New commits imported: 43cecda fmd: remove cc-aging support 867551d fmd: relocate aging_support field to maintain compatibility 08d3161 Merge pull request #75 in SDK/linux-devel from ~R52568/linux-devel:hotfix/QLINUX-3938-fmd-ioc_compat_fm_pcd_kg_scheme_spc_t to master 1cf7b57 fmd: PCD: fix offline port blocked by fragmented IPV6 295abbb fmd: move compat_copy_fm_pcd_kg_scheme_spc out of DPAA 11 def 976b0aa fmd: PCD: fix offline port blocked by fragmented IPV6 b242e9c fmd: Add support for IPF after HMan fd8e7c7 Merge pull request #69 in SDK/linux-devel from ~R27762/linux-devel:add_hm_node_stats to master 1c291fd Merge pull request #72 in SDK/linux-devel from ~B56421/linux-devel_sdk_ip-usb:bugfix/QSDK-2454 to master 50eea51 Merge pull request #48 in SDK/linux-devel from ~R52568/linux-devel:bugfix/QLINUX-3822-monitoring-any-match-inside-hash to master dd05648 arm: dts: ls1021a: Add dma-coherent property to usb3 node e8c288e drivers: dwc3: Set dma ops for xhci dev 56afee9 drivers: dwc3 : Add support for USB snooping 3f83306 drivers: dwc3: host: Change burst beat and outstanding pipelined transfers requests. 1666d68 fmd: add ioctl call to retrieve scheme counter - spc a9d0ea5 fmd: Render header manip node statistics available in user space e176e3f fmd: fix compat layer for aging_support 3a8d082 Merge pull request #65 in SDK/linux-devel from ~R01356/linux-devel:bugfix/QUSDPA-704-segfault-in-dma_mem-allocator.c to master a57ffba Merge pull request #64 in SDK/linux-devel from ~B06830/linux-devel:bugfix/QSDK-2404-cryptodev-linux-sha1-offloading to master 0586472 Merge pull request #63 in SDK/linux-devel from ~B29983/linux-devel-tyt:QSDK-2438-there-is-some-improper-information to master 61f0e11 Merge pull request #61 in SDK/linux-devel from ~R54964/linux-devel:bug-fix to master cfb981f dev-exporting dev functions to support ceetm e98f36e fsl_usdpaa: Remove TLB entries when unmapping memory 8700bab crypto: caam - Change kmalloc to kzalloc to avoid residual data 4fd8cac crypto: caam - fix memory corruption in ahash_final_ctx 16d4087 crypto: caam - Fix incorrect size when DMA unmapping buffer 270c770 crypto: caam - improve initalization for context state saves 67ef9d13 crypto: caam - enable LARGE_BURST for enhancing DMA transactions size 2c27666 crypto: caam - make write transactions bufferable on PPC platforms 418316c cpufreq: qoriq: fix output information error 300c3ff fsl-qbman: Handle memory leaks 56e16e2 fsl-pme: Check for null pointers 3ac31aa fsl-dce: Handle memory allocation failures e11b65b fsl-dce: Correct error paths and add new labels ea1259d fsl_qbman: null check for lni in qman_ceemt_sp_release e82a761 fmd: fix locks in ModifyKey function dedb4b3 fmd: support aging in CC 0506053 pci/layerscape: Do not check link status e03bbdd asf_tcp_nf: Patch to correct typecast in tcp timestamp copy. 91aade9 Revert "gianfar: Add dma transfer endian property support" 03dd666 kconfig-asf : asf_qos disabled with asf by default b8d41cb dev-asf-qos - correcting the qos hooks for dpaa linux qos 18b0779 dpaa_eth: fix bpool seeding for advanced drivers 7faf498 net, gianfar, rt: Update TX_TIMEOUT to 5HZ 502650e net, rt: limit net_rx napi calls in threaded ISR context e37ce25 powerpc/t1040/dts: Add interrupts for PHYs connected to L2 Switch ports 6da728b fmd: support aging in CC 5cd3a8f fmd: fix policer spelling issues b5d9899 fmd: Remove FM_PORT_PcdPrsModifyStartOffset routine - not supported 4c642e7 fsl_qbman: enable sending cscn to dcp for old qman revision b8887bd t4240dts: correct the ceetm node f33567a dpaa_eth: Add a configurable dpa_tx call e48b858 mmc: sdhci-of-esdhc: fix compile error on ARM platform fb9a06b QorIQ/TMU: add TMU node to device tree for QorIQ T1023/T1024 4fbd4cf QorIQ/TMU: add thermal management support based on TMU a9463a3 QorIQ/TMU: add TMU node to device tree for QorIQ T1040/T1042 6b43e2b thermal: fix cpu_cooling max_level behavior f0a70b9 mmc: sdhci-of-esdhc: add SDR50 mode support for SD/MMC Legacy Adapter Card 9ac6075 mmc: esdhc: add eMMC45 Adapter Card HS200 mode support e89634e QorIQ/TMU: add TMU node to device tree for LS1021A fd1b282 Merge branch 'feature/LS1-for-SDK' of ssh://sw-stash.freescale.net/dnnpi/ls1-linux-jason into LS1-SDKV04-SDK1.9 8d3a3ca t104xd4rdb: add DS26522 nodes to device tree aaa395f Merge branch 'qoriq-sdk' into LS1-SDK-Rev2.0 ec78b41 arm: ls1021a: mask interrupts before entering deep sleep 13d0daf arm: dts: ls1021a: Add transfer mode property in DSPI2 node 6f50fc1 arm: dts: ls1021atwr: Enable DSPI2 dts node 7a6dba4 arm: ls1021a: change the order of setting PMC interrupt registers 4e61417 crypto: caam - fix SEC ERA reading bcd5862 crypto: caam - fix snooping for write transactions eec616d1 arm/dts: Add node for ina220 on LS1021ATWR-Rev 3915d65 arm/defconfig: Add Atheros AT803X PHYs support e364aa2 QE/HDLC: modify QE-HDLC for both ARM and POWERPC bb21962 Test patch: Enable CAN SoC level loopback a4c59bc arm/dts: Add nodes for flexcan devices present on LS1021A-Rev2 SoC 6e9c5a3 can: flexcan: Add support for non RX-FIFO mode c466de8 can: flexcan: Remodel FlexCAN register r/w APIs for BE instances 306f67c can: flexcan: Add ls1021a flexcan device entry 7e71163d doc/bindings: Add 'endianess' optional-property for FlexCAN controller 1795472 Revert 'can: flexcan: Add ls1021a flexcan device entry' 863d717 Revert 'net: can: Remodel FlexCAN register read/write APIs for BE instances' 41cfacd Revert 'can: flexcan: Add support for non RX-FIFO mode and conditional ERRATA ERR005829 handling' 8d04086 mmc: sdhci: more efficient interrupt enable register handling e08d8bf pci/layerscape: update MSI code for ls1021 rev2 MSI support d217d96 mmc: sdhci-pltfm: enable interrupt mode to detect card 7549302 arm: ls1021a: enable sleep and deep sleep for rev 2.0 silicon 3aa9846 ls1021a: dts: Add eTSEC info for 2nd interrupt groups 5217966 Merge branch 'qoriq-sdk' into LS1-SDK-Rev2.0 02d571e ls102xa: nor: Adjust the partition for NOR flash according to the new mapping table 78fd893 arm: ls1021a: add deep sleep support on TWR-LS1021A-PB b65882a Revert "gianfar: Disable Rx snooping for LS1" 138b929 Revert "gianfar: Remove tx snooping support from LS1" 7b56cb1 arm: ls1021a: add platform notifier for dma-coherent requirement 62a2834 ls102xa: dts: Remove big-endian-regs property for SAI bebea77 Revert "ls102xa: audio: Workaround for SAI data transfer endian issue" a449b22 ls102xa: nor: Adjust the partition for NOR flash
Signed-off-by: Ting Liu <ting.liu@nxp.com> Signed-off-by: Zhenhua Luo <zhenhua.luo@nxp.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
New commit imported: eb3d4fc powerpc/mpc85xx: fix disable port for FMAN erratum A007273 63dfac1 powerpc/mpc85xx: restrict platform for FMAN erratum A007273 54e5fde powerpc/mpc85xx: Add workaround for FMAN erratum A007273 61ac6ba drivers:usb:fsl: Add affected SOCs for USB Erratum A007792 ced4086 fsl/ddr: updated ddr errata-A008378 for arm and power SoCs 17f724f fsl/errata: move fsl_errata.h to common directory 6acff65 arm: ls102x: add get_svr and IS_SVR_REV helper 0caf24c arm: ls102xa: Set fdt_high and initrd_high to the value of 0xffffffff 263a323 powerpc: e6500: Lock/unlock L2 cache instead of L1 as init_ram f28fd70 powerpc: configs: Fix init_ram physical address for several boards 46be0a4 ls102xa: fdt: Disable IFC in SD boot for QSPI f1ca68a ls102xa: etsec: Use proper settings for BE BDs e0712fe mpc85xx/u-boot*.lds: remove _GLOBAL_OFFSET_TABLE_ definition a04b8b3 mmc: fsl_esdhc: fix mmc read/write error on T4160/T4080 23a735b arm: ls1021a: Fix the patch check warnings for the previous commit 5a42d5d SECURE_BOOT: Correct reading of ITS bit f6310eb arm: ls1021a: Add QSPI or IFC support in SD boot ab2db1a arm: ls102xa: fix patch check issue. 7611d2c Merge pull request #40 in SDK/u-boot-devel from ~B29983/u-boot-devel-tyt:feature/QUBOOT-1181-the-u-boot-deepsleep-resume-time to master d406c4f arm: ls102xa: enable all the snoop signal for masters. 1686b84 Merge pull request #38 in SDK/u-boot-devel from ~B46683/u-boot-devel:bugfix/QUBOOT-1173-ls1021atwr-sd-boot-don-t-support to master 0258a33 Merge pull request #37 in SDK/u-boot-devel from ~B18965/my-u-boot-devel:master to master dfc49e9 arm: ls1021atwr: optimize the deep sleep latency 2023326 Merge pull request #34 in SDK/u-boot-devel from ~B40530/u-boot-devel:feature/QUBOOT-789-sd-boot-on-t4-rdb to master 3a8f547 Merge pull request #35 in SDK/u-boot-devel from ~B47093/u-boot-devel-sdhc:feature/QLINUX-3755-support-the-uhs-speed-mode-for to master e4957df Merge pull request #31 in SDK/u-boot-devel from DNNPI/ls1-uboot:feature/LS1-for-SDK1.9 to master 774b080 configs: ls1021atwr: Enable ID EEPROM for SD boot 95a9abb arm: mmu: Add missing volatile for reading SCTLR register c9a7868 mmc: fsl_esdhc: enable EVDD automatic control for SD/MMC legacy card 3fe5a18 Merge pull request #27 in SDK/u-boot-devel from ~B45475/u-boot-devel-fork:feature/QUBOOT-1057-support-qe-tdm-on-t1040rdbd4 to master 767a4ad powerpc/t1040qds: enable peripheral clock for eSDHC 0867fed powerpc/t1040qds: enable adapter card type identification support 26d6853 mmc: fsl_esdhc: enable dat[4:7] for eMMC4.5 acd4c74 powerpc/t4rdb: Add SD boot support for T4240RDB board b773c94 configs: ls1021atwr: Disable DSPI2 with NOR boot 98bf337 mtd: sf: Add support AT26DF081A chip 91bdf34 dm: dts: ls1021a-twr: Enable DSPI2 on LS1021ATWR 7289cb9 configs: ls1021atwr: Enable DSPI for LS1021ATWR 870af21 t1040d4rdb: assign muxed pins to qe-tdm when set hwconfig qe-tdm 300674e powerpc/t1023rdb: change default core frequency to 1200MHz 45f88eb powerpc/t1023rdb: eMMC boot without external SD card aee4875 board/t1023rdb: add support for T1023RDB RevC 2c86026 ls1021atwr: sdboot: change serdes 0x20 to 0x30 8d38d25 LS102XA:workaround:disable priorities within DDR 2621445 arm: ls1021a: allow OCRAM access permission as R/W in SPL 10bc2b0 crypto/fsl: fix "era" property value on LE platforms 6d5733d arm: ls102xa: enable snooping for CAAM transactions d3e6f11 Revert "Enable snooping on transactions from CAAM block" a5dc0f7 arm: ls1021a: Ensure LS1021 ARM Generic Timer CompareValue Set 64-bit 0669b57 LS102XA: Workaround: DDR require a register write 449ab6a ls102xa: Enable snoop and DVM message requests. 6ae8810 Revert "arm: ls102xa: workaround for cache coherency problem" 617f1fd Revert "sf: Add support for flag status register on Micron chips" 18bd7bb arm: ls102xa: Add LE mode support for audio IPs on Rev2.0 silicon 80a22d8 QE: modify the address of qe ucode 784f7f2 arm: ls1021atwr: program the regulator for deep sleep 8971e50 ls1021atwr: added deep sleep support in uboot 9f692ae ls102xa: etsec: Enable Tx Data and TxBD snooping 79da560 arm/ls102xa:Add support of conditional workaround implementation as per SoC ver b6876d0 arm: ls102xa: Adjust the load address of U-Boot for NOR boot
Remove the useless Fix-the-depend-race-issue.patch.
Signed-off-by: Ting Liu <ting.liu@nxp.com> Signed-off-by: Zhenhua Luo <zhenhua.luo@nxp.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Ting Liu [Tue, 22 Dec 2015 16:07:48 +0000 (00:07 +0800)]
usdpaa-apps: use in6_* macros from glibc instead of kernel
Both glibc and the kernel have in6_* macros definition. Use the one from glibc.
Kernel headers will check for previous libc definitions by including
include/linux/libc-compat.h.
Fix the below build error:
| [CC] xfrm_km.c (bin:srio_ipsec_offload)
| In file included from .../tmp/sysroots/b4860qds-64b/usr/include/linux/xfrm.h:4:0,
| from src/srio_ipsec_offload/xfrm_km.c:46:
| .../usr/include/netinet/in.h:99:5: error: expected identifier before numeric constant
| IPPROTO_HOPOPTS = 0, /* IPv6 Hop-by-Hop options. */
| ^
| In file included from .../tmp/sysroots/b4860qds-64b/usr/include/netinet/ip.h:24:0,
| from src/srio_ipsec_offload/xfrm_km.c:52:
| .../tmp/sysroots/b4860qds-64b/usr/include/netinet/in.h:209:8: error: redefinition of 'struct in6_addr'
| struct in6_addr
| ^
| In file included from .../tmp/sysroots/b4860qds-64b/usr/include/linux/xfrm.h:4:0,
| from src/srio_ipsec_offload/xfrm_km.c:46:
| .../tmp/sysroots/b4860qds-64b/usr/include/linux/in6.h:32:8: note: originally defined here
| struct in6_addr {
| ^
Signed-off-by: Ting Liu <ting.liu@freescale.com> Signed-off-by: Zhenhua Luo <zhenhua.luo@nxp.com> Acked-by: Otavio Salvador <otavio@ossystems.com.br>
Zhenhua Luo [Wed, 23 Sep 2015 13:35:39 +0000 (21:35 +0800)]
skmm-ep: fix the inline function definition with gcc 5.x
There are different semantics for inline functions for gcc-5.x compared to
previous gcc. Fix the following build error.
| skmm.c:(.text.startup+0xfc): undefined reference to `va_to_pa'
| skmm.c:(.text.startup+0x10c): undefined reference to `va_to_pa'
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com> Acked-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 2 Dec 2015 13:57:43 +0000 (11:57 -0200)]
xf86-video-imxfb-vivante: Stop using Git to write local version
The standard version does not use a Git repository so we should not
use Git to identify the commit of the build as it can end getting the
version from a wrong repository and can be misleading.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Mon, 23 Nov 2015 12:36:19 +0000 (10:36 -0200)]
kernel-module-imx-gpu-viv: Add 'galcore' to module autoloading
The module autoloading ensures 'galcore' to be probed as early as
possible during the boot. The order is important so the GPU can be
used as fast as possible, ensuring a reduced time until the initial
GPU-based application is loaded.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
The 4.3 version of kernel cause build errors on lttng and cannot be
included until those are fix. We are reverting it now and will add it
back once OE-Core is fixed to support it.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Mon, 9 Nov 2015 11:27:04 +0000 (09:27 -0200)]
linux-fslc: Upgrade to 4.3-based kernel
Besides the revision change, this synchronizes the defconfig used for
the i.MX and the Vybrid SoCs. The only SoCs using a specific defconfig
file are the MXS-based ones (i.MX23 and i.MX28) as the others are
consolidated on the single defconfig file.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
This is a library for using the i.MX6 VPU. It is an alternative to
libfslvpuwrap, hosted on Github, and has an API that features several
improvements over libfslvpuwrap, which include:
* User-defined context information associated with input frames, which is
passed on to corresponding output frames (to be able to identify which
input frame produced which output frame)
* Groundwork for future DMA-BUF/BMM/ION/CMA allocator integration, using
file descriptors instead of physical addresses
* Indicators for when it is safe to try to decode frames, which is
critical in multi-threaded playback cases
* Simplified, higher-level JPEG en/decoding API, based on the VPU MJPEG
codec; useful for picture viewing without the extra boilerplate for
VPU-based en/decoding
Signed-off-by: Carlos Rafael Giani <dv@pseudoterminal.org> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Javier Viguera [Mon, 2 Nov 2015 18:12:53 +0000 (19:12 +0100)]
gst1.0-fsl-plugin: fix FSL codecs, parser and vpuwrap detection
Patch gst1.0-fsl-plugin's configure script to add a trailing space in the
FSL_MM_CORE_CFLAGS and FSL_MM_CORE_LIBS variables between different
pkg-config executions.
These variables are used later to pass parameters to the compiler and
as they are not built correctly, the 'configure' script fails to detect
some of the functionality:
Otavio Salvador [Thu, 29 Oct 2015 11:22:55 +0000 (09:22 -0200)]
Remove DirectFB support for i.MX6
The DirectFB project is dead upstream for a while and the backend is
deprecated so it is better to drop it and avoid new users to depend on
it and instead to focus in the current technologies for new projects.
This commit removes everything related to the DirectFB support so if
someone wish to revert it locally it can be easily done.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Acked-by: Daiane Angolini <daiane.angolini@freescale.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
This recipe is based on the now-deleted linux-imx-rt recipe, using the
same set of patches.
The linux-fslc-mx6 kernel is currently at upstream version 3.14.54, for
which no realtime patch has been released, but patch 3.14.51-rt52
applies cleanly.
Zhenhua Luo [Tue, 10 Nov 2015 05:53:48 +0000 (13:53 +0800)]
linux-qoriq: fix the kernel hang issue with gcc-5.2
Backport a patch to fix the kernel hang issue when the uImage is built
by gcc-5.2.
The details can be found in
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=5e95235
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com>
(backported from meta-fsl-ppc:95d9f9aba81ccdae53011ac8c6fbbc2605b9cedc) Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 28 Oct 2015 14:14:08 +0000 (12:14 -0200)]
linux-imx-mfgtool: Remove leftover from 3.14.38 QP upgrade
The 33615de (linux-imx: Upgrade to 3.14.38-6QP_beta release) did not
remove the linux-imx-mfgtool-3.14.28/mx6/defconfig file, keeping it
unused. Remove it now.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
This includes the Congatec QMX6 SPL support which is targetting
inclusion for U-Boot 2016.01 release. The following patches are
included:
23575d2 cgtqmx6eval: Add DFU support 112dfb1 iomux-v3: Take MX6D in consideration for imx_iomux_v3_setup_pad() bf2ca16 cgtqmx6eval: Add SPL support 3bce451 cgtqmx6eval: Add fastboot support 77b786a cgtqmx6eval: Use SPI NOR to store the environment f6bc396 cgtqmx6eval: Add SPI NOR flash support 540baa3 spi: Add SPI NOR protection mechanism 6e057b9 spi: sf_ops: Add SPI protection mechanism from the kernel 3f150e4 powerpc: Remove __ilog2_u64 and ffs4 from bitops 2feb00c compat: Remove is_power_of_2() definition 254550e nios2: Use the generic bitops headers 7611cdc nds32: Use the generic bitops headers f9bfea4 openrisc: Use the generic bitops headers d2f21eb sparc: Use the generic bitops headers fde093e sandbox: Use the generic bitops headers ef582a7 sh: bitops: Use the generic bitops headers d068eae blackfin: bitops: Use the generic bitops headers 37d32df m68k: bitops: Use the generic bitops headers 902775c4 x86: bitops: Use the generic bitops headers 6ab6bf2 ARM: bitops: Use the generic bitops headers 8324309 include: Add generic bitops headers 5054c77 include: Add log2 header from the kernel
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Tue, 27 Oct 2015 11:15:49 +0000 (09:15 -0200)]
Remove Vybrid mcc related recipes
The mqxboot and mcc Linux kernel module requires the Timesys kernel to
work. The kernel is not in use in meta-fsl-arm anymore and in long
term the Vybrid will be using the remoteproc mechanism to share
resources across the different CPU, inside of the SoC.
As consequence of the update for the Linux 4.1 kernel, those recipes
ought to move for meta-fsl-arm-extra where 3rd party boards can keep
using them.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Fabio Estevam [Sun, 25 Oct 2015 22:07:01 +0000 (20:07 -0200)]
linux-fslc (4.1): Update to latest version fa868c9e
Commit fa868c9e ("dts: imx6: fix sd card gpio polarity specified in
device tree") missed to be applied into 4.1.11 causing esdhc regressions
on mx6, so let's apply it so that we can have SD card working with
4.1.11.
Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Sat, 24 Oct 2015 20:42:26 +0000 (18:42 -0200)]
linux-timesys: Remove unused kernel
The Vybrid reference board (twr-vf65gs10) now uses the linux-fslc
(4.1) kernel. This makes this kernel recipe unused. Remove it as it
has known security flaws, toolchain problems and is quite old.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Stefan Agner [Sat, 24 Oct 2015 05:49:34 +0000 (22:49 -0700)]
twr-vf65gs10: move to Linux 4.1 based linux-fslc tree
Change the default Kernel to the linux-fslc 4.1 kernel.
What seems to work:
- UART
- CPU/Memory (obviously) and L2 Cache
- ARM global timer as clock source
- Pinctrl/GPIO support
- SD-Card
- USB Host (according to bootlog, not tested)
- SoC internal RTC (SNVS RTC)
- Watchdog
Not tested, but should be working:
- Audio using SGTL5000 connected via SAI (hardware not available)
- DCU Framebuffer device (hardware not available)
- Ethernet PHYs (not sure, might be a hardware issue)
Known to be missing:
- Spansion Flash S25FL128S connected via QuadSPI
- MCC v1 and v2
- Touchscreen with TWR-LCD-RGB (according to Tony)
Signed-off-by: Stefan Agner <stefan.agner@toradex.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Andreas Müller [Thu, 22 Oct 2015 19:11:12 +0000 (21:11 +0200)]
efl: fix build with current master meta-oe/oe-core
fixes:
| In file included from /home/a.mueller/tmp/oe-core-glibc/sysroots/zeravarimx6q/usr/include/EGL/eglplatform.h:38:0,
| from /home/a.mueller/tmp/oe-core-glibc/sysroots/zeravarimx6q/usr/include/EGL/egl.h:39,
| from ../src/modules/evas/engines/gl_common/evas_gl_api_ext.h:7,
| from ../src/modules/evas/engines/gl_common/evas_gl_core_private.h:5,
| from modules/evas/engines/gl_generic/evas_engine.c:2:
| /home/a.mueller/tmp/oe-core-glibc/sysroots/zeravarimx6q/usr/include/EGL/eglvivante.h:266:2: error: #error "Platform not recognized"
| #error "Platform not recognized"
| ^
Signed-off-by: Andreas Müller <schnitzeltony@googlemail.com>
Daiane Angolini [Thu, 22 Oct 2015 16:10:22 +0000 (14:10 -0200)]
imx6ulevk: Configure machine to use u-boot SPL
At first, this machine was supported only by u-boot-imx U-Boot provider, however
as soon as u-boot-fslc included this support, the U-Boot provider was changed.
During the change of default U-Boot provider, the SPL configuration end up being
missed.
The needed SPL configuration is set up in this patch.
image_types_fsl.bbclass: Fix image generation for i.MX 6UltraLite
When the i.MX 6UltraLite (mx6ul) was separated from the SoC family mx6
(commit 24396b28f0137464cd82b9a19f705c1db180ead2) it started to create
invalid sdcard images.
This is because of SDCARD_GENERATION_COMMAND no more being triggered.
The solution in this moment is add a mx6ul override for the
SDCARD_GENERATION_COMMAND definition.
Signed-off-by: Felipe Provenzano <felipeprovenzano@maxtrack.com.br> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Tue, 20 Oct 2015 11:47:25 +0000 (09:47 -0200)]
Remove unused patch files
Those patch files were left orphan when upgrading the linux-imx and
u-boot-imx recipes, becoming unused. The removal of them cause no harm
and helps to keep the metadata clean.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Update the U-Boot fork to a 2015.10-rc5 based version. This brings
several fixes for i.MX and Vybrid based systems. Mainly:
cf04ad3 arm: vf610twr: improve memory layout d45fd01 colibri_vf: Fix bstlen field e24bb2b mtd: nand: vf610_nfc: resync with upstream Linux version f861f51 ls102xa: Fix reset hang f532727 imx_watchdog: Add a header file for watchdog registers 040ef8f pcm052: fix MTD partitioning 61903b7 imximage: fix commands other than write_data 835c30e imximage: header.length of 4 is valid
Change-Id: Ifdb607d9a94ef594f3fe37d2b6388dc9a8268106 Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 14 Oct 2015 18:26:53 +0000 (15:26 -0300)]
image_types_fsl.bbclass: Fail if a Device Tree file is missing
When generating the rootfs we need to include all the referenced
Device Tree files. The code were gracefully checking if the file
existed before building the image however the expected behavior is to
fail.
As consequence of this change, if someone is using a very old kernel
(no Device Tree based) and tries to generate an image, it will fail
until the machine definition is changed to not list Device Tree files
for installation.
Currently all machines supported by meta-fsl-arm BSP are consistent
for the default kernel in use however custom BSP may need to be
changed accordingly.
Change-Id: I548c65c61dfefe24154a455d097032b1598b14d3 Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
The code which copies the boot scripts and Device Tree files is
duplicated across the different image types. This patch reworks this
code to avoid this duplication.
Change-Id: Ib9405067a5419e1b295dfac9e3d3bd9b896d424e Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Stefan Agner [Thu, 15 Oct 2015 05:28:52 +0000 (22:28 -0700)]
linux-fslc (4.1): Add Vybrid SoC to compatible machines
Upstream Linux 4.1 supports the Vybrid SoC quite well, so enable
linux-fslc also for Vybrid SoC's.
Change-Id: I6745b67f4cc4b12ccc6f05754af9241e60731c6a Signed-off-by: Stefan Agner <stefan.agner@toradex.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 14 Oct 2015 17:33:07 +0000 (14:33 -0300)]
imx-base.inc: Set a default provider for u-boot-mxsboot-native
The O.S. Systems Embedded Linux provides an alternative version of
U-Boot and, for completeness, an alternative recipe for
u-boot-mxsboot-native.
The provider, in imx-base.inc, had no default thus raising a
warning. This turns the u-boot-fslc-mxsboot-native as the default
provider but allows custom distributions to override it.
Change-Id: I46799f9ec744a074d95c383561cf0cbddb26e151 Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 14 Oct 2015 14:36:04 +0000 (11:36 -0300)]
opencv: Use the 2.4 release only
The opencv 3.0 is not tested and does not have the i.MX6 specific
patch ported so we disable it for the i.MX6 SoC from Quad, DualLite,
SoloX and SoloLite families. The 2.4 is used instead.
This fixes following warning:
,----
| NOTE: multiple providers are available for runtime opencv-samples (opencv-samples, opencv)
| NOTE: consider defining a PREFERRED_PROVIDER entry to match opencv-samples
`----
Change-Id: I2f897cdd31dcee9046ee7d28ef4dcadb154ae88c Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 14 Oct 2015 14:26:01 +0000 (11:26 -0300)]
mesa-gl: Blacklist the recipe for i.MX6 SoCs with GPU
The mesa-gl is not in sure for the i.MX6 SoC from Quad, DualLite,
SoloX and SoloLite families. The full featured, but trimmed, mesa is
used instead.
This fixes following warning:
,----
| NOTE: multiple providers are available for runtime libgl-mesa-dev (mesa, mesa-gl)
| NOTE: consider defining a PREFERRED_PROVIDER entry to match libgl-mesa-dev
| NOTE: multiple providers are available for runtime mesa-driver-swrast (mesa, mesa-gl)
| NOTE: consider defining a PREFERRED_PROVIDER entry to match mesa-driver-swrast
`----
Change-Id: I082b8f7a739e4ea3cfe5f1760c7163d118e9883c Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Zhenhua Luo [Tue, 13 Oct 2015 06:15:55 +0000 (14:15 +0800)]
cryptodev-qoriq: Fix runtime providers so 'cryptodev' is not taken as provider
QorIQ targets should use FSL custom cryprodev instead of the one provided by
poky, fix the runtime providers to ensure 'cryptodev-qoriq' is taken as
provider.
Signed-off-by: Zhenhua Luo <zhenhua.luo@freescale.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Tue, 29 Sep 2015 21:22:22 +0000 (18:22 -0300)]
linux-fslc-mx6: Update revision to 964e5a3 revision
The 964e5a3 revision includes the 3.14.54 stable release fixes plus
following changes contributed by Gateworks:
7292dc7 ASoC: add tda1997x codec support aa4ee3d ipu3: keep buffer offset aligned 05608b5 ipu3: fix stripe calculation 30c6d90 thermal: imx: allow crit_delta to be passed in via kernel param 334fc35 arm: dts: add handles to anatop regulators c7a4d85 fec: Russell Kings corruption fix 0c2c11f net: fec: fix the warning found by dma debug b5b53b5 pci: imx6: Add delay after LTSSM transition 6dba6b0 fec_main: add mb() before setting status 05ab2a6 net: fec: fix NULL pointer dereference in fec_enet_timeout_work 07465db drivers: video: mxc: Add support for interlace on hdmi out 3eca3ca imx: thermal: use CPU temperature grade info for thresholds 172ad49 media: pci: tw6869: Implement ENUM_FRAMESIZES ioctl a7d429e media: pci: tw6869: add device registered dev info 957f3fa media: pci: tw6869: initial add cb61124 ARM: dts: imx: fix invalid #address-cells value ea53bf7f imx:busfreq: disable busfreq driver by default e3e114d video: mxc: ldb: use native mode for display timings fa6b13a net: igb: register mii_bus for SerDes w/ external phy 09359b6 net: igb: add phy read/write functions that accept phy addr d6daeac net: igb: add i210/i211 support for phy read/write 00b2ceb video: mxc: add adv739x TV encoder 3117a93 mxc_capture: add tda1997x HDMI receiver video support 9348099 imx: add imx tda1997x i2s DAI support b81d263 mfd: add tda1997x HDMI receiver support dba225f mxc_capture: check for init_camera_struct failure 09f7c1f mxc_capture: add device registration info 1db12fc mxc_hdmi: present sysfs edid as binary data instead of ascii bd0601d mxc_hdmi: enable overflow interrupt after initialization complete d6cc9e6 mxc_ipuv3: show device name for overlay fb a7f44be mxc_ipuv3: show ipu_id/disp_id of registered device e9ea090 regulator: Add LTC3676 support 3e12a168 regulator: core: Fix the init of DT defined fixed regulators 154b8be hwmon: mma8451: add fxos8700 magnetometer support e19d671 hwmon: mma8451: add fxos8700 accel support 760c31c cpufreq: imx: add a 25mV tollerance to the operating point voltages ef2eda1 cpufreq: imx: show soc voltage when debugging b205762d cpufreq: imx: init mutex before it can be used 04bc9ed input: edt-ft5x06 - add device-tree config for screen res and inversion 1ca89b6 Input: edt-ft5x06 - fix an i2c write for M09 support dc5f6c9 Input: edt-ft5x06 - add a missing condition f2854dd Input: edt-ft5x06 - add support for M09 firmware version 390394c Input: edt-ft5x06 - ignore touchdown events 6b60fb3 Input: edt-ft5x06 - adjust delays to conform datasheet 74bceb0 Input: edt-ft5x06 - add DT support 481da81 input: goodix: add invert and rotate support via device-tree 6297f18 input: goodix: add support for performing init sqequence 4746ef3 input: goodix: add device-tree support
For compatibility with the Vivante kernel module build, the following
commit has be also applied:
964e5a3 FSLC: Add busfreq-imx.h header for forward compatibility
Change-Id: I92b566943c7245751b484302f2320b42ec7d761c Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 7 Oct 2015 01:38:49 +0000 (22:38 -0300)]
kernel-module-imx-gpu-viv: Fork driver for FSL Community
The driver source code provided by Freescale needs fixes. This fork is
a community driven development and allow faster development and easier
integration of fixes.
The "+fslc" suffix is added to the module during probing to make a
clear identification.
Change-Id: I914f1a7731263effdef75827e3a9fb8daba37ec9 Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 7 Oct 2015 18:28:22 +0000 (15:28 -0300)]
imx-base.inc: Ensure MACHINE_HAS_VIVANTE_KERNEL_DRIVER_SUPPORT is expanded
The MACHINE_HAS_VIVANTE_KERNEL_DRIVER_SUPPORT needs to be expanded and
cannot have a default overriden later, otherwise it does not respect
the default set in the SoC.
Change-Id: I1d90d173603e6ee258f3e05c7a53e576b3b1f552 Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 16 Sep 2015 19:04:31 +0000 (16:04 -0300)]
u-boot-fslc: Update to 2015.10-based fork
This updates the fork based on upcoming 2015.10 release. This is
currently based on 2015.10-rc3 tag and offers a very solid release for
start the tests for Yocto Project 2.0 release.
Change-Id: I1d35409a9806a96f8dd538e2b8c2a3520e9104ba Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Gary Thomas [Tue, 6 Oct 2015 16:03:47 +0000 (10:03 -0600)]
kernel-module-imx-gpu-viv: Better work-around for change in name of busfreq-imx.h
The renaming of include/linux/busfreq-imx6.h to include/linux/busfreq-imx.h
is not consistent over the many kernel versions currently being built. This
changeset works around this inconsistency by creating a local symbolic link
to whatever file is actually present in the kernel sources.
Gary Thomas [Mon, 5 Oct 2015 22:09:17 +0000 (16:09 -0600)]
qt4: Fix *.la pollution
Recent [platform dependent] changes to QT_CONFIG_FLAGS allow
some C-preprocessor defines to creep into *.la (libtool) files.
This are not allowed (or even handled) and cause linker errors
when trying to link against these libraries.
This patch removes the incorrect/erroneous preprocessor directives
to leave the *.la files useable once again. The change is done in
a post-processing step to clean up the *.la files as they are only
used externally to this package.
Signed-off-by: Gary Thomas <gary@mlbassoc.com>
Upstream-status: Innapropriate [OE patch causes the problem] Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
--Upgrade gst1.0-fsl-plugin to version 4.0.7 released in 3.14.38-6UL_ga and 3.14.38-6QP_ga releases.
--Add gstreamer-plugins-bad as one of the dependency.
--Imxcompositor enhancement:
1. fix alpha blending capbility query issue of g2d.
2. add zorder feature.
3. add keep ratio feature.
4. find best output format by leverage the total complexity and picture quality.
Signed-off-by: Yuqing Zhu <b54851@freescale.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Mon, 21 Sep 2015 20:47:06 +0000 (17:47 -0300)]
Move mxs-base.inc contents to imx-base.inc
The consolidation of all i.MX related base settings allows a more
global view of the settings in place.
One of the reasons for this patch is the fragmentation caused by
i.MX23 and i.MX28 SoCs using mxs-base.inc.
The changes needed to get i.MX23 and i.MX28 SoCs working have been
included in imx-base.inc file already and some values required rework
to apply to specific SoC families to avoid regressions.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Thu, 17 Sep 2015 17:09:14 +0000 (14:09 -0300)]
imx-base.inc, mxs-base.inc: Change default image to sdcard.gz
Replace .sdcard image to sdcard.gz image as a default image type for
all machines.
The SDCard class has the feature of compress the generated .sdcard image.
It can be used to ease the exchange of this image among developers or machines.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Thu, 17 Sep 2015 14:14:10 +0000 (11:14 -0300)]
imx-base.inc, mxs-base.inc: Stop generating tar.bz2 image tarball by default
The tar.bz2 is useful for people using NFS or generating the SD card
image byhand, those can enable this in the board file or in the
local.conf when needed. By default the SD card image should be the
only image built as it is the most commonly used one for initial
development.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Wed, 23 Sep 2015 12:53:31 +0000 (09:53 -0300)]
image_types_fsl.bbclass: Fix DeviceTree copy for FAT partition
The rootfs generation script now runs outside of the deploy directory
so the code need to use the full path for the installed files. This
fixes the rootfs generation process so the files are copied to the FAT
partition, allowing the system to boot.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br> Tested-by: Daiane Angolini <daiane.angolini@freescale.com>
Otavio Salvador [Tue, 22 Sep 2015 22:20:04 +0000 (19:20 -0300)]
kernel-module-imx-gpu-viv: Fix build against 3.14-based kernels - again
The build fail about the bus frequency header (linux/busfreq-imx6.h)
not being found is caused by the mistaken check for the wrong kernel
version.
The regression has been introduced by d476c35
(kernel-module-imx-gpu-viv: Upgrade to 5.0.11.p7.1 for 3.14.38-6QP_ga
release) which has been not properly tested. This patch fixes it,
again, by adding back the removed patch.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Otavio Salvador [Tue, 11 Aug 2015 20:32:47 +0000 (17:32 -0300)]
Enable Vivante kernel driver dynamic provider
The linux-imx.inc based kernels now can dynamically provide the
Vivante kernel driver and satisfy the runtime dependency added for the
Vivante GPU user space libraries. The beauty of this solution is that
it makes ease for 3rd party kernel to reuse the same infrastructure
and use the external module easily.
Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
kernel-module-imx-gpu-viv: Upgrade to 5.0.11.p7.1 for 3.14.38-6QP_ga release
- Package name is changed to kernel-module-imx-gpu-viv.
- Patch platform-Fix-kernel-version-check-for-3.14-based-ker.patch is no longer required
as the fix has been applied upstream.
- Upgrade to graphics version 5.0.11.p7.1
- Makefile patch will be included in next release.
Signed-off-by: Neena Busireddy <neena.busireddy@freescale.com> Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Ting Liu [Wed, 12 Aug 2015 08:51:09 +0000 (16:51 +0800)]
adjust MACHINEOVERRIDES ordering
FILESPATH is in part built by using MACHINEOVERRIDES and the path
is built working right to left. Use =. instead of .= to prepend
(e6500-64b:qoriq-ppc) to (soc:board).
Take t2080rdb-64b as example, change from:
FILESOVERRIDES="powerpc64:t2:t2080:t2080rdb-64b:qoriq-ppc:e6500-64b:poky"
MACHINEOVERRIDES="t2:t2080:t2080rdb-64b:qoriq-ppc:e6500-64b"
To:
FILESOVERRIDES="powerpc64:e6500-64b:qoriq-ppc:t2:t2080:t2080rdb-64b:poky"
MACHINEOVERRIDES="e6500-64b:qoriq-ppc:t2:t2080:t2080rdb-64b"
