
> + if (status != gcvSTATUS_OK)
> +@@ -368,14 +366,17 @@ gckVGHARDWARE_Construct(
> +
> + gcmkERR_BREAK(gckOS_CreateMutex(Os, &hardware->powerMutex));
> + gcmkERR_BREAK(gckOS_CreateSignal(Os, gcvFALSE, &hardware->idleSignal));
> +-#if gcdPOWER_MANAGEMENT
> ++
> ++ /* Enable power management by default. */
> ++ hardware->powerManagement = gcvTRUE;
> ++
> + gcmkERR_BREAK(gckOS_StartThread(
> + hardware->os,
> + _TimeIdleThread,
> + hardware,
> + &hardware->timeIdleThread
> + ));
> +-#endif
> ++
> + /* Return pointer to the gckVGHARDWARE object. */
> + *Hardware = hardware;
> +
> +@@ -395,6 +396,8 @@ gckVGHARDWARE_Construct(
> + gcmkVERIFY_OK(gckOS_Free(Os, hardware));
> + }
> +
> ++ gcmkVERIFY_OK(gckOS_SetGPUPower(Os, gcvCORE_VG, gcvFALSE, gcvFALSE));
> ++
> + gcmkFOOTER();
> + /* Return the status. */
> + return status;
> +@@ -425,11 +428,10 @@ gckVGHARDWARE_Destroy(
> + /* Verify the arguments. */
> + gcmkVERIFY_OBJECT(Hardware, gcvOBJ_HARDWARE);
> +
> +-#if gcdPOWER_MANAGEMENT
> + Hardware->killThread = gcvTRUE;
> + gcmkVERIFY_OK(gckOS_Signal(Hardware->os, Hardware->idleSignal, gcvTRUE));
> + gcmkVERIFY_OK(gckOS_StopThread(Hardware->os, Hardware->timeIdleThread));
> +-#endif
> ++
> + /* Mark the object as unknown. */
> + Hardware->object.type = gcvOBJ_UNKNOWN;
> +
> +@@ -1432,7 +1434,6 @@ gckVGHARDWARE_ReadInterrupt(
> + return status;
> + }
> +
> +-#if gcdPOWER_MANAGEMENT
> + static gceSTATUS _CommandStall(
> + gckVGHARDWARE Hardware)
> + {
> +@@ -1477,7 +1478,6 @@ static gceSTATUS _CommandStall(
> + /* Return the status. */
> + return status;
> + }
> +-#endif
> +
> + /*******************************************************************************
> + **
> +@@ -1500,7 +1500,6 @@ gckVGHARDWARE_SetPowerManagementState(
> + IN gceCHIPPOWERSTATE State
> + )
> + {
> +-#if gcdPOWER_MANAGEMENT
> + gceSTATUS status;
> + gckVGCOMMAND command = gcvNULL;
> + gckOS os;
> +@@ -1600,6 +1599,12 @@ gckVGHARDWARE_SetPowerManagementState(
> + command = Hardware->kernel->command;
> + gcmkVERIFY_OBJECT(command, gcvOBJ_COMMAND);
> +
> ++ if (Hardware->powerManagement == gcvFALSE)
> ++ {
> ++ gcmkFOOTER_NO();
> ++ return gcvSTATUS_OK;
> ++ }
> ++
> + /* Start profiler. */
> + gcmkPROFILE_INIT(freq, time);
> +
> +@@ -1914,10 +1919,6 @@ OnError:
> + /* Return the status. */
> + gcmkFOOTER();
> + return status;
> +-#else /* gcdPOWER_MANAGEMENT */
> +- /* Do nothing */
> +- return gcvSTATUS_OK;
> +-#endif
> + }
> +
> + /*******************************************************************************
> +@@ -1955,6 +1956,40 @@ gckVGHARDWARE_QueryPowerManagementState(
> + return gcvSTATUS_OK;
> + }
> +
> ++/*******************************************************************************
> ++**
> ++** gckVGHARDWARE_SetPowerManagement
> ++**
> ++** Configure GPU power management function.
> ++** Only used in driver initialization stage.
> ++**
> ++** INPUT:
> ++**
> ++** gckVGHARDWARE Harwdare
> ++** Pointer to an gckHARDWARE object.
> ++**
> ++** gctBOOL PowerManagement
> ++** Power Mangement State.
> ++**
> ++*/
> ++gceSTATUS
> ++gckVGHARDWARE_SetPowerManagement(
> ++ IN gckVGHARDWARE Hardware,
> ++ IN gctBOOL PowerManagement
> ++ )
> ++{
> ++ gcmkHEADER_ARG("Hardware=0x%x", Hardware);
> ++
> ++ /* Verify the arguments. */
> ++ gcmkVERIFY_OBJECT(Hardware, gcvOBJ_HARDWARE);
> ++
> ++ Hardware->powerManagement = PowerManagement;
> ++
> ++ /* Success. */
> ++ gcmkFOOTER_NO();
> ++ return gcvSTATUS_OK;
> ++}
> ++
> + gceSTATUS
> + gckVGHARDWARE_SetPowerOffTimeout(
> + IN gckVGHARDWARE Hardware,
> +diff --git a/drivers/mxc/gpu-viv/arch/GC350/hal/kernel/gc_hal_kernel_hardware_vg.h b/drivers/mxc/gpu-viv/arch/GC350/hal/kernel/gc_hal_kernel_hardware_vg.h
> +index 83a603e..16b81ae 100644
> +--- a/drivers/mxc/gpu-viv/arch/GC350/hal/kernel/gc_hal_kernel_hardware_vg.h
> ++++ b/drivers/mxc/gpu-viv/arch/GC350/hal/kernel/gc_hal_kernel_hardware_vg.h
> +@@ -66,6 +66,8 @@ struct _gckVGHARDWARE
> + gctTHREAD timeIdleThread;
> + gctBOOL killThread;
> + gctPOINTER pageTableDirty;
> ++
> ++ gctBOOL powerManagement;
> + };
> +
> + #endif /* __gc_hal_kernel_hardware_h_ */
> +diff --git a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> +index ebd36fe..00f3839 100644
> +--- a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> ++++ b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.c
> +@@ -176,6 +176,7 @@ _IdentifyHardware(
> + Identity->chipMinorFeatures1 = 0;
> + Identity->chipMinorFeatures2 = 0;
> + Identity->chipMinorFeatures3 = 0;
> ++ Identity->chipMinorFeatures4 = 0;
> + }
> + else
> + {
> +@@ -207,13 +208,20 @@ _IdentifyHardware(
> + gckOS_ReadRegisterEx(Os, Core,
> + 0x00088,
> + &Identity->chipMinorFeatures3));
> ++
> ++ /* Read chip minor featuress register #4. */
> ++ gcmkONERROR(
> ++ gckOS_ReadRegisterEx(Os, Core,
> ++ 0x00094,
> ++ &Identity->chipMinorFeatures4));
> + }
> + else
> + {
> +- /* Chip doesn't has minor features register #1 or 2 or 3. */
> ++ /* Chip doesn't has minor features register #1 or 2 or 3 or 4. */
> + Identity->chipMinorFeatures1 = 0;
> + Identity->chipMinorFeatures2 = 0;
> + Identity->chipMinorFeatures3 = 0;
> ++ Identity->chipMinorFeatures4 = 0;
> + }
> + }
> +
> +@@ -234,14 +242,14 @@ _IdentifyHardware(
> +
> + /* Exception for GC1000, revision 5035 & GC800, revision 4612 */
> + if (((Identity->chipModel == gcv1000) && ((Identity->chipRevision == 0x5035)
> +- || (Identity->chipRevision == 0x5036)))
> ++ || (Identity->chipRevision == 0x5036)
> ++ || (Identity->chipRevision == 0x5037)))
> + || ((Identity->chipModel == gcv800) && (Identity->chipRevision == 0x4612)))
> + {
> + Identity->superTileMode = 1;
> + }
> +
> +
> +-
> + /* Disable HZ when EZ is present for older chips. */
> + if (!((((gctUINT32) (Identity->chipFeatures)) >> (0 ? 16:16) & ((gctUINT32) ((((1 ? 16:16) - (0 ? 16:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 16:16) - (0 ? 16:16) + 1)))))) == (0x1 & ((gctUINT32) ((((1 ? 16:16) - (0 ? 16:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 16:16) - (0 ? 16:16) + 1))))))))
> + {
> +@@ -285,6 +293,10 @@ _IdentifyHardware(
> + "Identity: chipMinorFeatures3=0x%08X",
> + Identity->chipMinorFeatures3);
> +
> ++ gcmkTRACE_ZONE(gcvLEVEL_INFO, gcvZONE_HARDWARE,
> ++ "Identity: chipMinorFeatures4=0x%08X",
> ++ Identity->chipMinorFeatures4);
> ++
> + /***************************************************************************
> + ** Get chip specs.
> + */
> +@@ -576,7 +588,6 @@ OnError:
> + return status;
> + }
> +
> +-#if gcdPOWER_MANAGEMENT
> + static gceSTATUS
> + _IsGPUPresent(
> + IN gckHARDWARE Hardware
> +@@ -631,7 +642,6 @@ OnError:
> + gcmkFOOTER();
> + return status;
> + }
> +-#endif
> +
> + /******************************************************************************\
> + ****************************** gckHARDWARE API code *****************************
> +@@ -708,6 +718,7 @@ gckHARDWARE_Construct(
> +
> + case gcv300:
> + case gcv320:
> ++ case gcv420:
> + hardware->type = gcvHARDWARE_2D;
> + /*set outstanding limit*/
> + gcmkONERROR(gckOS_ReadRegisterEx(Os, Core, 0x00414, &axi_ot));
> +@@ -795,6 +806,9 @@ gckHARDWARE_Construct(
> + hardware->linkQueue.count = 0;
> + #endif
> +
> ++ /* Enable power management by default. */
> ++ hardware->powerManagement = gcvTRUE;
> ++
> + /* Return pointer to the gckHARDWARE object. */
> + *Hardware = hardware;
> +
> +@@ -1404,6 +1418,7 @@ gckHARDWARE_QueryChipIdentity(
> + Identity->chipMinorFeatures1 = Hardware->identity.chipMinorFeatures1;
> + Identity->chipMinorFeatures2 = Hardware->identity.chipMinorFeatures2;
> + Identity->chipMinorFeatures3 = Hardware->identity.chipMinorFeatures3;
> ++ Identity->chipMinorFeatures4 = Hardware->identity.chipMinorFeatures4;
> +
> + /* Return chip specs. */
> + Identity->streamCount = Hardware->identity.streamCount;
> +@@ -3129,7 +3144,7 @@ gckHARDWARE_FlushMMU(
> + | ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 25:16) - (0 ? 25:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 25:16) - (0 ? 25:16) + 1))))))) << (0 ? 25:16))) | (((gctUINT32) ((gctUINT32) (1) & ((gctUINT32) ((((1 ? 25:16) - (0 ? 25:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 25:16) - (0 ? 25:16) + 1))))))) << (0 ? 25:16)));
> +
> + buffer[9]
> +- = (((((gctUINT32) (~0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 4:4) - (0 ? 4:4) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 4:4) - (0 ? 4:4) + 1))))))) << (0 ? 4:4))) | (((gctUINT32) (0x1 & ((gctUINT32) ((((1 ? 4:4) - (0 ? 4:4) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 4:4) - (0 ? 4:4) + 1))))))) << (0 ? 4:4))) & ((((gctUINT32) (~0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 7:7) - (0 ? 7:7) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 7:7) - (0 ? 7:7) + 1))))))) << (0 ? 7:7))) | (((gctUINT32) (0x0 & ((gctUINT32) ((((1 ? 7:7) - (0 ? 7:7) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 7:7) - (0 ? 7:7) + 1))))))) << (0 ? 7:7))) );
> ++ = (((((gctUINT32) (~0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 4:4) - (0 ? 4:4) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 4:4) - (0 ? 4:4) + 1))))))) << (0 ? 4:4))) | (((gctUINT32) (0x1 & ((gctUINT32) ((((1 ? 4:4) - (0 ? 4:4) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 4:4) - (0 ? 4:4) + 1))))))) << (0 ? 4:4))) & ((((gctUINT32) (~0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 7:7) - (0 ? 7:7) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 7:7) - (0 ? 7:7) + 1))))))) << (0 ? 7:7))) | (((gctUINT32) (0x0 & ((gctUINT32) ((((1 ? 7:7) - (0 ? 7:7) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 7:7) - (0 ? 7:7) + 1))))))) << (0 ? 7:7))));
> +
> + /* Arm the PE-FE Semaphore. */
> + buffer[10]
> +@@ -3660,7 +3675,7 @@ typedef enum
> + }
> + gcePOWER_FLAGS;
> +
> +-#if gcmIS_DEBUG(gcdDEBUG_TRACE) && gcdPOWER_MANAGEMENT
> ++#if gcmIS_DEBUG(gcdDEBUG_TRACE)
> + static gctCONST_STRING
> + _PowerEnum(gceCHIPPOWERSTATE State)
> + {
> +@@ -3709,7 +3724,6 @@ gckHARDWARE_SetPowerManagementState(
> + IN gceCHIPPOWERSTATE State
> + )
> + {
> +-#if gcdPOWER_MANAGEMENT
> + gceSTATUS status;
> + gckCOMMAND command = gcvNULL;
> + gckOS os;
> +@@ -3841,6 +3855,12 @@ gckHARDWARE_SetPowerManagementState(
> + command = Hardware->kernel->command;
> + gcmkVERIFY_OBJECT(command, gcvOBJ_COMMAND);
> +
> ++ if (Hardware->powerManagement == gcvFALSE)
> ++ {
> ++ gcmkFOOTER_NO();
> ++ return gcvSTATUS_OK;
> ++ }
> ++
> + /* Start profiler. */
> + gcmkPROFILE_INIT(freq, time);
> +
> +@@ -4491,10 +4511,6 @@ OnError:
> + /* Return the status. */
> + gcmkFOOTER();
> + return status;
> +-#else /* gcdPOWER_MANAGEMENT */
> +- /* Do nothing */
> +- return gcvSTATUS_OK;
> +-#endif
> + }
> +
> + /*******************************************************************************
> +@@ -4532,6 +4548,40 @@ gckHARDWARE_QueryPowerManagementState(
> + return gcvSTATUS_OK;
> + }
> +
> ++/*******************************************************************************
> ++**
> ++** gckHARDWARE_SetPowerManagement
> ++**
> ++** Configure GPU power management function.
> ++** Only used in driver initialization stage.
> ++**
> ++** INPUT:
> ++**
> ++** gckHARDWARE Harwdare
> ++** Pointer to an gckHARDWARE object.
> ++**
> ++** gctBOOL PowerManagement
> ++** Power Mangement State.
> ++**
> ++*/
> ++gceSTATUS
> ++gckHARDWARE_SetPowerManagement(
> ++ IN gckHARDWARE Hardware,
> ++ IN gctBOOL PowerManagement
> ++ )
> ++{
> ++ gcmkHEADER_ARG("Hardware=0x%x", Hardware);
> ++
> ++ /* Verify the arguments. */
> ++ gcmkVERIFY_OBJECT(Hardware, gcvOBJ_HARDWARE);
> ++
> ++ Hardware->powerManagement = PowerManagement;
> ++
> ++ /* Success. */
> ++ gcmkFOOTER_NO();
> ++ return gcvSTATUS_OK;
> ++}
> ++
> + #if gcdENABLE_FSCALE_VAL_ADJUST
> + gceSTATUS
> + gckHARDWARE_SetFscaleValue(
> +@@ -4767,6 +4817,21 @@ OnError:
> + GC_DEBUG_SIGNALS_##block##_Address, \
> + &profiler->data))
> +
> ++#define gcmkREAD_DEBUG_REGISTER_N(control, block, index, data) \
> ++ gcmkONERROR(\
> ++ gckOS_WriteRegisterEx(Hardware->os, \
> ++ Hardware->core, \
> ++ GC_DEBUG_CONTROL##control##_Address, \
> ++ gcmSETFIELD(0, \
> ++ GC_DEBUG_CONTROL##control, \
> ++ block, \
> ++ index))); \
> ++ gcmkONERROR(\
> ++ gckOS_ReadRegisterEx(Hardware->os, \
> ++ Hardware->core, \
> ++ GC_DEBUG_SIGNALS_##block##_Address, \
> ++ &data))
> ++
> + #define gcmkRESET_DEBUG_REGISTER(control, block) \
> + gcmkONERROR(\
> + gckOS_WriteRegisterEx(Hardware->os, \
> +@@ -4857,6 +4922,9 @@ gckHARDWARE_QueryProfileRegisters(
> + {
> + gceSTATUS status;
> + gcsPROFILER_COUNTERS * profiler = Counters;
> ++ gctUINT i, clock;
> ++ gctUINT32 colorKilled, colorDrawn, depthKilled, depthDrawn;
> ++ gctUINT32 totalRead, totalWrite;
> +
> + gcmkHEADER_ARG("Hardware=0x%x Counters=0x%x", Hardware, Counters);
> +
> +@@ -4867,16 +4935,6 @@ gckHARDWARE_QueryProfileRegisters(
> + gcmkONERROR(
> + gckOS_ReadRegisterEx(Hardware->os,
> + Hardware->core,
> +- 0x00040,
> +- &profiler->gpuTotalRead64BytesPerFrame));
> +- gcmkONERROR(
> +- gckOS_ReadRegisterEx(Hardware->os,
> +- Hardware->core,
> +- 0x00044,
> +- &profiler->gpuTotalWrite64BytesPerFrame));
> +- gcmkONERROR(
> +- gckOS_ReadRegisterEx(Hardware->os,
> +- Hardware->core,
> + 0x00438,
> + &profiler->gpuCyclesCounter));
> +
> +@@ -4892,8 +4950,63 @@ gckHARDWARE_QueryProfileRegisters(
> + 0x0007C,
> + &profiler->gpuIdleCyclesCounter));
> +
> +- if(Reset){
> +
> ++ /* Read clock control register. */
> ++ gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os,
> ++ Hardware->core,
> ++ 0x00000,
> ++ &clock));
> ++
> ++ profiler->gpuTotalRead64BytesPerFrame = 0;
> ++ profiler->gpuTotalWrite64BytesPerFrame = 0;
> ++ profiler->pe_pixel_count_killed_by_color_pipe = 0;
> ++ profiler->pe_pixel_count_killed_by_depth_pipe = 0;
> ++ profiler->pe_pixel_count_drawn_by_color_pipe = 0;
> ++ profiler->pe_pixel_count_drawn_by_depth_pipe = 0;
> ++
> ++ /* Walk through all avaiable pixel pipes. */
> ++ for (i = 0; i < Hardware->identity.pixelPipes; ++i)
> ++ {
> ++ /* Select proper pipe. */
> ++ gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os,
> ++ Hardware->core,
> ++ 0x00000,
> ++ ((((gctUINT32) (clock)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:20) - (0 ? 23:20) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:20) - (0 ? 23:20) + 1))))))) << (0 ? 23:20))) | (((gctUINT32) ((gctUINT32) (i) & ((gctUINT32) ((((1 ? 23:20) - (0 ? 23:20) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:20) - (0 ? 23:20) + 1))))))) << (0 ? 23:20)))));
> ++
> ++ /* BW */
> ++ gcmkONERROR(
> ++ gckOS_ReadRegisterEx(Hardware->os,
> ++ Hardware->core,
> ++ 0x00040,
> ++ &totalRead));
> ++ gcmkONERROR(
> ++ gckOS_ReadRegisterEx(Hardware->os,
> ++ Hardware->core,
> ++ 0x00044,
> ++ &totalWrite));
> ++
> ++ profiler->gpuTotalRead64BytesPerFrame += totalRead;
> ++ profiler->gpuTotalWrite64BytesPerFrame += totalWrite;
> ++
> ++ /* PE */
> ++ gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (0) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16)))));gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os, Hardware->core, 0x00454, &colorKilled));
> ++ gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (1) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16)))));gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os, Hardware->core, 0x00454, &depthKilled));
> ++ gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (2) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16)))));gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os, Hardware->core, 0x00454, &colorDrawn));
> ++ gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (3) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16)))));gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os, Hardware->core, 0x00454, &depthDrawn));
> ++
> ++ profiler->pe_pixel_count_killed_by_color_pipe += colorKilled;
> ++ profiler->pe_pixel_count_killed_by_depth_pipe += depthKilled;
> ++ profiler->pe_pixel_count_drawn_by_color_pipe += colorDrawn;
> ++ profiler->pe_pixel_count_drawn_by_depth_pipe += depthDrawn;
> ++ }
> ++
> ++ /* Reset clock control register. */
> ++ gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os,
> ++ Hardware->core,
> ++ 0x00000,
> ++ clock));
> ++
> ++ if(Reset){
> + /* Reset counters. */
> + gcmkONERROR(
> + gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x0003C, 1));
> +@@ -4903,19 +5016,10 @@ gckHARDWARE_QueryProfileRegisters(
> + gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00438, 0));
> + gcmkONERROR(
> + gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00078, 0));
> +- }
> +- /* PE */
> +- gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (0) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) ));
> +-gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os, Hardware->core, 0x00454, &profiler->pe_pixel_count_killed_by_color_pipe));
> +- gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (1) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) ));
> +-gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os, Hardware->core, 0x00454, &profiler->pe_pixel_count_killed_by_depth_pipe));
> +- gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (2) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) ));
> +-gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os, Hardware->core, 0x00454, &profiler->pe_pixel_count_drawn_by_color_pipe));
> +- gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (3) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) ));
> +-gcmkONERROR(gckOS_ReadRegisterEx(Hardware->os, Hardware->core, 0x00454, &profiler->pe_pixel_count_drawn_by_depth_pipe));
> +- if(Reset){ gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (15) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) ));
> ++ gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (15) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) ));
> + gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16))) | (((gctUINT32) ((gctUINT32) (0) & ((gctUINT32) ((((1 ? 23:16) - (0 ? 23:16) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 23:16) - (0 ? 23:16) + 1))))))) << (0 ? 23:16)))
> +-)); }
> ++));
> ++ }
> +
> + /* SH */
> + gcmkONERROR(gckOS_WriteRegisterEx(Hardware->os, Hardware->core, 0x00470, ((((gctUINT32) (0)) & ~(((gctUINT32) (((gctUINT32) ((((1 ? 31:24) - (0 ? 31:24) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 31:24) - (0 ? 31:24) + 1))))))) << (0 ? 31:24))) | (((gctUINT32) ((gctUINT32) (7) & ((gctUINT32) ((((1 ? 31:24) - (0 ? 31:24) + 1) == 32) ? ~0 : (~(~0 << ((1 ? 31:24) - (0 ? 31:24) + 1))))))) << (0 ? 31:24))) ));
> +diff --git a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.h b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.h
> +index 517b35c..37226b7 100644
> +--- a/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.h
> ++++ b/drivers/mxc/gpu-viv/arch/XAQ2/hal/kernel/gc_hal_kernel_hardware.h
> +@@ -90,6 +90,8 @@ struct _gckHARDWARE
> + #if gcdLINK_QUEUE_SIZE
> + struct _gckLINKQUEUE linkQueue;
> + #endif
> ++
> ++ gctBOOL powerManagement;
> + };
> +
> + gceSTATUS
> +diff --git a/drivers/mxc/gpu-viv/config b/drivers/mxc/gpu-viv/config
> +index 1196efa..cdd143e 100644
> +--- a/drivers/mxc/gpu-viv/config
> ++++ b/drivers/mxc/gpu-viv/config
> +@@ -22,7 +22,6 @@
> + ARCH_TYPE ?= arm
> + SDK_DIR ?= $(AQROOT)/build/sdk
> + USE_3D_VG ?= 1
> +-USE_POWER_MANAGEMENT ?= 1
> + FORCE_ALL_VIDEO_MEMORY_CACHED ?= 0
> + NONPAGED_MEMORY_CACHEABLE ?= 0
> + NONPAGED_MEMORY_BUFFERABLE ?= 1
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.c
> +index 7964585..b7b0d28 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel.c
> +@@ -904,9 +904,6 @@ gckKERNEL_Dispatch(
> + gctSIGNAL signal;
> + #endif
> +
> +- gcsDATABASE_RECORD record;
> +- gctPOINTER data;
> +-
> + gcmkHEADER_ARG("Kernel=0x%x FromUser=%d Interface=0x%x",
> + Kernel, FromUser, Interface);
> +
> +@@ -1940,249 +1937,133 @@ gckKERNEL_Dispatch(
> + #endif
> +
> + case gcvHAL_GET_SHARED_INFO:
> +- bytes = (gctSIZE_T) Interface->u.GetSharedInfo.size;
> +-
> +- if (Interface->u.GetSharedInfo.dataId != 0)
> ++ if (Interface->u.GetSharedInfo.data == gcvNULL)
> + {
> +- gcmkONERROR(gckKERNEL_FindProcessDB(Kernel,
> +- Interface->u.GetSharedInfo.pid,
> +- 0,
> +- gcvDB_SHARED_INFO,
> +- gcmINT2PTR(Interface->u.GetSharedInfo.dataId),
> +- &record));
> +-
> +- /* find a record in db, check size */
> +- if (record.bytes != bytes)
> +- {
> +- /* Size change is not allowed */
> +- gcmkONERROR(gcvSTATUS_INVALID_DATA);
> +- }
> +-
> +- /* fetch data */
> +- gcmkONERROR(gckOS_CopyToUserData(
> +- Kernel->os,
> +- record.physical,
> +- gcmUINT64_TO_PTR(Interface->u.GetSharedInfo.data),
> +- bytes
> +- ));
> +-
> ++ gcmkONERROR(gcvSTATUS_INVALID_ARGUMENT);
> + }
> +-
> +- if ((node = gcmUINT64_TO_PTR(Interface->u.GetSharedInfo.node)) != gcvNULL)
> ++ else
> + {
> +- switch (Interface->u.GetSharedInfo.infoType)
> +- {
> +- case gcvVIDMEM_INFO_GENERIC:
> +- { /* Generic data stored */
> +- if (node->VidMem.memory->object.type == gcvOBJ_VIDMEM)
> +- {
> +- data = &node->VidMem.sharedInfo;
> +-
> +- }
> +- else
> +- {
> +- data = &node->Virtual.sharedInfo;
> +- }
> ++ gctUINT32 pid = Interface->u.GetSharedInfo.pid;
> ++ gctUINT32 dataId = Interface->u.GetSharedInfo.dataId;
> ++ gctSIZE_T bytes = Interface->u.GetSharedInfo.bytes;
> ++ gctPOINTER data = Interface->u.GetSharedInfo.data;
> ++ gcsDATABASE_RECORD record;
> +
> +- gcmkONERROR(gckOS_CopyToUserData(
> +- Kernel->os,
> +- data,
> +- gcmUINT64_TO_PTR(Interface->u.GetSharedInfo.nodeData),
> +- sizeof(gcsVIDMEM_NODE_SHARED_INFO)
> +- ));
> +- }
> +- break;
> +-
> +- case gcvVIDMEM_INFO_DIRTY_RECTANGLE:
> +- { /* Dirty rectangle stored */
> +- gcsVIDMEM_NODE_SHARED_INFO *storedSharedInfo;
> +- gcsVIDMEM_NODE_SHARED_INFO alignedSharedInfo;
> +-
> +- if (node->VidMem.memory->object.type == gcvOBJ_VIDMEM)
> +- {
> +- storedSharedInfo = &node->VidMem.sharedInfo;
> +- }
> +- else
> +- {
> +- storedSharedInfo = &node->Virtual.sharedInfo;
> +- }
> +-
> +- /* Stored shared info holds the unaligned dirty rectangle.
> +- Align it first. */
> +-
> +- /* Hardware requires 64-byte aligned address, and 16x4 pixel aligned rectsize.
> +- We simply align to 32 pixels which covers both 16- and 32-bpp formats. */
> +-
> +- /* Make sure we have a legit rectangle. */
> +- gcmkASSERT((storedSharedInfo->RectSize.width != 0) && (storedSharedInfo->RectSize.height != 0));
> +-
> +- alignedSharedInfo.SrcOrigin.x = gcmALIGN_BASE(storedSharedInfo->SrcOrigin.x, 32);
> +- alignedSharedInfo.RectSize.width = gcmALIGN((storedSharedInfo->RectSize.width + (storedSharedInfo->SrcOrigin.x - alignedSharedInfo.SrcOrigin.x)), 16);
> +-
> +- alignedSharedInfo.SrcOrigin.y = gcmALIGN_BASE(storedSharedInfo->SrcOrigin.y, 4);
> +- alignedSharedInfo.RectSize.height = gcmALIGN((storedSharedInfo->RectSize.height + (storedSharedInfo->SrcOrigin.y - alignedSharedInfo.SrcOrigin.y)), 4);
> +-
> +- gcmkONERROR(gckOS_CopyToUserData(
> +- Kernel->os,
> +- &alignedSharedInfo,
> +- gcmUINT64_TO_PTR(Interface->u.GetSharedInfo.nodeData),
> +- sizeof(gcsVIDMEM_NODE_SHARED_INFO)
> +- ));
> +-
> +- gcmkTRACE_ZONE(gcvLEVEL_VERBOSE, gcvZONE_KERNEL,
> +- "Node = %p, unaligned rectangle (l=%d, t=%d, w=%d, h=%d) aligned to (l=%d, t=%d, w=%d, h=%d)", node,
> +- storedSharedInfo->SrcOrigin.x, storedSharedInfo->SrcOrigin.y,
> +- storedSharedInfo->RectSize.width, storedSharedInfo->RectSize.height,
> +- alignedSharedInfo.SrcOrigin.x, alignedSharedInfo.SrcOrigin.y,
> +- alignedSharedInfo.RectSize.width, alignedSharedInfo.RectSize.height);
> ++ /* Find record. */
> ++ gcmkONERROR(
> ++ gckKERNEL_FindProcessDB(Kernel,
> ++ pid,
> ++ 0,
> ++ gcvDB_SHARED_INFO,
> ++ gcmINT2PTR(dataId),
> ++ &record));
> ++
> ++ /* Check memory size. */
> ++ if (bytes < record.bytes)
> ++ {
> ++ /* Insufficient memory to hold shared data. */
> ++ gcmkONERROR(gcvSTATUS_INVALID_ARGUMENT);
> ++ }
> +
> +- /* Rectangle */
> +- storedSharedInfo->SrcOrigin.x =
> +- storedSharedInfo->SrcOrigin.y =
> +- storedSharedInfo->RectSize.width =
> +- storedSharedInfo->RectSize.height = 0;
> +- }
> +- break;
> +- }
> ++ /* Copy to user. */
> ++ status = gckOS_CopyToUserData(Kernel->os,
> ++ record.physical,
> ++ data,
> ++ record.bytes);
> ++
> ++ /*
> ++ * Remove from process db.
> ++ * Every time when shared info is taken, the record is erased in
> ++ * kernel side.
> ++ */
> ++ gcmkVERIFY_OK(
> ++ gckKERNEL_RemoveProcessDB(Kernel,
> ++ pid,
> ++ gcvDB_SHARED_INFO,
> ++ gcmINT2PTR(dataId)));
> ++ /* Free existed data. */
> ++ gcmkVERIFY_OK(
> ++ gckOS_FreeMemory(Kernel->os, record.physical));
> + }
> + break;
> +
> + case gcvHAL_SET_SHARED_INFO:
> +- bytes = (gctSIZE_T) Interface->u.SetSharedInfo.size;
> +-
> +- if (Interface->u.SetSharedInfo.dataId != 0)
> + {
> +- status = gckKERNEL_FindProcessDB(Kernel, processID, 0,
> +- gcvDB_SHARED_INFO,
> +- gcmINT2PTR(Interface->u.SetSharedInfo.dataId),
> +- &record);
> +-
> +- if (status == gcvSTATUS_INVALID_DATA)
> +- {
> +- /* private data has not been created yet */
> +- /* Note: we count on DestoryProcessDB to free it */
> +- gcmkONERROR(gckOS_AllocateMemory(
> +- Kernel->os,
> +- bytes,
> +- &data
> +- ));
> +-
> +- gcmkONERROR(
> +- gckKERNEL_AddProcessDB(Kernel, processID,
> +- gcvDB_SHARED_INFO,
> +- gcmINT2PTR(Interface->u.SetSharedInfo.dataId),
> +- data,
> +- bytes
> +- ));
> +- }
> +- else
> ++ gctUINT32 dataId = Interface->u.SetSharedInfo.dataId;
> ++ gctPOINTER data = Interface->u.SetSharedInfo.data;
> ++ gctUINT32 bytes = Interface->u.SetSharedInfo.bytes;
> ++ gctPOINTER memory = gcvNULL;
> ++ gcsDATABASE_RECORD record;
> ++
> ++ if (gcmIS_SUCCESS(gckKERNEL_FindProcessDB(Kernel,
> ++ processID,
> ++ 0,
> ++ gcvDB_SHARED_INFO,
> ++ gcmINT2PTR(dataId),
> ++ &record)))
> + {
> +- /* bail on other errors */
> +- gcmkONERROR(status);
> +-
> +- /* find a record in db, check size */
> +- if (record.bytes != bytes)
> ++ /* Find a record with the same id. */
> ++ if (bytes != record.bytes)
> + {
> +- /* Size change is not allowed */
> +- gcmkONERROR(gcvSTATUS_INVALID_DATA);
> ++ /* Remove from process db. */
> ++ gcmkVERIFY_OK(
> ++ gckKERNEL_RemoveProcessDB(Kernel,
> ++ processID,
> ++ gcvDB_SHARED_INFO,
> ++ gcmINT2PTR(dataId)));
> ++
> ++ /* Free existed data. */
> ++ gcmkVERIFY_OK(
> ++ gckOS_FreeMemory(Kernel->os, record.physical));
> + }
> +-
> +- /* get storage address */
> +- data = record.physical;
> +- }
> +-
> +- gcmkONERROR(gckOS_CopyFromUserData(
> +- Kernel->os,
> +- data,
> +- gcmUINT64_TO_PTR(Interface->u.SetSharedInfo.data),
> +- bytes
> +- ));
> +- }
> +-
> +- if ((node = gcmUINT64_TO_PTR(Interface->u.SetSharedInfo.node)) != gcvNULL)
> +- {
> +- switch (Interface->u.SetSharedInfo.infoType)
> ++ else
> + {
> +- case gcvVIDMEM_INFO_GENERIC:
> +- { /* Generic data stored */
> +- if (node->VidMem.memory->object.type == gcvOBJ_VIDMEM)
> +- {
> +- data = &node->VidMem.sharedInfo;
> +- }
> +- else
> +- {
> +- data = &node->Virtual.sharedInfo;
> +- }
> +-
> +- gcmkONERROR(gckOS_CopyFromUserData(
> +- Kernel->os,
> +- data,
> +- gcmUINT64_TO_PTR(Interface->u.SetSharedInfo.nodeData),
> +- sizeof(gcsVIDMEM_NODE_SHARED_INFO)
> +- ));
> +- }
> +- break;
> ++ /* Re-use allocated memory. */
> ++ memory = record.physical;
> ++ }
> ++ }
> +
> +- case gcvVIDMEM_INFO_DIRTY_RECTANGLE:
> +- { /* Dirty rectangle stored */
> +- gcsVIDMEM_NODE_SHARED_INFO newSharedInfo;
> +- gcsVIDMEM_NODE_SHARED_INFO *currentSharedInfo;
> +- gctINT dirtyX, dirtyY, right, bottom;
> +-
> +- /* Expand the dirty rectangle stored in the node to include the rectangle passed in. */
> +- gcmkONERROR(gckOS_CopyFromUserData(
> +- Kernel->os,
> +- &newSharedInfo,
> +- gcmUINT64_TO_PTR(Interface->u.SetSharedInfo.nodeData),
> +- gcmSIZEOF(gcsVIDMEM_NODE_SHARED_INFO)
> +- ));
> +-
> +- if (node->VidMem.memory->object.type == gcvOBJ_VIDMEM)
> +- {
> +- currentSharedInfo = &node->VidMem.sharedInfo;
> +- }
> +- else
> +- {
> +- currentSharedInfo = &node->Virtual.sharedInfo;
> +- }
> ++ if ((data == gcvNULL) || (bytes == 0))
> ++ {
> ++ /* Nothing to record. */
> ++ break;
> ++ }
> +
> +- gcmkTRACE_ZONE(gcvLEVEL_VERBOSE, gcvZONE_KERNEL, "Node = %p Stored rectangle (l=%d, t=%d, w=%d, h=%d)", node,
> +- currentSharedInfo->SrcOrigin.x, currentSharedInfo->SrcOrigin.y,
> +- currentSharedInfo->RectSize.width, currentSharedInfo->RectSize.height);
> ++ if (bytes > 1024)
> ++ {
> ++ /* Limite data size. */
> ++ gcmkONERROR(gcvSTATUS_TOO_COMPLEX);
> ++ }
> +
> +- gcmkTRACE_ZONE(gcvLEVEL_VERBOSE, gcvZONE_KERNEL, "To combine with (l=%d, t=%d, w=%d, h=%d)",
> +- newSharedInfo.SrcOrigin.x, newSharedInfo.SrcOrigin.y,
> +- newSharedInfo.RectSize.width, newSharedInfo.RectSize.height);
> ++ if (memory == gcvNULL)
> ++ {
> ++ /* Allocate memory for holding shared data. */
> ++ gcmkONERROR(
> ++ gckOS_AllocateMemory(Kernel->os, bytes, &memory));
> +
> +- if ((currentSharedInfo->RectSize.width == 0) || (currentSharedInfo->RectSize.height == 0))
> +- { /* Setting it for the first time */
> +- currentSharedInfo->SrcOrigin.x = newSharedInfo.SrcOrigin.x;
> +- currentSharedInfo->SrcOrigin.y = newSharedInfo.SrcOrigin.y;
> +- currentSharedInfo->RectSize.width = newSharedInfo.RectSize.width;
> +- currentSharedInfo->RectSize.height = newSharedInfo.RectSize.height;
> +- }
> +- else
> +- {
> +- /* Expand the stored rectangle to include newly locked rectangle */
> +- dirtyX = (newSharedInfo.SrcOrigin.x < currentSharedInfo->SrcOrigin.x) ? newSharedInfo.SrcOrigin.x : currentSharedInfo->SrcOrigin.x;
> +- right = gcmMAX((currentSharedInfo->SrcOrigin.x + currentSharedInfo->RectSize.width), (newSharedInfo.SrcOrigin.x + newSharedInfo.RectSize.width));
> +- currentSharedInfo->RectSize.width = right - dirtyX;
> +- currentSharedInfo->SrcOrigin.x = dirtyX;
> +-
> +- dirtyY = (newSharedInfo.SrcOrigin.y < currentSharedInfo->SrcOrigin.y) ? newSharedInfo.SrcOrigin.y : currentSharedInfo->SrcOrigin.y;
> +- bottom = gcmMAX((currentSharedInfo->SrcOrigin.y + currentSharedInfo->RectSize.height), (newSharedInfo.SrcOrigin.y + newSharedInfo.RectSize.height));
> +- currentSharedInfo->RectSize.height = bottom - dirtyY;
> +- currentSharedInfo->SrcOrigin.y = dirtyY;
> +- }
> ++ /* Add to process db. */
> ++ status = gckKERNEL_AddProcessDB(Kernel,
> ++ processID,
> ++ gcvDB_SHARED_INFO,
> ++ gcmINT2PTR(dataId),
> ++ memory,
> ++ bytes);
> +
> +- gcmkTRACE_ZONE(gcvLEVEL_VERBOSE, gcvZONE_KERNEL, "Combined rectangle (l=%d, t=%d, w=%d, h=%d)",
> +- currentSharedInfo->SrcOrigin.x, currentSharedInfo->SrcOrigin.y,
> +- currentSharedInfo->RectSize.width, currentSharedInfo->RectSize.height);
> +- }
> ++ if (gcmIS_ERROR(status))
> ++ {
> ++ /* Failed to add process db. Free allocated memory. */
> ++ gcmkVERIFY_OK(gckOS_FreeMemory(Kernel->os, memory));
> + break;
> + }
> +- }
> ++ }
> +
> ++ /* Copy shared data to kernel memory. */
> ++ gcmkONERROR(
> ++ gckOS_CopyFromUserData(Kernel->os,
> ++ memory,
> ++ data,
> ++ bytes));
> ++ }
> + break;
> +
> + case gcvHAL_SET_FSCALE_VALUE:
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c
> +index 66ce0d1..9ee9ea1 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c
> +@@ -2047,14 +2047,14 @@ gckCOMMAND_Commit(
> + EventQueue = nextEventRecord;
> + }
> +
> +-#if gcdPOWER_MANAGEMENT
> +- if (Command->kernel->eventObj->queueHead == gcvNULL)
> ++ if (Command->kernel->eventObj->queueHead == gcvNULL
> ++ && Command->kernel->hardware->powerManagement == gcvTRUE
> ++ )
> + {
> + /* Commit done event by which work thread knows all jobs done. */
> + gcmkVERIFY_OK(
> + gckEVENT_CommitDone(Command->kernel->eventObj, gcvKERNEL_PIXEL));
> + }
> +-#endif
> +
> + /* Submit events. */
> + status = gckEVENT_Submit(Command->kernel->eventObj, gcvTRUE, gcvFALSE);
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command_vg.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command_vg.c
> +index 9685a5d..76c1c10 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command_vg.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command_vg.c
> +@@ -1234,7 +1234,6 @@ _EventHandler_BusError(
> + return gcvSTATUS_OK;
> + }
> +
> +-#if gcdPOWER_MANAGEMENT
> + /******************************************************************************\
> + ****************************** Power Stall Handler *******************************
> + \******************************************************************************/
> +@@ -1250,7 +1249,6 @@ _EventHandler_PowerStall(
> + Kernel->command->powerStallSignal,
> + gcvTRUE);
> + }
> +-#endif
> +
> + /******************************************************************************\
> + ******************************** Task Routines *********************************
> +@@ -1965,15 +1963,12 @@ gcmDECLARE_INTERRUPT_HANDLER(COMMAND, 0)
> + );
> + }
> + }
> +-#if gcdPOWER_MANAGEMENT
> + else
> + {
> +-
> + status = gckVGHARDWARE_SetPowerManagementState(
> + Kernel->command->hardware, gcvPOWER_IDLE_BROADCAST
> + );
> + }
> +-#endif
> +
> + /* Break out of the loop. */
> + break;
> +@@ -2848,7 +2843,7 @@ gckVGCOMMAND_Construct(
> + _EventHandler_BusError
> + ));
> +
> +-#if gcdPOWER_MANAGEMENT
> ++
> + command->powerStallInt = 30;
> + /* Enable the interrupt. */
> + gcmkERR_BREAK(gckVGINTERRUPT_Enable(
> +@@ -2856,7 +2851,6 @@ gckVGCOMMAND_Construct(
> + &command->powerStallInt,
> + _EventHandler_PowerStall
> + ));
> +-#endif
> +
> + /***********************************************************************
> + ** Task management initialization.
> +@@ -3419,7 +3413,6 @@ gckVGCOMMAND_Commit(
> + gcvINFINITE
> + ));
> +
> +-#if gcdPOWER_MANAGEMENT
> + status = gckVGHARDWARE_SetPowerManagementState(
> + Command->hardware, gcvPOWER_ON_AUTO);
> +
> +@@ -3447,7 +3440,7 @@ gckVGCOMMAND_Commit(
> +
> + break;
> + }
> +-#endif
> ++
> + gcmkERR_BREAK(_FlushMMU(Command));
> +
> + do
> +@@ -3676,10 +3669,9 @@ gckVGCOMMAND_Commit(
> + }
> + while (gcvFALSE);
> +
> +-#if gcdPOWER_MANAGEMENT
> + gcmkVERIFY_OK(gckOS_ReleaseSemaphore(
> + Command->os, Command->powerSemaphore));
> +-#endif
> ++
> + /* Release the mutex. */
> + gcmkCHECK_STATUS(gckOS_ReleaseMutex(
> + Command->os,
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c
> +index bc5f083..673d4f7 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_db.c
> +@@ -1303,9 +1303,9 @@ gckKERNEL_DestroyProcessDB(
> + gcmPTR2INT(record->data), status);
> + break;
> +
> +- case gcvDB_SHARED_INFO:
> +- status = gckOS_FreeMemory(Kernel->os, record->physical);
> +- break;
> ++ case gcvDB_SHARED_INFO:
> ++ status = gckOS_FreeMemory(Kernel->os, record->physical);
> ++ break;
> +
> + default:
> + gcmkTRACE_ZONE(gcvLEVEL_ERROR, gcvZONE_DATABASE,
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c
> +index 43c9297..c7f67c7 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_mmu.c
> +@@ -97,6 +97,14 @@ static gcsMirrorPageTable_PTR mirrorPageTable = gcvNULL;
> + static gctPOINTER mirrorPageTableMutex = gcvNULL;
> + #endif
> +
> ++typedef struct _gcsDynamicSpaceNode * gcsDynamicSpaceNode_PTR;
> ++typedef struct _gcsDynamicSpaceNode
> ++{
> ++ gctUINT32 start;
> ++ gctINT32 entries;
> ++}
> ++gcsDynamicSpaceNode;
> ++
> + static void
> + _WritePageEntry(
> + IN gctUINT32_PTR PageEntry,
> +@@ -482,30 +490,117 @@ OnError:
> + }
> +
> + static gceSTATUS
> ++_FindDynamicSpace(
> ++ IN gckMMU Mmu,
> ++ OUT gcsDynamicSpaceNode_PTR *Array,
> ++ OUT gctINT * Size
> ++ )
> ++{
> ++ gceSTATUS status = gcvSTATUS_OK;
> ++ gctPOINTER pointer = gcvNULL;
> ++ gcsDynamicSpaceNode_PTR array = gcvNULL;
> ++ gctINT size = 0;
> ++ gctINT i = 0, nodeStart = -1, nodeEntries = 0;
> ++
> ++ /* Allocate memory for the array. */
> ++ gcmkONERROR(gckOS_Allocate(Mmu->os,
> ++ gcmSIZEOF(*array) * (gcdMMU_MTLB_ENTRY_NUM / 2),
> ++ &pointer));
> ++
> ++ array = (gcsDynamicSpaceNode_PTR)pointer;
> ++
> ++ /* Loop all the entries. */
> ++ while (i < gcdMMU_MTLB_ENTRY_NUM)
> ++ {
> ++ if (!Mmu->mtlbLogical[i])
> ++ {
> ++ if (nodeStart < 0)
> ++ {
> ++ /* This is the first entry of the dynamic space. */
> ++ nodeStart = i;
> ++ nodeEntries = 1;
> ++ }
> ++ else
> ++ {
> ++ /* Other entries of the dynamic space. */
> ++ nodeEntries++;
> ++ }
> ++ }
> ++ else if (nodeStart >= 0)
> ++ {
> ++ /* Save the previous node. */
> ++ array[size].start = nodeStart;
> ++ array[size].entries = nodeEntries;
> ++ size++;
> ++
> ++ /* Reset the start. */
> ++ nodeStart = -1;
> ++ nodeEntries = 0;
> ++ }
> ++
> ++ i++;
> ++ }
> ++
> ++ /* Save the previous node. */
> ++ if (nodeStart >= 0)
> ++ {
> ++ array[size].start = nodeStart;
> ++ array[size].entries = nodeEntries;
> ++ size++;
> ++ }
> ++
> ++#if gcdMMU_TABLE_DUMP
> ++ for (i = 0; i < size; i++)
> ++ {
> ++ gckOS_Print("%s(%d): [%d]: start=%d, entries=%d.\n",
> ++ __FUNCTION__, __LINE__,
> ++ i,
> ++ array[i].start,
> ++ array[i].entries);
> ++ }
> ++#endif
> ++
> ++ *Array = array;
> ++ *Size = size;
> ++
> ++ return gcvSTATUS_OK;
> ++
> ++OnError:
> ++ if (pointer != gcvNULL)
> ++ {
> ++ gckOS_Free(Mmu->os, pointer);
> ++ }
> ++
> ++ return status;
> ++}
> ++
> ++static gceSTATUS
> + _SetupDynamicSpace(
> + IN gckMMU Mmu
> + )
> + {
> + gceSTATUS status;
> +- gctINT i;
> ++ gcsDynamicSpaceNode_PTR nodeArray = gcvNULL;
> ++ gctINT i, nodeArraySize = 0;
> + gctUINT32 physical;
> +- gctINT numEntries;
> ++ gctINT numEntries = 0;
> + gctUINT32_PTR pageTable;
> + gctBOOL acquired = gcvFALSE;
> +
> +- /* find the start of dynamic address space. */
> +- for (i = 0; i < gcdMMU_MTLB_ENTRY_NUM; i++)
> ++ /* Find all the dynamic address space. */
> ++ gcmkONERROR(_FindDynamicSpace(Mmu, &nodeArray, &nodeArraySize));
> ++
> ++ /* TODO: We only use the largest one for now. */
> ++ for (i = 0; i < nodeArraySize; i++)
> + {
> +- if (!Mmu->mtlbLogical[i])
> ++ if (nodeArray[i].entries > numEntries)
> + {
> +- break;
> ++ Mmu->dynamicMappingStart = nodeArray[i].start;
> ++ numEntries = nodeArray[i].entries;
> + }
> + }
> +
> +- Mmu->dynamicMappingStart = i;
> +-
> +- /* Number of entries in Master TLB for dynamic mapping. */
> +- numEntries = gcdMMU_MTLB_ENTRY_NUM - i;
> ++ gckOS_Free(Mmu->os, (gctPOINTER)nodeArray);
> +
> + Mmu->pageTableSize = numEntries * 4096;
> +
> +@@ -545,7 +640,9 @@ _SetupDynamicSpace(
> + acquired = gcvTRUE;
> +
> + /* Map to Master TLB. */
> +- for (; i < gcdMMU_MTLB_ENTRY_NUM; i++)
> ++ for (i = (gctINT)Mmu->dynamicMappingStart;
> ++ i < (gctINT)Mmu->dynamicMappingStart + numEntries;
> ++ i++)
> + {
> + _WritePageEntry(Mmu->mtlbLogical + i,
> + physical
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_video_memory.c b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_video_memory.c
> +index 8a442a2..8b8bbdc 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_video_memory.c
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_video_memory.c
> +@@ -2144,6 +2144,9 @@ gckVIDMEM_Unlock(
> +
> + if (!Node->Virtual.contiguous
> + && (Node->Virtual.lockeds[Kernel->core] == 1)
> ++#if gcdENABLE_VG
> ++ && (Kernel->vg == gcvNULL)
> ++#endif
> + )
> + {
> + if (Type == gcvSURF_BITMAP)
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal.h
> +index 7077412..4406d7e 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal.h
> +@@ -2072,6 +2072,12 @@ gckHARDWARE_QueryPowerManagementState(
> + OUT gceCHIPPOWERSTATE* State
> + );
> +
> ++gceSTATUS
> ++gckHARDWARE_SetPowerManagement(
> ++ IN gckHARDWARE Hardware,
> ++ IN gctBOOL PowerManagement
> ++ );
> ++
> + #if gcdENABLE_FSCALE_VAL_ADJUST
> + gceSTATUS
> + gckHARDWARE_SetFscaleValue(
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_base.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_base.h
> +index ac86399..44689b0 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_base.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_base.h
> +@@ -74,7 +74,6 @@ typedef struct _gcsSYNC_CONTEXT * gcsSYNC_CONTEXT_PTR;
> + /******************************************************************************\
> + ******************************* Process local storage *************************
> + \******************************************************************************/
> +-
> + typedef struct _gcsPLS * gcsPLS_PTR;
> + typedef struct _gcsPLS
> + {
> +@@ -107,6 +106,7 @@ typedef struct _gcsPLS
> +
> + /* Reference count for destructor. */
> + gcsATOM_PTR reference;
> ++ gctBOOL bKFS;
> + #if gcdUSE_NPOT_PATCH
> + gctBOOL bNeedSupportNP2Texture;
> + #endif
> +@@ -123,7 +123,7 @@ extern gcsPLS gcPLS;
> + typedef struct _gcsTLS * gcsTLS_PTR;
> +
> + typedef void (* gctTLS_DESTRUCTOR) (
> +- gcsTLS_PTR TLS
> ++ gcsTLS_PTR
> + );
> +
> + typedef struct _gcsTLS
> +@@ -658,8 +658,6 @@ gcoHAL_QueryChipFeature(
> + IN gceFEATURE Feature);
> +
> + #endif
> +-
> +-
> + /******************************************************************************\
> + ********************************** gcoOS Object *********************************
> + \******************************************************************************/
> +@@ -1775,20 +1773,6 @@ gcoSURF_QueryVidMemNode(
> + OUT gctUINT_PTR Bytes
> + );
> +
> +-/* Set usage attribute of a surface. */
> +-gceSTATUS
> +-gcoSURF_SetUsage(
> +- IN gcoSURF Surface,
> +- IN gceSURF_USAGE Usage
> +- );
> +-
> +-/* Return usage attribute of a surface. */
> +-gceSTATUS
> +-gcoSURF_QueryUsage(
> +- IN gcoSURF Surface,
> +- OUT gceSURF_USAGE *Usage
> +- );
> +-
> + /* Set the color type of the surface. */
> + gceSTATUS
> + gcoSURF_SetColorType(
> +@@ -1975,6 +1959,14 @@ gcoSURF_SetWindow(
> + IN gctUINT Height
> + );
> +
> ++/* Set width/height alignment of the surface directly and calculate stride/size. This is only for dri backend now. Please be careful before use. */
> ++gceSTATUS
> ++gcoSURF_SetAlignment(
> ++ IN gcoSURF Surface,
> ++ IN gctUINT Width,
> ++ IN gctUINT Height
> ++ );
> ++
> + /* Increase reference count of the surface. */
> + gceSTATUS
> + gcoSURF_ReferenceSurface(
> +@@ -2009,6 +2001,12 @@ gcoSURF_SetOffset(
> + );
> +
> + gceSTATUS
> ++gcoSURF_GetOffset(
> ++ IN gcoSURF Surface,
> ++ OUT gctUINT *Offset
> ++ );
> ++
> ++gceSTATUS
> + gcoSURF_NODE_Cache(
> + IN gcsSURF_NODE_PTR Node,
> + IN gctPOINTER Logical,
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_compiler.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_compiler.h
> +index 4a0870f..8693c37 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_compiler.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_compiler.h
> +@@ -36,12 +36,16 @@ extern "C" {
> + #endif
> +
> + #ifndef GC_ENABLE_LOADTIME_OPT
> +-#define GC_ENABLE_LOADTIME_OPT 1
> ++#define GC_ENABLE_LOADTIME_OPT 1
> + #endif
> +
> + #define TEMP_OPT_CONSTANT_TEXLD_COORD 1
> +
> +-#define TEMP_SHADER_PATCH 1
> ++#define TEMP_SHADER_PATCH 1
> ++
> ++#define ADD_PRE_ROTATION_TO_VS 0
> ++
> ++#define TEMP_INLINE_ALL_EXPANSION 1
> + /******************************* IR VERSION ******************/
> + #define gcdSL_IR_VERSION gcmCC('\0','\0','\0','\1')
> +
> +@@ -683,6 +687,13 @@ typedef enum _gceSHADER_FLAGS
> + gcvSHADER_USE_ALPHA_KILL = 0x100,
> + #endif
> +
> ++#if ADD_PRE_ROTATION_TO_VS
> ++ gcvSHADER_VS_PRE_ROTATION = 0x200,
> ++#endif
> ++
> ++#if TEMP_INLINE_ALL_EXPANSION
> ++ gcvSHADER_INLINE_ALL_EXPANSION = 0x200,
> ++#endif
> + }
> + gceSHADER_FLAGS;
> +
> +@@ -771,10 +782,15 @@ typedef enum _gceSHADER_OPTIMIZATION
> + /* optimize varying packing */
> + gcvOPTIMIZATION_VARYINGPACKING = 1 << 22,
> +
> ++#if TEMP_INLINE_ALL_EXPANSION
> ++ gcvOPTIMIZATION_INLINE_ALL_EXPANSION = 1 << 23,
> ++#endif
> ++
> + /* Full optimization. */
> + /* Note that gcvOPTIMIZATION_LOAD_SW_WORKAROUND is off. */
> + gcvOPTIMIZATION_FULL = 0x7FFFFFFF &
> + ~gcvOPTIMIZATION_LOAD_SW_WORKAROUND &
> ++ ~gcvOPTIMIZATION_INLINE_ALL_EXPANSION &
> + ~gcvOPTIMIZATION_POWER_OPTIMIZATION,
> +
> + /* Optimization Unit Test flag. */
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_driver.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_driver.h
> +index 028bbd1..b056c52 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_driver.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_driver.h
> +@@ -210,6 +210,9 @@ typedef struct _gcsHAL_QUERY_CHIP_IDENTITY
> + /* Supported minor feature 3 fields. */
> + gctUINT32 chipMinorFeatures3;
> +
> ++ /* Supported minor feature 4 fields. */
> ++ gctUINT32 chipMinorFeatures4;
> ++
> + /* Number of streams supported. */
> + gctUINT32 streamCount;
> +
> +@@ -929,30 +932,30 @@ typedef struct _gcsHAL_INTERFACE
> +
> + struct _gcsHAL_GET_SHARED_INFO
> + {
> ++ /* Process id. */
> + IN gctUINT32 pid;
> ++
> ++ /* Data id. */
> + IN gctUINT32 dataId;
> +- /* gcuVIDMEM_NODE_PTR */
> +- IN gctUINT64 node;
> +- /* gctUINT8_PTR */
> +- OUT gctUINT64 data;
> +- /* fix size. gctUINT8_PTR*/
> +- OUT gctUINT64 nodeData;
> +- gctUINT64 size;
> +- IN gceVIDMEM_NODE_SHARED_INFO_TYPE infoType;
> ++
> ++ /* Data size. */
> ++ IN gctSIZE_T bytes;
> ++
> ++ /* Pointer to save the shared data. */
> ++ OUT gctPOINTER data;
> + }
> + GetSharedInfo;
> +
> + struct _gcsHAL_SET_SHARED_INFO
> + {
> ++ /* Data id. */
> + IN gctUINT32 dataId;
> +- /* gcuVIDMEM_NODE_PTR */
> +- IN gctUINT64 node;
> +- /* gctUINT8_PTR */
> +- IN gctUINT64 data;
> +- /* gctUINT8_PTR */
> +- IN gctUINT64 nodeData;
> +- IN gctUINT64 size;
> +- IN gceVIDMEM_NODE_SHARED_INFO_TYPE infoType;
> ++
> ++ /* Data to be shared. */
> ++ IN gctPOINTER data;
> ++
> ++ /* Data size. */
> ++ IN gctSIZE_T bytes;
> + }
> + SetSharedInfo;
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h
> +index 249b61b..8481375 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_engine.h
> +@@ -323,50 +323,6 @@ gcoSURF_Resolve(
> + IN gcoSURF DestSurface
> + );
> +
> +-/* Export the render target. */
> +-gceSTATUS
> +-gcoSURF_ExportRenderTarget(
> +- IN gcoSURF SrcSurface
> +-);
> +-
> +-/* Import the render target. */
> +-gceSTATUS
> +-gcoSURF_ImportRenderTarget(
> +- IN gctUINT32 Pid,
> +- IN gcoSURF SrcSurface
> +-);
> +-
> +-/* Save the Resolve info to kernel. */
> +-gceSTATUS
> +-gcoSURF_PrepareRemoteResolveRect(
> +- IN gcoSURF SrcSurface,
> +- IN gcsPOINT_PTR SrcOrigin,
> +- IN gcsPOINT_PTR DestOrigin,
> +- IN gcsPOINT_PTR RectSize
> +- );
> +-
> +-/* Resolve using the rectangle info previously saved in the vid mem node. */
> +-gceSTATUS
> +-gcoSURF_ResolveFromStoredRect(
> +- IN gcoSURF SrcSurface,
> +- IN gcoSURF DestSurface
> +- );
> +-
> +-/* Using the info that Process Pid saved to do resolve. */
> +-gceSTATUS
> +-gcoSURF_RemoteResolveRect(
> +- IN gcoSURF SrcSurface,
> +- IN gcoSURF DestSurface,
> +- IN gctBOOL *resolveDiscarded
> +- );
> +-
> +-/* Return the "resolve submitted indicator" signal. */
> +-gceSTATUS
> +-gcoSURF_GetRTSignal(
> +- IN gcoSURF RTSurface,
> +- OUT gctSIGNAL * resolveSubmittedSignal
> +- );
> +-
> + /* Resolve rectangular area of a surface. */
> + gceSTATUS
> + gcoSURF_ResolveRect(
> +@@ -1684,6 +1640,12 @@ gcoTEXTURE_IsRenderable(
> + );
> +
> + gceSTATUS
> ++gcoTEXTURE_IsRenderableEx(
> ++ IN gcoTEXTURE Texture,
> ++ IN gctUINT Level
> ++ );
> ++
> ++gceSTATUS
> + gcoTEXTURE_IsComplete(
> + IN gcoTEXTURE Texture,
> + IN gctINT MaxLevel
> +@@ -2028,21 +1990,15 @@ gceSTATUS
> + gcoHAL_GetSharedInfo(
> + IN gctUINT32 Pid,
> + IN gctUINT32 DataId,
> +- OUT gctUINT8_PTR Data,
> + IN gctSIZE_T Bytes,
> +- IN gctUINT64 Node,
> +- OUT gctUINT8_PTR NodeData,
> +- IN gceVIDMEM_NODE_SHARED_INFO_TYPE SharedInfoType
> ++ OUT gctPOINTER Data
> + );
> +
> + gceSTATUS
> + gcoHAL_SetSharedInfo(
> + IN gctUINT32 DataId,
> +- IN gctUINT8_PTR Data,
> +- IN gctSIZE_T Bytes,
> +- IN gctUINT64 Node,
> +- IN gctUINT8_PTR NodeData,
> +- IN gceVIDMEM_NODE_SHARED_INFO_TYPE SharedInfoType
> ++ IN gctPOINTER Data,
> ++ IN gctSIZE_T Bytes
> + );
> +
> + #ifdef __cplusplus
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_enum.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_enum.h
> +index cf6b425..a1d9ae5 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_enum.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_enum.h
> +@@ -181,13 +181,6 @@ typedef enum _gceCACHEOPERATION
> + }
> + gceCACHEOPERATION;
> +
> +-typedef enum _gceVIDMEM_NODE_SHARED_INFO_TYPE
> +-{
> +- gcvVIDMEM_INFO_GENERIC,
> +- gcvVIDMEM_INFO_DIRTY_RECTANGLE
> +-}
> +-gceVIDMEM_NODE_SHARED_INFO_TYPE;
> +-
> + /* Surface types. */
> + typedef enum _gceSURF_TYPE
> + {
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h
> +index afe83d0..9e2a8db 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_options.h
> +@@ -391,15 +391,6 @@
> + #endif
> +
> + /*
> +- gcdPOWER_MANAGEMENT
> +-
> +- This define enables the power management code.
> +-*/
> +-#ifndef gcdPOWER_MANAGEMENT
> +-# define gcdPOWER_MANAGEMENT 1
> +-#endif
> +-
> +-/*
> + gcdPOWER_SUSNPEND_WHEN_IDLE
> +
> + Set to 1 to make GPU enter gcvPOWER_SUSPEND when idle detected,
> +@@ -428,7 +419,7 @@
> + If the value is 0, no timeout will be checked for.
> + */
> + #ifndef gcdGPU_TIMEOUT
> +-# if gcdFPGA_BUILD
> ++#if gcdFPGA_BUILD
> + # define gcdGPU_TIMEOUT 0
> + # else
> + # define gcdGPU_TIMEOUT 20000
> +@@ -726,31 +717,13 @@
> +
> + Support swap with a specific rectangle.
> +
> +- Set the rectangle with eglSetSwapRectangleANDROID api.
> ++ Set the rectangle with eglSetSwapRectangleVIV api.
> + */
> + #ifndef gcdSUPPORT_SWAP_RECTANGLE
> + # define gcdSUPPORT_SWAP_RECTANGLE 0
> + #endif
> +
> + /*
> +- gcdDEFER_RESOLVES
> +-
> +- Support deferred resolves for 3D apps.
> +-*/
> +-#ifndef gcdDEFER_RESOLVES
> +-# define gcdDEFER_RESOLVES 0
> +-#endif
> +-
> +-/*
> +- gcdCOPYBLT_OPTIMIZATION
> +-
> +- Combine dirty areas resulting from Android's copyBlt.
> +-*/
> +-#ifndef gcdCOPYBLT_OPTIMIZATION
> +-# define gcdCOPYBLT_OPTIMIZATION 0
> +-#endif
> +-
> +-/*
> + gcdGPU_LINEAR_BUFFER_ENABLED
> +
> + Use linear buffer for GPU apps so HWC can do 2D composition.
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h
> +index 808fde0..03cb4d6 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_version.h
> +@@ -28,7 +28,7 @@
> +
> + #define gcvVERSION_PATCH 9
> +
> +-#define gcvVERSION_BUILD 4651
> ++#define gcvVERSION_BUILD 6622
> +
> + #define gcvVERSION_DATE __DATE__
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_vg.h b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_vg.h
> +index 5ff0281..2a910e8 100644
> +--- a/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_vg.h
> ++++ b/drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_vg.h
> +@@ -552,6 +552,12 @@ gckVGHARDWARE_QueryPowerManagementState(
> + );
> +
> + gceSTATUS
> ++gckVGHARDWARE_SetPowerManagement(
> ++ IN gckVGHARDWARE Hardware,
> ++ IN gctBOOL PowerManagement
> ++ );
> ++
> ++gceSTATUS
> + gckVGHARDWARE_SetPowerOffTimeout(
> + IN gckVGHARDWARE Hardware,
> + IN gctUINT32 Timeout
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_device.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_device.c
> +index 7168f0e..168987a 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_device.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_device.c
> +@@ -304,6 +304,7 @@ gckGALDEVICE_Construct(
> + IN gctINT Signal,
> + IN gctUINT LogFileSize,
> + IN struct device *pdev,
> ++ IN gctINT PowerManagement,
> + OUT gckGALDEVICE *Device
> + )
> + {
> +@@ -538,6 +539,9 @@ gckGALDEVICE_Construct(
> + device->kernels[gcvCORE_MAJOR]->hardware, FastClear, Compression
> + ));
> +
> ++ gcmkONERROR(gckHARDWARE_SetPowerManagement(
> ++ device->kernels[gcvCORE_MAJOR]->hardware, PowerManagement
> ++ ));
> +
> + #if COMMAND_PROCESSOR_VERSION == 1
> + /* Start the command queue. */
> +@@ -593,6 +597,10 @@ gckGALDEVICE_Construct(
> + device
> + ));
> +
> ++ gcmkONERROR(gckHARDWARE_SetPowerManagement(
> ++ device->kernels[gcvCORE_2D]->hardware, PowerManagement
> ++ ));
> ++
> + #if COMMAND_PROCESSOR_VERSION == 1
> + /* Start the command queue. */
> + gcmkONERROR(gckCOMMAND_Start(device->kernels[gcvCORE_2D]->command));
> +@@ -624,6 +632,11 @@ gckGALDEVICE_Construct(
> + device->coreMapping[gcvHARDWARE_VG] = gcvCORE_VG;
> + }
> +
> ++
> ++ gcmkONERROR(gckVGHARDWARE_SetPowerManagement(
> ++ device->kernels[gcvCORE_VG]->vg->hardware,
> ++ PowerManagement
> ++ ));
> + #endif
> + }
> + else
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_device.h b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_device.h
> +index 460f022..d488fc8 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_device.h
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_device.h
> +@@ -169,6 +169,7 @@ gceSTATUS gckGALDEVICE_Construct(
> + IN gctINT Signal,
> + IN gctUINT LogFileSize,
> + IN struct device *pdev,
> ++ IN gctINT PowerManagement,
> + OUT gckGALDEVICE *Device
> + );
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +index 64cace1..183000d 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +@@ -131,6 +131,9 @@ module_param(fastClear, int, 0644);
> + static int compression = -1;
> + module_param(compression, int, 0644);
> +
> ++static int powerManagement = 1;
> ++module_param(powerManagement, int, 0644);
> ++
> + static int signal = 48;
> + module_param(signal, int, 0644);
> +
> +@@ -781,6 +784,9 @@ static int drv_init(struct device *pdev)
> + }
> + #endif
> +
> ++ printk(KERN_INFO "Galcore version %d.%d.%d.%d\n",
> ++ gcvVERSION_MAJOR, gcvVERSION_MINOR, gcvVERSION_PATCH, gcvVERSION_BUILD);
> ++
> + if (showArgs)
> + {
> + printk("galcore options:\n");
> +@@ -810,7 +816,8 @@ static int drv_init(struct device *pdev)
> + printk(" signal = %d\n", signal);
> + printk(" baseAddress = 0x%08lX\n", baseAddress);
> + printk(" physSize = 0x%08lX\n", physSize);
> +- printk(" logFileSize = %d KB \n", logFileSize);
> ++ printk(" logFileSize = %d KB \n", logFileSize);
> ++ printk(" powerManagement = %d\n", powerManagement);
> + #if ENABLE_GPU_CLOCK_BY_DRIVER
> + printk(" coreClock = %lu\n", coreClock);
> + #endif
> +@@ -833,6 +840,7 @@ static int drv_init(struct device *pdev)
> + bankSize, fastClear, compression, baseAddress, physSize, signal,
> + logFileSize,
> + pdev,
> ++ powerManagement,
> + &device
> + ));
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> +index dfbc699..6a0295d 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.c
> +@@ -55,6 +55,7 @@ const char * _PLATFORM = "\n\0$PLATFORM$Linux$\n";
> + #endif
> +
> + #define USER_SIGNAL_TABLE_LEN_INIT 64
> ++#define gcdSUPPRESS_OOM_MESSAGE 1
> +
> + #define MEMORY_LOCK(os) \
> + gcmkVERIFY_OK(gckOS_AcquireMutex( \
> +@@ -85,6 +86,12 @@ const char * _PLATFORM = "\n\0$PLATFORM$Linux$\n";
> + #define gcmkNONPAGED_MEMROY_PROT(x) pgprot_noncached(x)
> + #endif
> +
> ++#if gcdSUPPRESS_OOM_MESSAGE
> ++#define gcdNOWARN __GFP_NOWARN
> ++#else
> ++#define gcdNOWARN 0
> ++#endif
> ++
> + #define gcdINFINITE_TIMEOUT (60 * 1000)
> + #define gcdDETECT_TIMEOUT 0
> + #define gcdDETECT_DMA_ADDRESS 1
> +@@ -261,7 +268,7 @@ _CreateMdl(
> +
> + gcmkHEADER_ARG("ProcessID=%d", ProcessID);
> +
> +- mdl = (PLINUX_MDL)kzalloc(sizeof(struct _LINUX_MDL), GFP_KERNEL | __GFP_NOWARN);
> ++ mdl = (PLINUX_MDL)kzalloc(sizeof(struct _LINUX_MDL), GFP_KERNEL | gcdNOWARN);
> + if (mdl == gcvNULL)
> + {
> + gcmkFOOTER_NO();
> +@@ -322,7 +329,7 @@ _CreateMdlMap(
> +
> + gcmkHEADER_ARG("Mdl=0x%X ProcessID=%d", Mdl, ProcessID);
> +
> +- mdlMap = (PLINUX_MDL_MAP)kmalloc(sizeof(struct _LINUX_MDL_MAP), GFP_KERNEL | __GFP_NOWARN);
> ++ mdlMap = (PLINUX_MDL_MAP)kmalloc(sizeof(struct _LINUX_MDL_MAP), GFP_KERNEL | gcdNOWARN);
> + if (mdlMap == gcvNULL)
> + {
> + gcmkFOOTER_NO();
> +@@ -481,7 +488,7 @@ _NonContiguousAlloc(
> +
> + size = NumPages * sizeof(struct page *);
> +
> +- pages = kmalloc(size, GFP_KERNEL | __GFP_NOWARN);
> ++ pages = kmalloc(size, GFP_KERNEL | gcdNOWARN);
> +
> + if (!pages)
> + {
> +@@ -496,7 +503,7 @@ _NonContiguousAlloc(
> +
> + for (i = 0; i < NumPages; i++)
> + {
> +- p = alloc_page(GFP_KERNEL | __GFP_HIGHMEM | __GFP_NOWARN);
> ++ p = alloc_page(GFP_KERNEL | __GFP_HIGHMEM | gcdNOWARN);
> +
> + if (!p)
> + {
> +@@ -762,7 +769,7 @@ _AllocateIntegerId(
> + int result;
> +
> + again:
> +- if (idr_pre_get(&Database->idr, GFP_KERNEL | __GFP_NOWARN) == 0)
> ++ if (idr_pre_get(&Database->idr, GFP_KERNEL | gcdNOWARN) == 0)
> + {
> + return gcvSTATUS_OUT_OF_MEMORY;
> + }
> +@@ -954,7 +961,7 @@ gckOS_Construct(
> + gcmkVERIFY_ARGUMENT(Os != gcvNULL);
> +
> + /* Allocate the gckOS object. */
> +- os = (gckOS) kmalloc(gcmSIZEOF(struct _gckOS), GFP_KERNEL | __GFP_NOWARN);
> ++ os = (gckOS) kmalloc(gcmSIZEOF(struct _gckOS), GFP_KERNEL | gcdNOWARN);
> +
> + if (os == gcvNULL)
> + {
> +@@ -1171,7 +1178,7 @@ _CreateKernelVirtualMapping(
> +
> + if (Mdl->contiguous)
> + {
> +- pages = kmalloc(sizeof(struct page *) * numPages, GFP_KERNEL | __GFP_NOWARN);
> ++ pages = kmalloc(sizeof(struct page *) * numPages, GFP_KERNEL | gcdNOWARN);
> +
> + if (!pages)
> + {
> +@@ -1385,7 +1392,7 @@ gckOS_AllocateMemory(
> + }
> + else
> + {
> +- memory = (gctPOINTER) kmalloc(Bytes, GFP_KERNEL | __GFP_NOWARN);
> ++ memory = (gctPOINTER) kmalloc(Bytes, GFP_KERNEL | gcdNOWARN);
> + }
> +
> + if (memory == gcvNULL)
> +@@ -1904,7 +1911,7 @@ gckOS_AllocateNonPagedMemory(
> + addr = dma_alloc_coherent(gcvNULL,
> + mdl->numPages * PAGE_SIZE,
> + &mdl->dmaHandle,
> +- GFP_KERNEL | __GFP_NOWARN);
> ++ GFP_KERNEL | gcdNOWARN);
> + }
> + #else
> + size = mdl->numPages * PAGE_SIZE;
> +@@ -1915,7 +1922,7 @@ gckOS_AllocateNonPagedMemory(
> + if (page == gcvNULL)
> + #endif
> + {
> +- page = alloc_pages(GFP_KERNEL | __GFP_NOWARN, order);
> ++ page = alloc_pages(GFP_KERNEL | gcdNOWARN, order);
> + }
> +
> + if (page == gcvNULL)
> +@@ -3848,6 +3855,9 @@ gckOS_AllocatePagedMemoryEx(
> + gctSIZE_T bytes;
> + gctBOOL locked = gcvFALSE;
> + gceSTATUS status;
> ++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 27)
> ++ gctPOINTER addr = gcvNULL;
> ++#endif
> +
> + gcmkHEADER_ARG("Os=0x%X Contiguous=%d Bytes=%lu", Os, Contiguous, Bytes);
> +
> +@@ -3873,13 +3883,27 @@ gckOS_AllocatePagedMemoryEx(
> + {
> + /* Get contiguous pages, and suppress warning (stack dump) from kernel when
> + we run out of memory. */
> +- mdl->u.contiguousPages =
> +- alloc_pages(GFP_KERNEL | __GFP_NOWARN | __GFP_NORETRY, GetOrder(numPages));
> ++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 27)
> ++ addr =
> ++ alloc_pages_exact(numPages * PAGE_SIZE, GFP_KERNEL | gcdNOWARN | __GFP_NORETRY);
> +
> ++ mdl->u.contiguousPages = addr
> ++ ? virt_to_page(addr)
> ++ : gcvNULL;
> ++
> ++ mdl->exact = gcvTRUE;
> ++#else
> ++ mdl->u.contiguousPages =
> ++ alloc_pages(GFP_KERNEL | gcdNOWARN | __GFP_NORETRY, GetOrder(numPages));
> ++#endif
> + if (mdl->u.contiguousPages == gcvNULL)
> + {
> + mdl->u.contiguousPages =
> +- alloc_pages(GFP_KERNEL | __GFP_HIGHMEM | __GFP_NOWARN, GetOrder(numPages));
> ++ alloc_pages(GFP_KERNEL | __GFP_HIGHMEM | gcdNOWARN, GetOrder(numPages));
> ++
> ++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 27)
> ++ mdl->exact = gcvFALSE;
> ++#endif
> + }
> + }
> + else
> +@@ -4024,7 +4048,16 @@ gckOS_FreePagedMemory(
> +
> + if (mdl->contiguous)
> + {
> +- __free_pages(mdl->u.contiguousPages, GetOrder(mdl->numPages));
> ++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 27)
> ++ if (mdl->exact == gcvTRUE)
> ++ {
> ++ free_pages_exact(page_address(mdl->u.contiguousPages), mdl->numPages * PAGE_SIZE);
> ++ }
> ++ else
> ++#endif
> ++ {
> ++ __free_pages(mdl->u.contiguousPages, GetOrder(mdl->numPages));
> ++ }
> + }
> + else
> + {
> +@@ -4859,7 +4892,7 @@ gckOS_MapUserPointer(
> + gcmkVERIFY_ARGUMENT(Size > 0);
> + gcmkVERIFY_ARGUMENT(KernelPointer != gcvNULL);
> +
> +- buf = kmalloc(Size, GFP_KERNEL | __GFP_NOWARN);
> ++ buf = kmalloc(Size, GFP_KERNEL | gcdNOWARN);
> + if (buf == gcvNULL)
> + {
> + gcmkTRACE(
> +@@ -5274,7 +5307,7 @@ OnError:
> + MEMORY_MAP_LOCK(Os);
> +
> + /* Allocate the Info struct. */
> +- info = (gcsPageInfo_PTR)kmalloc(sizeof(gcsPageInfo), GFP_KERNEL | __GFP_NOWARN);
> ++ info = (gcsPageInfo_PTR)kmalloc(sizeof(gcsPageInfo), GFP_KERNEL | gcdNOWARN);
> +
> + if (info == gcvNULL)
> + {
> +@@ -5283,7 +5316,7 @@ OnError:
> + }
> +
> + /* Allocate the array of page addresses. */
> +- pages = (struct page **)kmalloc(pageCount * sizeof(struct page *), GFP_KERNEL | __GFP_NOWARN);
> ++ pages = (struct page **)kmalloc(pageCount * sizeof(struct page *), GFP_KERNEL | gcdNOWARN);
> +
> + if (pages == gcvNULL)
> + {
> +@@ -6502,7 +6535,7 @@ gckOS_CreateSemaphore(
> + gcmkVERIFY_ARGUMENT(Semaphore != gcvNULL);
> +
> + /* Allocate the semaphore structure. */
> +- sem = (struct semaphore *)kmalloc(gcmSIZEOF(struct semaphore), GFP_KERNEL | __GFP_NOWARN);
> ++ sem = (struct semaphore *)kmalloc(gcmSIZEOF(struct semaphore), GFP_KERNEL | gcdNOWARN);
> + if (sem == gcvNULL)
> + {
> + gcmkONERROR(gcvSTATUS_OUT_OF_MEMORY);
> +@@ -6942,6 +6975,7 @@ gckOS_SetGPUPower(
> + #else
> + imx_gpc_power_up_pu(false);
> + #endif
> ++
> + }
> + /* TODO: Put your code here. */
> + gcmkFOOTER_NO();
> +@@ -7255,7 +7289,7 @@ gckOS_CreateSignal(
> + gcmkVERIFY_ARGUMENT(Signal != gcvNULL);
> +
> + /* Create an event structure. */
> +- signal = (gcsSIGNAL_PTR) kmalloc(sizeof(gcsSIGNAL), GFP_KERNEL | __GFP_NOWARN);
> ++ signal = (gcsSIGNAL_PTR) kmalloc(sizeof(gcsSIGNAL), GFP_KERNEL | gcdNOWARN);
> +
> + if (signal == gcvNULL)
> + {
> +@@ -8000,7 +8034,7 @@ gckOS_CreateSemaphoreVG(
> + do
> + {
> + /* Allocate the semaphore structure. */
> +- newSemaphore = (struct semaphore *)kmalloc(gcmSIZEOF(struct semaphore), GFP_KERNEL | __GFP_NOWARN);
> ++ newSemaphore = (struct semaphore *)kmalloc(gcmSIZEOF(struct semaphore), GFP_KERNEL | gcdNOWARN);
> + if (newSemaphore == gcvNULL)
> + {
> + gcmkERR_BREAK(gcvSTATUS_OUT_OF_MEMORY);
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.h b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.h
> +index e970477..006632c 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.h
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_os.h
> +@@ -54,6 +54,9 @@ typedef struct _LINUX_MDL
> + gctINT numPages;
> + gctINT pagedMem;
> + gctBOOL contiguous;
> ++#if LINUX_VERSION_CODE >= KERNEL_VERSION(2, 6, 27)
> ++ gctBOOL exact;
> ++#endif
> + dma_addr_t dmaHandle;
> + PLINUX_MDL_MAP maps;
> + struct _LINUX_MDL * prev;
> +--
> +1.8.3.2
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0004-ENGR00271136-Fix-build-break-when-CONFIG_CLK_DEBUG-i.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0004-ENGR00271136-Fix-build-break-when-CONFIG_CLK_DEBUG-i.patch
> new file mode 100644
> index 0000000..7316351
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0004-ENGR00271136-Fix-build-break-when-CONFIG_CLK_DEBUG-i.patch
> @@ -0,0 +1,43 @@
> +From d8601292ae25e0af47aa4486055221ab44113f0e Mon Sep 17 00:00:00 2001
> +From: Mahesh Mahadevan <Mahesh.Mahadevan at freescale.com>
> +Date: Mon, 15 Jul 2013 15:34:54 -0500
> +Subject: [PATCH] ENGR00271136 Fix build break when CONFIG_CLK_DEBUG is
> + disabled
> +Organization: O.S. Systems Software LTDA.
> +
> +clk structure member name is defined only when CONFIG_CLK_DEBUG is enabled.
> +Hence need to encapsulate the code with this config.
> +
> +Patch received from imx community:
> +https://community.freescale.com/thread/308482
> +
> +Upstream-Status: Pending
> +
> +Signed-off-by: xiongweihuang
> +Signed-off-by: Mahesh Mahadevan <Mahesh.Mahadevan at freescale.com>
> +---
> + arch/arm/plat-mxc/clock.c | 4 ++--
> + 1 file changed, 2 insertions(+), 2 deletions(-)
> +
> +diff --git a/arch/arm/plat-mxc/clock.c b/arch/arm/plat-mxc/clock.c
> +index 93347eb..1aa2664 100755
> +--- a/arch/arm/plat-mxc/clock.c
> ++++ b/arch/arm/plat-mxc/clock.c
> +@@ -58,12 +58,12 @@ static void __clk_disable(struct clk *clk)
> + {
> + if (clk == NULL || IS_ERR(clk))
> + return;
> +-
> ++#ifdef CONFIG_CLK_DEBUG
> + if (!clk->usecount) {
> + WARN(1, "clock enable/disable mismatch! clk %s\n", clk->name);
> + return;
> + }
> +-
> ++#endif
> + if (!(--clk->usecount)) {
> + if (clk->disable)
> + clk->disable(clk);
> +--
> +1.8.4.rc3
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0005-ENGR00264275-GPU-Correct-suspend-resume-calling-afte.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0005-ENGR00264275-GPU-Correct-suspend-resume-calling-afte.patch
> new file mode 100644
> index 0000000..fa937e4
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0005-ENGR00264275-GPU-Correct-suspend-resume-calling-afte.patch
> @@ -0,0 +1,61 @@
> +From a845abe4589366d3e37b5a646be6337984074b28 Mon Sep 17 00:00:00 2001
> +From: Loren HUANG <b02279 at freescale.com>
> +Date: Wed, 22 May 2013 17:21:30 +0800
> +Subject: [PATCH 5/6] ENGR00264275 [GPU]Correct suspend/resume calling after
> + adding runtime pm.
> +
> +After enabling runtime pm the suspend/resume entry is changed.
> +
> +-Add new entry for suspend/resume in runtime pm frame work.
> +-Add static define for all runtime pm function.
> +
> +Upstream-Status: Backport [3.5.7-1.0.0]
> +
> +Signed-off-by: Loren HUANG <b02279 at freescale.com>
> +Acked-by: Lily Zhang
> +---
> + .../gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c | 16 ++++++++++++++--
> + 1 file changed, 14 insertions(+), 2 deletions(-)
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +index 183000d..3632a6c 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +@@ -1252,20 +1252,32 @@ static const struct of_device_id mxs_gpu_dt_ids[] = {
> + MODULE_DEVICE_TABLE(of, mxs_gpu_dt_ids);
> +
> + #ifdef CONFIG_PM
> +-int gpu_runtime_suspend(struct device *dev)
> ++static int gpu_runtime_suspend(struct device *dev)
> + {
> + release_bus_freq(BUS_FREQ_HIGH);
> + return 0;
> + }
> +
> +-int gpu_runtime_resume(struct device *dev)
> ++static int gpu_runtime_resume(struct device *dev)
> + {
> + request_bus_freq(BUS_FREQ_HIGH);
> + return 0;
> + }
> +
> ++static int gpu_system_suspend(struct device *dev)
> ++{
> ++ pm_message_t state={0};
> ++ return gpu_suspend(to_platform_device(dev), state);
> ++}
> ++
> ++static int gpu_system_resume(struct device *dev)
> ++{
> ++ return gpu_resume(to_platform_device(dev));
> ++}
> ++
> + static const struct dev_pm_ops gpu_pm_ops = {
> + SET_RUNTIME_PM_OPS(gpu_runtime_suspend, gpu_runtime_resume, NULL)
> ++ SET_SYSTEM_SLEEP_PM_OPS(gpu_system_suspend, gpu_system_resume)
> + };
> + #endif
> + #endif
> +--
> +1.8.3.2
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0005-ENGR00271359-Add-Multi-touch-support.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0005-ENGR00271359-Add-Multi-touch-support.patch
> new file mode 100644
> index 0000000..cb20198
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0005-ENGR00271359-Add-Multi-touch-support.patch
> @@ -0,0 +1,98 @@
> +From 538f4bb2f7a51f267395550a5be9f0ab2e426712 Mon Sep 17 00:00:00 2001
> +From: Erik Boto <erik.boto at pelagicore.com>
> +Date: Tue, 16 Jul 2013 12:06:05 -0500
> +Subject: [PATCH] ENGR00271359 Add Multi-touch support
> +Organization: O.S. Systems Software LTDA.
> +
> +The previous behavior of the driver did not work properly with Qt5
> +QtQuick multi touch-point gestures, due to how touch-points are
> +reported when removing a touch-point. My interpretation of the
> +available documentation [1] was that the driver should report all
> +touch-points between SYN_REPORTs, but it is not explicitly stated so.
> +I've found another mail-thread [2] where the creator of the protocol
> +states:
> +
> +"The protocol defines a generic way of sending a variable amount of
> +contacts. The contact count is obtained by counting the number of
> +non-empty finger packets between SYN_REPORT events."-Henrik Rydberg
> +
> +I think this verifies my assumption that all touch-points should be
> +reported between SYN_REPORTs, otherwise it can not be used to obtain
> +the count.
> +
> +[1] https://www.kernel.org/doc/Documentation/input/multi-touch-protocol.txt
> +[2] http://lists.x.org/archives/xorg-devel/2010-March/006466.html
> +
> +Upstream-Status: Pending
> +
> +Signed-off-by: Erik Boto <erik.boto at pelagicore.com>
> +Signed-off-by: Mahesh Mahadevan <Mahesh.Mahadevan at freescale.com>
> +(cherry picked from commit 7cba001c5a502680f6dbf902821726779a9c9287)
> +---
> + drivers/input/touchscreen/egalax_ts.c | 36 +++++++++++++++++------------------
> + 1 file changed, 18 insertions(+), 18 deletions(-)
> +
> +diff --git a/drivers/input/touchscreen/egalax_ts.c b/drivers/input/touchscreen/egalax_ts.c
> +index 0b6cde7..271f820 100644
> +--- a/drivers/input/touchscreen/egalax_ts.c
> ++++ b/drivers/input/touchscreen/egalax_ts.c
> +@@ -133,7 +133,6 @@ retry:
> + }
> +
> + if (down) {
> +- /* should also report old pointers */
> + events[id].valid = valid;
> + events[id].status = down;
> + events[id].x = x;
> +@@ -144,23 +143,6 @@ retry:
> + input_report_abs(input_dev, ABS_Y, y);
> + input_event(data->input_dev, EV_KEY, BTN_TOUCH, 1);
> + input_report_abs(input_dev, ABS_PRESSURE, 1);
> +-#else
> +- for (i = 0; i < MAX_SUPPORT_POINTS; i++) {
> +- if (!events[i].valid)
> +- continue;
> +- dev_dbg(&client->dev, "report id:%d valid:%d x:%d y:%d",
> +- i, valid, x, y);
> +-
> +- input_report_abs(input_dev,
> +- ABS_MT_TRACKING_ID, i);
> +- input_report_abs(input_dev,
> +- ABS_MT_TOUCH_MAJOR, 1);
> +- input_report_abs(input_dev,
> +- ABS_MT_POSITION_X, events[i].x);
> +- input_report_abs(input_dev,
> +- ABS_MT_POSITION_Y, events[i].y);
> +- input_mt_sync(input_dev);
> +- }
> + #endif
> + } else {
> + dev_dbg(&client->dev, "release id:%d\n", id);
> +@@ -176,6 +158,24 @@ retry:
> + #endif
> + }
> +
> ++#ifndef CONFIG_TOUCHSCREEN_EGALAX_SINGLE_TOUCH
> ++ /* report all pointers */
> ++ for (i = 0; i < MAX_SUPPORT_POINTS; i++) {
> ++ if (!events[i].valid)
> ++ continue;
> ++ dev_dbg(&client->dev, "report id:%d valid:%d x:%d y:%d",
> ++ i, valid, x, y);
> ++ input_report_abs(input_dev,
> ++ ABS_MT_TRACKING_ID, i);
> ++ input_report_abs(input_dev,
> ++ ABS_MT_TOUCH_MAJOR, 1);
> ++ input_report_abs(input_dev,
> ++ ABS_MT_POSITION_X, events[i].x);
> ++ input_report_abs(input_dev,
> ++ ABS_MT_POSITION_Y, events[i].y);
> ++ input_mt_sync(input_dev);
> ++ }
> ++#endif
> + input_sync(input_dev);
> + return IRQ_HANDLED;
> + }
> +--
> +1.8.4.rc3
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0006-Add-support-for-DVI-monitors.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0006-Add-support-for-DVI-monitors.patch
> new file mode 100644
> index 0000000..00a6b5c
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0006-Add-support-for-DVI-monitors.patch
> @@ -0,0 +1,227 @@
> +From 3e6441d113f72b412081a2c87f39011e4c253a35 Mon Sep 17 00:00:00 2001
> +From: Robert Winkler <robert.winkler at boundarydevices.com>
> +Date: Fri, 19 Jul 2013 19:00:41 -0700
> +Subject: [PATCH] Add support for DVI monitors
> +Organization: O.S. Systems Software LTDA.
> +
> +Upstream-Status: Pending
> +
> +Signed-off-by: Robert Winkler <robert.winkler at boundarydevices.com>
> +---
> + arch/arm/plat-mxc/include/mach/mxc_hdmi.h | 7 +++
> + drivers/video/mxc_hdmi.c | 98 +++++++++++++------------------
> + 2 files changed, 49 insertions(+), 56 deletions(-)
> +
> +diff --git a/arch/arm/plat-mxc/include/mach/mxc_hdmi.h b/arch/arm/plat-mxc/include/mach/mxc_hdmi.h
> +index 94f7638..af59c62 100644
> +--- a/arch/arm/plat-mxc/include/mach/mxc_hdmi.h
> ++++ b/arch/arm/plat-mxc/include/mach/mxc_hdmi.h
> +@@ -605,6 +605,10 @@ enum {
> + HDMI_IH_MUTE_PHY_STAT0_TX_PHY_LOCK = 0x2,
> + HDMI_IH_MUTE_PHY_STAT0_HPD = 0x1,
> +
> ++/* IH and IH_MUTE convenience macro RX_SENSE | HPD*/
> ++ HDMI_DVI_IH_STAT = 0x3D,
> ++
> ++
> + /* IH_AHBDMAAUD_STAT0 field values */
> + HDMI_IH_AHBDMAAUD_STAT0_ERROR = 0x20,
> + HDMI_IH_AHBDMAAUD_STAT0_LOST = 0x10,
> +@@ -903,6 +907,9 @@ enum {
> + HDMI_PHY_HPD = 0x02,
> + HDMI_PHY_TX_PHY_LOCK = 0x01,
> +
> ++/* HDMI STAT convenience RX_SENSE | HPD */
> ++ HDMI_DVI_STAT = 0xF2,
> ++
> + /* PHY_I2CM_SLAVE_ADDR field values */
> + HDMI_PHY_I2CM_SLAVE_ADDR_PHY_GEN2 = 0x69,
> + HDMI_PHY_I2CM_SLAVE_ADDR_HEAC_PHY = 0x49,
> +diff --git a/drivers/video/mxc_hdmi.c b/drivers/video/mxc_hdmi.c
> +index c5069aa..544f352 100644
> +--- a/drivers/video/mxc_hdmi.c
> ++++ b/drivers/video/mxc_hdmi.c
> +@@ -180,7 +180,6 @@ struct mxc_hdmi {
> + bool dft_mode_set;
> + char *dft_mode_str;
> + int default_bpp;
> +- u8 latest_intr_stat;
> + bool irq_enabled;
> + spinlock_t irq_lock;
> + bool phy_enabled;
> +@@ -1996,58 +1995,48 @@ static void hotplug_worker(struct work_struct *work)
> + struct delayed_work *delay_work = to_delayed_work(work);
> + struct mxc_hdmi *hdmi =
> + container_of(delay_work, struct mxc_hdmi, hotplug_work);
> +- u32 phy_int_stat, phy_int_pol, phy_int_mask;
> +- u8 val;
> ++ u32 hdmi_phy_stat0, hdmi_phy_pol0, hdmi_phy_mask0;
> + unsigned long flags;
> + char event_string[32];
> + char *envp[] = { event_string, NULL };
> +
> +- phy_int_stat = hdmi->latest_intr_stat;
> +- phy_int_pol = hdmi_readb(HDMI_PHY_POL0);
> +
> +- dev_dbg(&hdmi->pdev->dev, "phy_int_stat=0x%x, phy_int_pol=0x%x\n",
> +- phy_int_stat, phy_int_pol);
> ++ hdmi_phy_stat0 = hdmi_readb(HDMI_PHY_STAT0);
> ++ hdmi_phy_pol0 = hdmi_readb(HDMI_PHY_POL0);
> ++
> ++ dev_dbg(&hdmi->pdev->dev, "hdmi_phy_stat0=0x%x, hdmi_phy_pol0=0x%x\n",
> ++ hdmi_phy_stat0, hdmi_phy_pol0);
> ++
> ++ /* Make HPD intr active low to capture unplug event or
> ++ * active high to capture plugin event */
> ++ hdmi_writeb((HDMI_DVI_STAT & ~hdmi_phy_stat0), HDMI_PHY_POL0);
> +
> + /* check cable status */
> +- if (phy_int_stat & HDMI_IH_PHY_STAT0_HPD) {
> +- /* cable connection changes */
> +- if (phy_int_pol & HDMI_PHY_HPD) {
> +- /* Plugin event */
> +- dev_dbg(&hdmi->pdev->dev, "EVENT=plugin\n");
> +- mxc_hdmi_cable_connected(hdmi);
> +-
> +- /* Make HPD intr active low to capture unplug event */
> +- val = hdmi_readb(HDMI_PHY_POL0);
> +- val &= ~HDMI_PHY_HPD;
> +- hdmi_writeb(val, HDMI_PHY_POL0);
> +-
> +- sprintf(event_string, "EVENT=plugin");
> +- kobject_uevent_env(&hdmi->pdev->dev.kobj, KOBJ_CHANGE, envp);
> ++ if (hdmi_phy_stat0 & HDMI_DVI_STAT) {
> ++ /* Plugin event */
> ++ dev_dbg(&hdmi->pdev->dev, "EVENT=plugin\n");
> ++ mxc_hdmi_cable_connected(hdmi);
> ++
> ++ sprintf(event_string, "EVENT=plugin");
> ++ kobject_uevent_env(&hdmi->pdev->dev.kobj, KOBJ_CHANGE, envp);
> + #ifdef CONFIG_MXC_HDMI_CEC
> +- mxc_hdmi_cec_handle(0x80);
> ++ mxc_hdmi_cec_handle(0x80);
> + #endif
> +- hdmi_set_cable_state(1);
> +-
> +- } else if (!(phy_int_pol & HDMI_PHY_HPD)) {
> +- /* Plugout event */
> +- dev_dbg(&hdmi->pdev->dev, "EVENT=plugout\n");
> +- hdmi_set_cable_state(0);
> +- mxc_hdmi_abort_stream();
> +- mxc_hdmi_cable_disconnected(hdmi);
> ++ hdmi_set_cable_state(1);
> +
> +- /* Make HPD intr active high to capture plugin event */
> +- val = hdmi_readb(HDMI_PHY_POL0);
> +- val |= HDMI_PHY_HPD;
> +- hdmi_writeb(val, HDMI_PHY_POL0);
> ++ } else {
> ++ /* Plugout event */
> ++ dev_dbg(&hdmi->pdev->dev, "EVENT=plugout\n");
> ++ hdmi_set_cable_state(0);
> ++ mxc_hdmi_abort_stream();
> ++ mxc_hdmi_cable_disconnected(hdmi);
> +
> +- sprintf(event_string, "EVENT=plugout");
> +- kobject_uevent_env(&hdmi->pdev->dev.kobj, KOBJ_CHANGE, envp);
> ++ sprintf(event_string, "EVENT=plugout");
> ++ kobject_uevent_env(&hdmi->pdev->dev.kobj, KOBJ_CHANGE, envp);
> + #ifdef CONFIG_MXC_HDMI_CEC
> +- mxc_hdmi_cec_handle(0x100);
> ++ mxc_hdmi_cec_handle(0x100);
> + #endif
> +
> +- } else
> +- dev_dbg(&hdmi->pdev->dev, "EVENT=none?\n");
> + }
> +
> + /* Lock here to ensure full powerdown sequence
> +@@ -2055,12 +2044,12 @@ static void hotplug_worker(struct work_struct *work)
> + spin_lock_irqsave(&hdmi->irq_lock, flags);
> +
> + /* Re-enable HPD interrupts */
> +- phy_int_mask = hdmi_readb(HDMI_PHY_MASK0);
> +- phy_int_mask &= ~HDMI_PHY_HPD;
> +- hdmi_writeb(phy_int_mask, HDMI_PHY_MASK0);
> ++ hdmi_phy_mask0 = hdmi_readb(HDMI_PHY_MASK0);
> ++ hdmi_phy_mask0 &= ~HDMI_DVI_STAT;
> ++ hdmi_writeb(hdmi_phy_mask0, HDMI_PHY_MASK0);
> +
> + /* Unmute interrupts */
> +- hdmi_writeb(~HDMI_IH_MUTE_PHY_STAT0_HPD, HDMI_IH_MUTE_PHY_STAT0);
> ++ hdmi_writeb(~HDMI_DVI_IH_STAT, HDMI_IH_MUTE_PHY_STAT0);
> +
> + if (hdmi_readb(HDMI_IH_FC_STAT2) & HDMI_IH_FC_STAT2_OVERFLOW_MASK)
> + mxc_hdmi_clear_overflow();
> +@@ -2086,7 +2075,7 @@ static void hdcp_hdp_worker(struct work_struct *work)
> + static irqreturn_t mxc_hdmi_hotplug(int irq, void *data)
> + {
> + struct mxc_hdmi *hdmi = data;
> +- u8 val, intr_stat;
> ++ u8 val;
> + unsigned long flags;
> +
> + spin_lock_irqsave(&hdmi->irq_lock, flags);
> +@@ -2108,25 +2097,22 @@ static irqreturn_t mxc_hdmi_hotplug(int irq, void *data)
> + * HDMI registers.
> + */
> + /* Capture status - used in hotplug_worker ISR */
> +- intr_stat = hdmi_readb(HDMI_IH_PHY_STAT0);
> +-
> +- if (intr_stat & HDMI_IH_PHY_STAT0_HPD) {
> ++ if (hdmi_readb(HDMI_IH_PHY_STAT0) & HDMI_DVI_IH_STAT) {
> +
> + dev_dbg(&hdmi->pdev->dev, "Hotplug interrupt received\n");
> +- hdmi->latest_intr_stat = intr_stat;
> +
> + /* Mute interrupts until handled */
> +
> + val = hdmi_readb(HDMI_IH_MUTE_PHY_STAT0);
> +- val |= HDMI_IH_MUTE_PHY_STAT0_HPD;
> ++ val |= HDMI_DVI_IH_STAT;
> + hdmi_writeb(val, HDMI_IH_MUTE_PHY_STAT0);
> +
> + val = hdmi_readb(HDMI_PHY_MASK0);
> +- val |= HDMI_PHY_HPD;
> ++ val |= HDMI_DVI_STAT;
> + hdmi_writeb(val, HDMI_PHY_MASK0);
> +
> + /* Clear Hotplug interrupts */
> +- hdmi_writeb(HDMI_IH_PHY_STAT0_HPD, HDMI_IH_PHY_STAT0);
> ++ hdmi_writeb(HDMI_DVI_IH_STAT, HDMI_IH_PHY_STAT0);
> +
> + schedule_delayed_work(&(hdmi->hotplug_work), msecs_to_jiffies(20));
> + }
> +@@ -2282,13 +2268,13 @@ static void mxc_hdmi_fb_registered(struct mxc_hdmi *hdmi)
> + HDMI_PHY_I2CM_CTLINT_ADDR);
> +
> + /* enable cable hot plug irq */
> +- hdmi_writeb((u8)~HDMI_PHY_HPD, HDMI_PHY_MASK0);
> ++ hdmi_writeb((u8)~HDMI_DVI_STAT, HDMI_PHY_MASK0);
> +
> + /* Clear Hotplug interrupts */
> +- hdmi_writeb(HDMI_IH_PHY_STAT0_HPD, HDMI_IH_PHY_STAT0);
> ++ hdmi_writeb(HDMI_DVI_IH_STAT, HDMI_IH_PHY_STAT0);
> +
> + /* Unmute interrupts */
> +- hdmi_writeb(~HDMI_IH_MUTE_PHY_STAT0_HPD, HDMI_IH_MUTE_PHY_STAT0);
> ++ hdmi_writeb(~HDMI_DVI_IH_STAT, HDMI_IH_MUTE_PHY_STAT0);
> +
> + hdmi->fb_reg = true;
> +
> +@@ -2522,10 +2508,10 @@ static int mxc_hdmi_disp_init(struct mxc_dispdrv_handle *disp,
> +
> + /* Configure registers related to HDMI interrupt
> + * generation before registering IRQ. */
> +- hdmi_writeb(HDMI_PHY_HPD, HDMI_PHY_POL0);
> ++ hdmi_writeb(HDMI_DVI_STAT, HDMI_PHY_POL0);
> +
> + /* Clear Hotplug interrupts */
> +- hdmi_writeb(HDMI_IH_PHY_STAT0_HPD, HDMI_IH_PHY_STAT0);
> ++ hdmi_writeb(HDMI_DVI_IH_STAT, HDMI_IH_PHY_STAT0);
> +
> + hdmi->nb.notifier_call = mxc_hdmi_fb_event;
> + ret = fb_register_client(&hdmi->nb);
> +--
> +1.8.4.rc3
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0006-ENGR00265130-gpu-Correct-section-mismatch-in-gpu-ker.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0006-ENGR00265130-gpu-Correct-section-mismatch-in-gpu-ker.patch
> new file mode 100644
> index 0000000..43407a1
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0006-ENGR00265130-gpu-Correct-section-mismatch-in-gpu-ker.patch
> @@ -0,0 +1,60 @@
> +From 376d63e9b981118f83646a836ce6626e541de1a3 Mon Sep 17 00:00:00 2001
> +From: Loren HUANG <b02279 at freescale.com>
> +Date: Fri, 31 May 2013 18:29:58 +0800
> +Subject: [PATCH 6/6] ENGR00265130 gpu:Correct section mismatch in gpu kernel
> + driver
> +
> +-Remove the __devinit for suspend/resume function.
> +-Replace __devinit to __devexit for remove function.
> +
> +Upstream-Status: Backport [3.5.7-1.0.0]
> +
> +Signed-off-by: Loren HUANG <b02279 at freescale.com>
> +Acked-by: Lily Zhang
> +---
> + drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c | 8 ++++----
> + 1 file changed, 4 insertions(+), 4 deletions(-)
> +
> +diff --git a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +index 3632a6c..9d9dc57 100644
> +--- a/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> ++++ b/drivers/mxc/gpu-viv/hal/os/linux/kernel/gc_hal_kernel_driver.c
> +@@ -1111,7 +1111,7 @@ static int __devinit gpu_probe(struct platform_device *pdev)
> + return ret;
> + }
> +
> +-static int __devinit gpu_remove(struct platform_device *pdev)
> ++static int __devexit gpu_remove(struct platform_device *pdev)
> + {
> + gcmkHEADER();
> + #if gcdENABLE_FSCALE_VAL_ADJUST
> +@@ -1123,7 +1123,7 @@ static int __devinit gpu_remove(struct platform_device *pdev)
> + return 0;
> + }
> +
> +-static int __devinit gpu_suspend(struct platform_device *dev, pm_message_t state)
> ++static int gpu_suspend(struct platform_device *dev, pm_message_t state)
> + {
> + gceSTATUS status;
> + gckGALDEVICE device;
> +@@ -1173,7 +1173,7 @@ static int __devinit gpu_suspend(struct platform_device *dev, pm_message_t state
> + return 0;
> + }
> +
> +-static int __devinit gpu_resume(struct platform_device *dev)
> ++static int gpu_resume(struct platform_device *dev)
> + {
> + gceSTATUS status;
> + gckGALDEVICE device;
> +@@ -1284,7 +1284,7 @@ static const struct dev_pm_ops gpu_pm_ops = {
> +
> + static struct platform_driver gpu_driver = {
> + .probe = gpu_probe,
> +- .remove = gpu_remove,
> ++ .remove = __devexit_p(gpu_remove),
> +
> + .suspend = gpu_suspend,
> + .resume = gpu_resume,
> +--
> +1.8.3.2
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/0007-ARM-mach-mx6-board-mx6q_sabresd-Register-SDHC3-first.patch b/recipes-kernel/linux/linux-udoo-3.0.35/0007-ARM-mach-mx6-board-mx6q_sabresd-Register-SDHC3-first.patch
> new file mode 100644
> index 0000000..d02aa40
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/0007-ARM-mach-mx6-board-mx6q_sabresd-Register-SDHC3-first.patch
> @@ -0,0 +1,38 @@
> +From cd31abbe08372fa870fac78ae845edd4859f8835 Mon Sep 17 00:00:00 2001
> +From: Fabio Estevam <fabio.estevam at freescale.com>
> +Date: Sat, 28 Sep 2013 18:46:18 -0300
> +Subject: [PATCH] ARM: mach-mx6: board-mx6q_sabresd: Register SDHC3 first
> +Organization: O.S. Systems Software LTDA.
> +
> +On sabresd boards we boot from SDHC3, so let's register it as mmc0.
> +
> +Currently eMMC is mmc0 and mmc1 can be SDHC3 or SDHC2 (if present).
> +
> +Registering SDHC3 is safer as we can always find the rootfs.
> +
> +Signed-off-by: Fabio Estevam <fabio.estevam at freescale.com>
> +---
> + arch/arm/mach-mx6/board-mx6q_sabresd.c | 5 +----
> + 1 file changed, 1 insertion(+), 4 deletions(-)
> +
> +diff --git a/arch/arm/mach-mx6/board-mx6q_sabresd.c b/arch/arm/mach-mx6/board-mx6q_sabresd.c
> +index 3f9a845..4e6b323 100644
> +--- a/arch/arm/mach-mx6/board-mx6q_sabresd.c
> ++++ b/arch/arm/mach-mx6/board-mx6q_sabresd.c
> +@@ -1847,12 +1847,9 @@ static void __init mx6_sabresd_board_init(void)
> +
> + imx6q_add_pm_imx(0, &mx6q_sabresd_pm_data);
> +
> +- /* Move sd4 to first because sd4 connect to emmc.
> +- Mfgtools want emmc is mmcblk0 and other sd card is mmcblk1.
> +- */
> ++ imx6q_add_sdhci_usdhc_imx(2, &mx6q_sabresd_sd3_data);
> + imx6q_add_sdhci_usdhc_imx(3, &mx6q_sabresd_sd4_data);
> + imx6q_add_sdhci_usdhc_imx(1, &mx6q_sabresd_sd2_data);
> +- imx6q_add_sdhci_usdhc_imx(2, &mx6q_sabresd_sd3_data);
> + imx_add_viv_gpu(&imx6_gpu_data, &imx6q_gpu_pdata);
> + imx6q_sabresd_init_usb();
> + /* SATA is not supported by MX6DL/Solo */
> +--
> +1.8.4.rc3
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/defconfig b/recipes-kernel/linux/linux-udoo-3.0.35/defconfig
> new file mode 100644
> index 0000000..6eb1443
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/defconfig
> @@ -0,0 +1,246 @@
> +CONFIG_EXPERIMENTAL=y
> +CONFIG_LOCALVERSION="-3.0.35-unico"
> +CONFIG_SYSVIPC=y
> +CONFIG_RCU_FANOUT=32
> +CONFIG_IKCONFIG=y
> +CONFIG_IKCONFIG_PROC=y
> +CONFIG_LOG_BUF_SHIFT=14
> +CONFIG_EMBEDDED=y
> +CONFIG_MODULES=y
> +CONFIG_MODULE_UNLOAD=y
> +CONFIG_MODULE_FORCE_UNLOAD=y
> +CONFIG_MODVERSIONS=y
> +# CONFIG_BLK_DEV_BSG is not set
> +CONFIG_NO_HZ=y
> +CONFIG_HIGH_RES_TIMERS=y
> +CONFIG_SMP=y
> +CONFIG_NR_CPUS=4
> +CONFIG_PREEMPT=y
> +CONFIG_COMPACTION=y
> +CONFIG_KSM=y
> +# CONFIG_SECCOMP is not set
> +CONFIG_SUSPEND_DEVICE_TIME_DEBUG=y
> +CONFIG_PM_RUNTIME=y
> +CONFIG_PM_DEBUG=y
> +CONFIG_CPU_FREQ=y
> +CONFIG_CPU_FREQ_DEFAULT_GOV_INTERACTIVE=y
> +CONFIG_CPU_FREQ_GOV_PERFORMANCE=y
> +CONFIG_CPU_FREQ_GOV_POWERSAVE=y
> +CONFIG_CPU_FREQ_GOV_USERSPACE=y
> +CONFIG_CPU_FREQ_GOV_ONDEMAND=y
> +CONFIG_CPU_FREQ_GOV_CONSERVATIVE=y
> +# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
> +CONFIG_NET=y
> +CONFIG_PACKET=y
> +CONFIG_UNIX=y
> +CONFIG_INET=y
> +CONFIG_IP_MULTICAST=y
> +CONFIG_IP_PNP=y
> +CONFIG_IP_PNP_DHCP=y
> +CONFIG_IP_PNP_BOOTP=y
> +# CONFIG_INET_LRO is not set
> +# CONFIG_IPV6 is not set
> +CONFIG_LLC2=y
> +CONFIG_CAN=y
> +CONFIG_CAN_RAW=y
> +CONFIG_CAN_BCM=y
> +CONFIG_CAN_VCAN=y
> +CONFIG_BT=y
> +CONFIG_BT_L2CAP=y
> +CONFIG_BT_SCO=y
> +CONFIG_BT_RFCOMM=y
> +CONFIG_BT_RFCOMM_TTY=y
> +CONFIG_BT_BNEP=y
> +CONFIG_BT_BNEP_MC_FILTER=y
> +CONFIG_BT_BNEP_PROTO_FILTER=y
> +CONFIG_BT_HIDP=y
> +CONFIG_BT_HCIBTUSB=y
> +CONFIG_BT_HCIUART=y
> +CONFIG_BT_HCIUART_ATH3K=y
> +CONFIG_BT_HCIVHCI=y
> +CONFIG_CFG80211=y
> +CONFIG_MAC80211=y
> +CONFIG_RFKILL=y
> +CONFIG_RFKILL_INPUT=y
> +CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug"
> +CONFIG_DEVTMPFS=y
> +CONFIG_DEVTMPFS_MOUNT=y
> +CONFIG_CONNECTOR=y
> +CONFIG_MTD=y
> +CONFIG_MTD_CMDLINE_PARTS=y
> +CONFIG_MTD_CHAR=y
> +CONFIG_MTD_BLOCK=y
> +CONFIG_MTD_CFI=y
> +CONFIG_MTD_CFI_AMDSTD=y
> +CONFIG_MTD_PHYSMAP=y
> +CONFIG_MTD_M25P80=y
> +CONFIG_MTD_NAND=y
> +CONFIG_MTD_UBI=y
> +CONFIG_BLK_DEV_LOOP=y
> +CONFIG_MISC_DEVICES=y
> +CONFIG_BLK_DEV_SD=y
> +CONFIG_SCSI_MULTI_LUN=y
> +CONFIG_ATA=y
> +# CONFIG_SATA_PMP is not set
> +CONFIG_SATA_AHCI=y
> +CONFIG_SATA_AHCI_PLATFORM=y
> +CONFIG_NETDEVICES=y
> +CONFIG_MII=y
> +CONFIG_PHYLIB=y
> +CONFIG_MICREL_PHY=y
> +CONFIG_NET_ETHERNET=y
> +# CONFIG_NETDEV_1000 is not set
> +# CONFIG_NETDEV_10000 is not set
> +CONFIG_ATH_COMMON=m
> +CONFIG_ATH6KL=m
> +CONFIG_HOSTAP=y
> +CONFIG_RT2X00=y
> +CONFIG_RT2800USB=y
> +CONFIG_RT2800USB_RT35XX=y
> +CONFIG_RT2800USB_RT53XX=y
> +CONFIG_RT2800USB_UNKNOWN=y
> +CONFIG_INPUT_POLLDEV=y
> +CONFIG_INPUT_EVDEV=y
> +CONFIG_KEYBOARD_GPIO=y
> +# CONFIG_INPUT_MOUSE is not set
> +CONFIG_INPUT_TOUCHSCREEN=y
> +CONFIG_TOUCHSCREEN_EGALAX=y
> +CONFIG_TOUCHSCREEN_ELAN=y
> +CONFIG_TOUCHSCREEN_MAX11801=y
> +CONFIG_TOUCHSCREEN_TSC2006=y
> +CONFIG_INPUT_MISC=y
> +CONFIG_INPUT_UINPUT=y
> +CONFIG_INPUT_ISL29023=y
> +CONFIG_VT_HW_CONSOLE_BINDING=y
> +CONFIG_HW_RANDOM=y
> +CONFIG_I2C=y
> +CONFIG_I2C_CHARDEV=y
> +CONFIG_SPI=y
> +CONFIG_SPI_BITBANG=y
> +CONFIG_GPIOLIB=y
> +CONFIG_GPIO_SYSFS=y
> +CONFIG_GPIO_PCA953X=y
> +CONFIG_SABRESD_MAX8903=y
> +CONFIG_SENSORS_MAX17135=y
> +CONFIG_SENSORS_MAG3110=y
> +# CONFIG_MXC_MMA8450 is not set
> +CONFIG_WATCHDOG=y
> +CONFIG_WATCHDOG_NOWAYOUT=y
> +CONFIG_MFD_WM8994=y
> +CONFIG_MFD_PFUZE=y
> +CONFIG_MFD_MAX17135=y
> +CONFIG_REGULATOR=y
> +CONFIG_REGULATOR_FIXED_VOLTAGE=y
> +CONFIG_REGULATOR_PFUZE100=y
> +CONFIG_REGULATOR_MAX17135=y
> +CONFIG_MEDIA_SUPPORT=y
> +CONFIG_VIDEO_DEV=y
> +# CONFIG_RC_CORE is not set
> +# CONFIG_MEDIA_TUNER_CUSTOMISE is not set
> +CONFIG_VIDEO_VIVI=y
> +CONFIG_USB_VIDEO_CLASS=y
> +# CONFIG_USB_GSPCA is not set
> +# CONFIG_RADIO_ADAPTERS is not set
> +CONFIG_DRM=m
> +CONFIG_DRM_VIVANTE=m
> +CONFIG_FB=y
> +CONFIG_FB_MODE_HELPERS=y
> +CONFIG_BACKLIGHT_LCD_SUPPORT=y
> +# CONFIG_LCD_CLASS_DEVICE is not set
> +CONFIG_BACKLIGHT_CLASS_DEVICE=y
> +# CONFIG_BACKLIGHT_GENERIC is not set
> +CONFIG_FRAMEBUFFER_CONSOLE=y
> +CONFIG_FONTS=y
> +CONFIG_LOGO=y
> +CONFIG_SOUND=y
> +CONFIG_SND=y
> +CONFIG_SND_USB_AUDIO=y
> +CONFIG_SND_SOC=y
> +CONFIG_HIDRAW=y
> +CONFIG_HID_A4TECH=y
> +CONFIG_HID_APPLE=y
> +CONFIG_HID_BELKIN=y
> +CONFIG_HID_CHERRY=y
> +CONFIG_HID_CHICONY=y
> +CONFIG_HID_CYPRESS=y
> +CONFIG_HID_EZKEY=y
> +CONFIG_HID_GYRATION=y
> +CONFIG_HID_LOGITECH=y
> +CONFIG_HID_MICROSOFT=y
> +CONFIG_HID_MONTEREY=y
> +CONFIG_HID_PANTHERLORD=y
> +CONFIG_HID_PETALYNX=y
> +CONFIG_HID_SAMSUNG=y
> +CONFIG_HID_SONY=y
> +CONFIG_HID_SUNPLUS=y
> +CONFIG_USB=y
> +# CONFIG_USB_DEVICE_CLASS is not set
> +CONFIG_USB_SUSPEND=y
> +CONFIG_USB_OTG=y
> +# CONFIG_USB_OTG_WHITELIST is not set
> +CONFIG_USB_EHCI_HCD=y
> +CONFIG_USB_EHCI_ROOT_HUB_TT=y
> +# CONFIG_USB_EHCI_TT_NEWSCHED is not set
> +CONFIG_USB_STORAGE=y
> +CONFIG_USB_SERIAL=y
> +CONFIG_USB_SERIAL_CONSOLE=y
> +CONFIG_USB_SERIAL_GENERIC=y
> +CONFIG_USB_SERIAL_FTDI_SIO=m
> +CONFIG_USB_SERIAL_DEBUG=y
> +CONFIG_USB_GADGET=y
> +CONFIG_USB_ETH=y
> +CONFIG_MMC=y
> +CONFIG_MMC_UNSAFE_RESUME=y
> +CONFIG_MMC_SDHCI=y
> +CONFIG_MMC_SDHCI_PLTFM=y
> +CONFIG_NEW_LEDS=y
> +CONFIG_LEDS_CLASS=y
> +CONFIG_LEDS_GPIO=y
> +CONFIG_LEDS_TRIGGERS=y
> +CONFIG_LEDS_TRIGGER_GPIO=y
> +CONFIG_LEDS_TRIGGER_DEFAULT_ON=y
> +CONFIG_RTC_CLASS=y
> +CONFIG_RTC_INTF_DEV_UIE_EMUL=y
> +CONFIG_RTC_DRV_PCF2123=y
> +# CONFIG_RTC_DRV_CMOS is not set
> +CONFIG_DMADEVICES=y
> +CONFIG_EXT2_FS=y
> +CONFIG_EXT3_FS=y
> +# CONFIG_EXT3_DEFAULTS_TO_ORDERED is not set
> +CONFIG_EXT4_FS=y
> +CONFIG_AUTOFS4_FS=y
> +CONFIG_MSDOS_FS=y
> +CONFIG_VFAT_FS=y
> +CONFIG_TMPFS=y
> +CONFIG_JFFS2_FS=y
> +CONFIG_UBIFS_FS=y
> +CONFIG_CRAMFS=y
> +CONFIG_NFS_FS=y
> +CONFIG_NFS_V3=y
> +CONFIG_ROOT_NFS=y
> +CONFIG_PARTITION_ADVANCED=y
> +CONFIG_EFI_PARTITION=y
> +CONFIG_NLS_CODEPAGE_437=y
> +CONFIG_NLS_ASCII=m
> +CONFIG_NLS_ISO8859_1=y
> +CONFIG_NLS_UTF8=m
> +CONFIG_PRINTK_TIME=y
> +CONFIG_FRAME_WARN=1024
> +CONFIG_MAGIC_SYSRQ=y
> +# CONFIG_UNUSED_SYMBOLS is not set
> +CONFIG_DEBUG_FS=y
> +CONFIG_SYSCTL_SYSCALL_CHECK=y
> +# CONFIG_CRYPTO_MANAGER_DISABLE_TESTS is not set
> +CONFIG_CRYPTO_AUTHENC=y
> +CONFIG_CRYPTO_TEST=m
> +CONFIG_CRYPTO_CCM=y
> +CONFIG_CRYPTO_GCM=y
> +CONFIG_CRYPTO_CBC=y
> +CONFIG_CRYPTO_CTS=y
> +CONFIG_CRYPTO_LRW=y
> +CONFIG_CRYPTO_PCBC=y
> +CONFIG_CRYPTO_XTS=y
> +CONFIG_CRYPTO_DES=y
> +# CONFIG_CRYPTO_ANSI_CPRNG is not set
> +CONFIG_CRYPTO_DEV_FSL_CAAM_SM=y
> +CONFIG_CRYPTO_DEV_FSL_CAAM_SM_TEST=y
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/drm-vivante-Add-00-sufix-in-returned-bus-Id.patch b/recipes-kernel/linux/linux-udoo-3.0.35/drm-vivante-Add-00-sufix-in-returned-bus-Id.patch
> new file mode 100644
> index 0000000..815d02c
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/drm-vivante-Add-00-sufix-in-returned-bus-Id.patch
> @@ -0,0 +1,31 @@
> +From b37a944f55a5010bd08297a63db0275540922f32 Mon Sep 17 00:00:00 2001
> +From: Otavio Salvador <otavio at ossystems.com.br>
> +Date: Thu, 22 Aug 2013 16:31:29 -0300
> +Subject: [PATCH] drm/vivante: Add ":00" sufix in returned bus Id
> +
> +This makes the 3.0.35 compatible with a Xorg driver build for 3.5.7 or
> +newer kernels.
> +
> +Upstream-Status: Inapropriate [embedded specific]
> +
> +Signed-off-by: Otavio Salvador <otavio at ossystems.com.br>
> +---
> + drivers/gpu/drm/vivante/vivante_drv.c | 2 +-
> + 1 file changed, 1 insertion(+), 1 deletion(-)
> +
> +diff --git a/drivers/gpu/drm/vivante/vivante_drv.c b/drivers/gpu/drm/vivante/vivante_drv.c
> +index 4224608..cea360d 100644
> +--- a/drivers/gpu/drm/vivante/vivante_drv.c
> ++++ b/drivers/gpu/drm/vivante/vivante_drv.c
> +@@ -55,7 +55,7 @@
> +
> + #include "drm_pciids.h"
> +
> +-static char platformdevicename[] = "Vivante GCCore";
> ++static char platformdevicename[] = "Vivante GCCore:00";
> + static struct platform_device *pplatformdev;
> +
> + static struct drm_driver driver = {
> +--
> +1.8.4.rc1
> +
> diff --git a/recipes-kernel/linux/linux-udoo-3.0.35/epdc-Rename-mxcfb_epdc_kernel.h-to-mxc_epdc.h.patch b/recipes-kernel/linux/linux-udoo-3.0.35/epdc-Rename-mxcfb_epdc_kernel.h-to-mxc_epdc.h.patch
> new file mode 100644
> index 0000000..0a20b3f
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo-3.0.35/epdc-Rename-mxcfb_epdc_kernel.h-to-mxc_epdc.h.patch
> @@ -0,0 +1,143 @@
> +From 149545df26169d257b144ff78934ce9cb5b6818b Mon Sep 17 00:00:00 2001
> +From: Otavio Salvador <otavio at ossystems.com.br>
> +Date: Sat, 19 Oct 2013 10:55:11 -0300
> +Subject: [PATCH] epdc: Rename mxcfb_epdc_kernel.h to mxc_epdc.h
> +Organization: O.S. Systems Software LTDA.
> +
> +This allow for forward compatibility with imx-test >= 3.10.9-1.0.0.
> +
> +Signed-off-by: Otavio Salvador <otavio at ossystems.com.br>
> +---
> + drivers/video/mxc/mxc_epdc_fb.c | 2 +-
> + include/linux/mxcfb_epdc.h | 49 +++++++++++++++++++++++++++++++++++++++
> + include/linux/mxcfb_epdc_kernel.h | 49 ---------------------------------------
> + 3 files changed, 50 insertions(+), 50 deletions(-)
> + create mode 100644 include/linux/mxcfb_epdc.h
> + delete mode 100644 include/linux/mxcfb_epdc_kernel.h
> +
> +diff --git a/drivers/video/mxc/mxc_epdc_fb.c b/drivers/video/mxc/mxc_epdc_fb.c
> +index 4103498..b3ef8ea 100644
> +--- a/drivers/video/mxc/mxc_epdc_fb.c
> ++++ b/drivers/video/mxc/mxc_epdc_fb.c
> +@@ -43,7 +43,7 @@
> + #include <linux/dmaengine.h>
> + #include <linux/pxp_dma.h>
> + #include <linux/mxcfb.h>
> +-#include <linux/mxcfb_epdc_kernel.h>
> ++#include <linux/mxcfb_epdc.h>
> + #include <linux/gpio.h>
> + #include <linux/regulator/driver.h>
> + #include <linux/fsl_devices.h>
> +diff --git a/include/linux/mxcfb_epdc.h b/include/linux/mxcfb_epdc.h
> +new file mode 100644
> +index 0000000..06fea6f
> +--- /dev/null
> ++++ b/include/linux/mxcfb_epdc.h
> +@@ -0,0 +1,49 @@
> ++/*
> ++ * Copyright (C) 2010-2012 Freescale Semiconductor, Inc. All Rights Reserved.
> ++ *
> ++ * This program is free software; you can redistribute it and/or modify
> ++ * it under the terms of the GNU General Public License as published by
> ++ * the Free Software Foundation; either version 2 of the License, or
> ++ * (at your option) any later version.
> ++ *
> ++ * This program is distributed in the hope that it will be useful,
> ++ * but WITHOUT ANY WARRANTY; without even the implied warranty of
> ++ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> ++ * GNU General Public License for more details.
> ++ *
> ++ * You should have received a copy of the GNU General Public License
> ++ * along with this program; if not, write to the Free Software
> ++ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
> ++ *
> ++ */
> ++#ifndef _MXCFB_EPDC_KERNEL
> ++#define _MXCFB_EPDC_KERNEL
> ++
> ++void mxc_epdc_fb_set_waveform_modes(struct mxcfb_waveform_modes *modes,
> ++ struct fb_info *info);
> ++int mxc_epdc_fb_set_temperature(int temperature, struct fb_info *info);
> ++int mxc_epdc_fb_set_auto_update(u32 auto_mode, struct fb_info *info);
> ++int mxc_epdc_fb_send_update(struct mxcfb_update_data *upd_data,
> ++ struct fb_info *info);
> ++int mxc_epdc_fb_wait_update_complete(
> ++ struct mxcfb_update_marker_data *marker_data,
> ++ struct fb_info *info);
> ++int mxc_epdc_fb_set_pwrdown_delay(u32 pwrdown_delay,
> ++ struct fb_info *info);
> ++int mxc_epdc_get_pwrdown_delay(struct fb_info *info);
> ++int mxc_epdc_fb_set_upd_scheme(u32 upd_scheme, struct fb_info *info);
> ++
> ++void mxc_spdc_fb_set_waveform_modes(struct mxcfb_waveform_modes *modes,
> ++ struct fb_info *info);
> ++int mxc_spdc_fb_set_temperature(int temperature, struct fb_info *info);
> ++int mxc_spdc_fb_set_auto_update(u32 auto_mode, struct fb_info *info);
> ++int mxc_spdc_fb_send_update(struct mxcfb_update_data *upd_data,
> ++ struct fb_info *info);
> ++int mxc_spdc_fb_wait_update_complete(
> ++ struct mxcfb_update_marker_data *marker_data,
> ++ struct fb_info *info);
> ++int mxc_spdc_fb_set_pwrdown_delay(u32 pwrdown_delay,
> ++ struct fb_info *info);
> ++int mxc_spdc_get_pwrdown_delay(struct fb_info *info);
> ++int mxc_spdc_fb_set_upd_scheme(u32 upd_scheme, struct fb_info *info);
> ++#endif
> +diff --git a/include/linux/mxcfb_epdc_kernel.h b/include/linux/mxcfb_epdc_kernel.h
> +deleted file mode 100644
> +index 06fea6f..0000000
> +--- a/include/linux/mxcfb_epdc_kernel.h
> ++++ /dev/null
> +@@ -1,49 +0,0 @@
> +-/*
> +- * Copyright (C) 2010-2012 Freescale Semiconductor, Inc. All Rights Reserved.
> +- *
> +- * This program is free software; you can redistribute it and/or modify
> +- * it under the terms of the GNU General Public License as published by
> +- * the Free Software Foundation; either version 2 of the License, or
> +- * (at your option) any later version.
> +- *
> +- * This program is distributed in the hope that it will be useful,
> +- * but WITHOUT ANY WARRANTY; without even the implied warranty of
> +- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> +- * GNU General Public License for more details.
> +- *
> +- * You should have received a copy of the GNU General Public License
> +- * along with this program; if not, write to the Free Software
> +- * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
> +- *
> +- */
> +-#ifndef _MXCFB_EPDC_KERNEL
> +-#define _MXCFB_EPDC_KERNEL
> +-
> +-void mxc_epdc_fb_set_waveform_modes(struct mxcfb_waveform_modes *modes,
> +- struct fb_info *info);
> +-int mxc_epdc_fb_set_temperature(int temperature, struct fb_info *info);
> +-int mxc_epdc_fb_set_auto_update(u32 auto_mode, struct fb_info *info);
> +-int mxc_epdc_fb_send_update(struct mxcfb_update_data *upd_data,
> +- struct fb_info *info);
> +-int mxc_epdc_fb_wait_update_complete(
> +- struct mxcfb_update_marker_data *marker_data,
> +- struct fb_info *info);
> +-int mxc_epdc_fb_set_pwrdown_delay(u32 pwrdown_delay,
> +- struct fb_info *info);
> +-int mxc_epdc_get_pwrdown_delay(struct fb_info *info);
> +-int mxc_epdc_fb_set_upd_scheme(u32 upd_scheme, struct fb_info *info);
> +-
> +-void mxc_spdc_fb_set_waveform_modes(struct mxcfb_waveform_modes *modes,
> +- struct fb_info *info);
> +-int mxc_spdc_fb_set_temperature(int temperature, struct fb_info *info);
> +-int mxc_spdc_fb_set_auto_update(u32 auto_mode, struct fb_info *info);
> +-int mxc_spdc_fb_send_update(struct mxcfb_update_data *upd_data,
> +- struct fb_info *info);
> +-int mxc_spdc_fb_wait_update_complete(
> +- struct mxcfb_update_marker_data *marker_data,
> +- struct fb_info *info);
> +-int mxc_spdc_fb_set_pwrdown_delay(u32 pwrdown_delay,
> +- struct fb_info *info);
> +-int mxc_spdc_get_pwrdown_delay(struct fb_info *info);
> +-int mxc_spdc_fb_set_upd_scheme(u32 upd_scheme, struct fb_info *info);
> +-#endif
> +--
> +1.8.4.rc3
> +
> diff --git a/recipes-kernel/linux/linux-udoo.inc b/recipes-kernel/linux/linux-udoo.inc
> new file mode 100644
> index 0000000..9fb946a
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo.inc
> @@ -0,0 +1,17 @@
> +# Adapted from linux-imx.inc, copyright (C) 2012, 2013 O.S. Systems Software LTDA
> +# Released under the MIT license (see COPYING.MIT for the terms)
> +
> +DESCRIPTION = "Linux kernel UNICO 2013 for UDOO"
> +LICENSE = "GPLv2"
> +LIC_FILES_CHKSUM = "file://COPYING;md5=d7810fab7487fb0aad327b76f1be7cd7"
> +
> +require recipes-kernel/linux/linux-imx.inc
> +
> +# Put a local version until we have a true SRCREV to point to
> +LOCALVERSION ?= "+yocto"
> +SCMVERSION ?= "y"
> +
> +# SRC_URI for wandboard kernel
> +SRC_URI = "git://github.com/UDOOboard/Kernel_Unico.git \
> + file://defconfig \
> +"
> diff --git a/recipes-kernel/linux/linux-udoo_3.0.35.bb b/recipes-kernel/linux/linux-udoo_3.0.35.bb
> new file mode 100644
> index 0000000..6a412eb
> --- /dev/null
> +++ b/recipes-kernel/linux/linux-udoo_3.0.35.bb
> @@ -0,0 +1,26 @@
> +# Adapted from linux-imx.inc, copyright (C) 2012, 2013 O.S. Systems Software LTDA
I'm not sure it was adapted from linux-imx.inc, or from
linux-imx_${PV}.bb instead.
Copyright here should be 2013 only, it's a new file being created.
> +# Released under the MIT license (see COPYING.MIT for the terms)
> +
> +include linux-udoo.inc
> +
> +# Kernel_Unico is a branch from 3.0.35 so we apply the 3.0.35 patches
> +SRCREV = "029e85168c08c8c9a8a6dd308b388599be608948"
> +LOCALVERSION = "-3.0.35-udoo"
> +
> +# Patches need for Yocto and not applied by Freescale when doing 4.1.0 branch
> +SRC_URI += "file://drm-vivante-Add-00-sufix-in-returned-bus-Id.patch \
> + file://epdc-Rename-mxcfb_epdc_kernel.h-to-mxc_epdc.h.patch\
> + file://0001-perf-tools-Fix-getrusage-related-build-failure-on-gl.patch \
> + file://0002-ARM-7668-1-fix-memset-related-crashes-caused-by-rece.patch \
> + file://0003-ARM-7670-1-fix-the-memset-fix.patch \
> + file://0004-ENGR00271136-Fix-build-break-when-CONFIG_CLK_DEBUG-i.patch \
> + file://0005-ENGR00271359-Add-Multi-touch-support.patch \
> + file://0007-ARM-mach-mx6-board-mx6q_sabresd-Register-SDHC3-first.patch\
> + file://0001-ENGR00255688-4.6.9p11.1-gpu-GPU-Kernel-driver-integr.patch \
> + file://0002-ENGR00265465-gpu-Add-global-value-for-minimum-3D-clo.patch \
> + file://0003-ENGR00261814-4-gpu-use-new-PU-power-on-off-interface.patch \
> + file://0004-ENGR00264288-1-GPU-Integrate-4.6.9p12-release-kernel.patch \
> + file://0005-ENGR00264275-GPU-Correct-suspend-resume-calling-afte.patch \
> + file://0006-ENGR00265130-gpu-Correct-section-mismatch-in-gpu-ker.patch"
> +
> +COMPATIBLE_MACHINE = "(mx6)"
>
--
Daiane
More information about the meta-freescale
mailing list