Message ID | alpine.LFD.2.00.1104040913320.20731@right.am.freescale.net |
---|---|
State | Accepted |
Delegated to: | Kumar Gala |
Headers | show |
On Mon, Apr 4, 2011 at 9:14 AM, Kumar Gala <galak@kernel.crashing.org> wrote: > Timur Tabi (2): > p1022ds: allow for board-specific ngPIXIS functions > fsl: obsolete NXID v0 EEPROMs, automatically upgrade them to NXID v1 What about: powerpc: clean up DIU macro definitions for Freescale reference boards powerpc: fix implementation of out_8 to match the other out_XX functions powerpc/85xx: introduce 'fdt verify' command You can find them at http://patchwork.ozlabs.org/project/uboot/list/?submitter=timur
On Apr 4, 2011, at 12:48 PM, Timur Tabi wrote: > On Mon, Apr 4, 2011 at 9:14 AM, Kumar Gala <galak@kernel.crashing.org> wrote: > >> Timur Tabi (2): >> p1022ds: allow for board-specific ngPIXIS functions >> fsl: obsolete NXID v0 EEPROMs, automatically upgrade them to NXID v1 > > What about: > > powerpc: clean up DIU macro definitions for Freescale reference boards need WD's ack. > powerpc: fix implementation of out_8 to match the other out_XX functions Will pull in. > powerpc/85xx: introduce 'fdt verify' command I'm wondering if its better to just set the dtb props instead of verifying. - k
On Mon, 4 Apr 2011 13:20:57 -0500 Kumar Gala <galak@kernel.crashing.org> wrote: > On Apr 4, 2011, at 12:48 PM, Timur Tabi wrote: > > > powerpc/85xx: introduce 'fdt verify' command > > I'm wondering if its better to just set the dtb props instead of verifying. It's not realistic to patch everything (if we do that, might as well just generate the tree from scratch), and likewise fdt verify doesn't check everything. It checks some important bits to detect common errors such as mixing 32/36 bit. Patching a tree with such an error would not result in a working tree. -Scott
Kumar Gala wrote: >> > powerpc/85xx: introduce 'fdt verify' command > I'm wondering if its better to just set the dtb props instead of verifying. In the long run, I agree. But this code is available now and it could help. The verify command is easier to implement because it doesn't require that a particular PCI address match a particular PCI node. That is, I search all PCI nodes and all addresses within a node to find a match. Setting the dtb props would require hard-coded matches between the U-Boot macros and the PCI nodes. It would be a significant amount of work, and each board would need to be updated, I think.
Dear Kumar Gala, In message <16CF6CAD-477C-4E04-93E0-1DADCF05C724@kernel.crashing.org> you wrote: > > > powerpc: fix implementation of out_8 to match the other out_XX functions > > Will pull in. Why you? This is in no way 85xx specific. Why does this come up only now? Why has it not been mentioned in the v2010.12 and v2011.03 relese cycles? Timur: does this fix any real problem? > > powerpc/85xx: introduce 'fdt verify' command > > I'm wondering if its better to just set the dtb props instead of verifying. Unfortunately Timur's posting broke the mail thread because he failed to add keep reference headers in his patch postings. I'm not happy with this implementation either. In any case this needs the ACK from the responsible custodian (FDT => gvb). Best regards, Wolfgang Denk
Wolfgang Denk wrote: >>> powerpc: fix implementation of out_8 to match the other out_XX functions >> > >> > Will pull in. > Why you? This is in no way 85xx specific. > > Why does this come up only now? Why has it not been mentioned in the > v2010.12 and v2011.03 relese cycles? I posted the patch four months ago. Did I miss something? > Timur: does this fix any real problem? Not to my knowledge, but I suspect that people want out_8 to have the same synchronization behavior as the other out_xx functions. >>> > > powerpc/85xx: introduce 'fdt verify' command >> > >> > I'm wondering if its better to just set the dtb props instead of verifying. > Unfortunately Timur's posting broke the mail thread because he failed > to add keep reference headers in his patch postings. I'm not doing anything special when I send email, so I don't know what you're talking about. > I'm not happy > with this implementation either. Can you be more specific? I posted this patch almost five months ago. I addressed all of your posted concerns from v2 of this patch. Other than a few additional fdt helper functions, the code is isolated to 85xx.
On Mon, 4 Apr 2011 21:42:27 +0200 Wolfgang Denk <wd@denx.de> wrote: > Dear Kumar Gala, > > In message <16CF6CAD-477C-4E04-93E0-1DADCF05C724@kernel.crashing.org> you wrote: > > > > > powerpc: fix implementation of out_8 to match the other out_XX functions > > > > Will pull in. > > Why you? This is in no way 85xx specific. > > Why does this come up only now? Why has it not been mentioned in the > v2010.12 and v2011.03 relese cycles? > > Timur: does this fix any real problem? It came up here: http://www.mail-archive.com/u-boot@lists.denx.de/msg43557.html I don't know of a visible symptom of this as the codebase currently stands, but it's a bug waiting to happen. The semantics are different from every other accessor, and different from Linux (with which U-Boot shares a fair amount of driver code). -Scott
Dear Scott Wood, In message <20110404145528.1bd34cea@schlenkerla.am.freescale.net> you wrote: > > > Why does this come up only now? Why has it not been mentioned in the > > v2010.12 and v2011.03 relese cycles? > > > > Timur: does this fix any real problem? > > It came up here: > http://www.mail-archive.com/u-boot@lists.denx.de/msg43557.html > > I don't know of a visible symptom of this as the codebase currently stands, > but it's a bug waiting to happen. The semantics are different from every > other accessor, and different from Linux (with which U-Boot shares a fair > amount of driver code). Agreed. Timur, can you please add this information to the commit message? Best regards, Wolfgang Denk
On Apr 4, 2011, at 2:42 PM, Wolfgang Denk wrote: > Dear Kumar Gala, > > In message <16CF6CAD-477C-4E04-93E0-1DADCF05C724@kernel.crashing.org> you wrote: >> >>> powerpc: fix implementation of out_8 to match the other out_XX functions >> >> Will pull in. > > Why you? This is in no way 85xx specific. Only because it was assigned to me on patchworks. (I might have done that but assumed you did). I'm happy to leave this to you as a generic PPC issue. - k
Dear Kumar Gala, In message <alpine.LFD.2.00.1104040913320.20731@right.am.freescale.net> you wrote: > The following changes since commit 19b54a701811220221fc4d5089a2bb18892018ca: > > Prepare v2011.03 (2011-03-31 23:45:36 +0200) > > are available in the git repository at: > git://git.denx.de/u-boot-mpc85xx.git master Pulling fails with a merge conflict in drivers/usb/host/ehci-pci.c; please check. > Dipen Dudhat (1): > powerpc/85xx: Add support for Integrated Flash Controller (IFC) > > Haiying Wang (2): > powerpc/85xx: Refactor Qman/Portal support to be shared between SoCs > powerpc/85xx: load ucode from nand flash before qe_init > > Jerry Huang (1): > powerpc/85xx: Add eSDHC support on P2020DS > > Jiang Yutang (1): > powerpc/85xx: Add 36-bit address map support to P1022DS > > Kumar Gala (16): > powerpc/85xx: Remove DATARATE_*_MHZ defines in static ddr init > powerpc/85xx: Declare fsl_ddr_set_memctl_regs in <asm/fsl_ddr_sdram.h> > powerpc/85xx: Add some defines for P2040, P3041, P5010, P5020 > powerpc: Move cpu specific lmb reserve to arch_lmb_reserve > powerpc/85xx: Remove config.mk for nand linker script > powerpc/8xxx: Replace fsl_ddr_get_mem_data_rate with get_ddr_freq() > powerpc/8xxx: Refactor fsl_ddr_get_spd into common code from board > powerpc/85xx: Cleanup some QE related defines > powerpc: Add cpu_secondary_init_r to allow for initialization post env setup > powerpc/85xx: Don't build read_tlbcam_entry for CONFIG_NAND_SPL > powerpc/8xxx: Fix LAW init to respect pre-initialized entries > powerpc/85xx: Support for Freescale P1024/P1025 processor > powerpc/85xx: Specify CONFIG_SYS_FM_MURAM_SIZE > powerpc/85xx: Fixup determining PME, FMan freq > powerpc/85xx: Fix determining Fman freq on P1023 > powerpc/8xxx: Fix typo for address hashing message > > Kyle Moffett (2): > fsl_ddr: Don't use full 64-bit divides on 32-bit PowerPC > mpc8xxx: DDR2/DDR3: Clean up DIMM-type switch statements > > Laurentiu TUDOR (1): > powerpc/85xx: Fix setting of LIODN prop in PCIe nodes on P3041/P5020 > > Li Yang (2): > powerpc/85xx: Update P2020DS default env settings > tsec: add AR8021 PHY support > > Pankaj Chauhan (1): > powerpc/85xx: Extend ethernet device tree stashing parameters for "fsl,etsec2" > > Poonam Aggrwal (6): > powerpc/85xx: Use BR_PHYS_ADDR macro to setup BRs on P1_P2_RDB > fsl_ddr: Adds 16 bit DDR Data width option > powerpc/85xx: Adds some P1010/P1014 SoC configuration defines > powerpc/85xx: Removed P1/P2 RDB RevB support > powerpc/85xx: Optimized DDR settings for 800MT/s on P1/P2 RDB > powerpc/85xx: Add 36-bit physical addressing support for P1_P2_RDB > > Prabhakar Kushwaha (2): > powerpc/85xx: Add SERDES support for P1010/P1014 > fsl_pci: Add support for FSL PCIe controllers v2.x > > Priyanka Jain (3): > powerpc/85xx: Corrected sdhc clock value for P1010 > powerpc/85xx: Use DDR for RAMBOOT instead of L2 SRAM on p1_p2_rdb > powerpc/85xx: Read board switch settings on p1_p2_rdb > > Roy Zang (1): > powerpc/85xx: Add support for Freescale P1023/P1017 Processors > > Timur Tabi (2): > p1022ds: allow for board-specific ngPIXIS functions > fsl: obsolete NXID v0 EEPROMs, automatically upgrade them to NXID v1 > > York Sun (2): > powerpc/85xx: Update fixed DDR3 timing table for P4080DS > powerpc/8xxx: Display DIMM model > > Zhao Chenhui (4): > echi: add ULI1575 PCI ID > powerpc/85xx: Replace memctl_intlv_ctl with hwconfig on MPC8572DS > powerpc/85xx: Disable ECC in considering performance on MPC8572DS > powerpc/85xx: Add support for ULI1575 PCI EHCI module on MPC8572DS > > bhaskar upadhaya (1): > powerpc/85xx: Update timer-frequency prop in ptp_timer node of device tree > > arch/powerpc/cpu/mpc85xx/Makefile | 14 +- > arch/powerpc/cpu/mpc85xx/cpu.c | 7 +- > arch/powerpc/cpu/mpc85xx/cpu_init.c | 32 +- > arch/powerpc/cpu/mpc85xx/fdt.c | 12 + > arch/powerpc/cpu/mpc85xx/p1010_serdes.c | 75 ++ > arch/powerpc/cpu/mpc85xx/p1023_serdes.c | 53 ++ > arch/powerpc/cpu/mpc85xx/p3041_ids.c | 10 +- > arch/powerpc/cpu/mpc85xx/p4080_ids.c | 8 +- > arch/powerpc/cpu/mpc85xx/p5020_ids.c | 10 +- > arch/powerpc/cpu/mpc85xx/portals.c | 32 +- > arch/powerpc/cpu/mpc85xx/speed.c | 48 +- > arch/powerpc/cpu/mpc85xx/tlb.c | 4 +- > arch/powerpc/cpu/mpc8xxx/Makefile | 1 + > arch/powerpc/cpu/mpc8xxx/cpu.c | 14 +- > arch/powerpc/cpu/mpc8xxx/ddr/ctrl_regs.c | 4 +- > arch/powerpc/cpu/mpc8xxx/ddr/ddr.h | 1 - > arch/powerpc/cpu/mpc8xxx/ddr/ddr2_dimm_params.c | 23 +- > arch/powerpc/cpu/mpc8xxx/ddr/ddr3_dimm_params.c | 43 +- > .../cpu/mpc8xxx/ddr/lc_common_dimm_params.c | 11 +- > arch/powerpc/cpu/mpc8xxx/ddr/main.c | 63 ++- > arch/powerpc/cpu/mpc8xxx/ddr/util.c | 60 +- > arch/powerpc/cpu/mpc8xxx/fsl_ifc.c | 84 ++ > arch/powerpc/include/asm/config.h | 2 + > arch/powerpc/include/asm/config_mpc85xx.h | 117 +++ > arch/powerpc/include/asm/fsl_ddr_sdram.h | 3 + > arch/powerpc/include/asm/fsl_ifc.h | 957 ++++++++++++++++++++ > arch/powerpc/include/asm/fsl_law.h | 4 +- > arch/powerpc/include/asm/fsl_liodn.h | 14 +- > arch/powerpc/include/asm/fsl_pci.h | 20 +- > arch/powerpc/include/asm/immap_85xx.h | 35 +- > arch/powerpc/include/asm/immap_qe.h | 98 +-- > arch/powerpc/include/asm/processor.h | 12 + > arch/powerpc/lib/board.c | 14 + > arch/powerpc/lib/bootm.c | 5 + > board/freescale/common/ngpixis.c | 56 +- > board/freescale/common/ngpixis.h | 8 +- > board/freescale/common/sys_eeprom.c | 50 +- > board/freescale/corenet_ds/corenet_ds.c | 11 +- > board/freescale/corenet_ds/ddr.c | 49 +- > board/freescale/corenet_ds/p4080ds_ddr.c | 24 +- > board/freescale/mpc8536ds/config.mk | 30 - > board/freescale/mpc8536ds/ddr.c | 26 - > board/freescale/mpc8540ads/ddr.c | 30 - > board/freescale/mpc8541cds/ddr.c | 26 - > board/freescale/mpc8544ds/ddr.c | 27 - > board/freescale/mpc8548cds/ddr.c | 27 - > board/freescale/mpc8555cds/ddr.c | 26 - > board/freescale/mpc8560ads/ddr.c | 30 - > board/freescale/mpc8568mds/ddr.c | 28 - > board/freescale/mpc8569mds/config.mk | 30 - > board/freescale/mpc8569mds/ddr.c | 28 - > board/freescale/mpc8572ds/config.mk | 30 - > board/freescale/mpc8572ds/ddr.c | 28 - > board/freescale/mpc8572ds/mpc8572ds.c | 11 +- > board/freescale/mpc8610hpcd/ddr.c | 26 - > board/freescale/mpc8641hpcn/ddr.c | 37 +- > board/freescale/mpc8641hpcn/mpc8641hpcn.c | 11 +- > board/freescale/p1022ds/ddr.c | 23 - > board/freescale/p1022ds/diu.c | 66 ++- > board/freescale/p1022ds/p1022ds.c | 13 +- > board/freescale/p1_p2_rdb/config.mk | 31 - > board/freescale/p1_p2_rdb/ddr.c | 47 +- > board/freescale/p1_p2_rdb/p1_p2_rdb.c | 66 +- > board/freescale/p1_p2_rdb/tlb.c | 22 +- > board/freescale/p2020ds/ddr.c | 24 - > board/freescale/p2020ds/p2020ds.c | 23 +- > board/sbc8548/ddr.c | 27 - > board/sbc8560/ddr.c | 30 - > board/sbc8641d/ddr.c | 35 - > board/sbc8641d/sbc8641d.c | 9 - > board/socrates/ddr.c | 27 - > board/stx/stxgp3/ddr.c | 30 - > board/stx/stxssa/ddr.c | 29 - > board/xes/xpedite517x/ddr.c | 32 +- > board/xes/xpedite517x/xpedite517x.c | 9 - > board/xes/xpedite520x/ddr.c | 22 +- > board/xes/xpedite537x/ddr.c | 22 +- > board/xes/xpedite537x/xpedite537x.c | 9 - > board/xes/xpedite550x/ddr.c | 20 +- > board/xes/xpedite550x/xpedite550x.c | 9 - > boards.cfg | 13 + > common/ddr_spd.c | 2 +- > drivers/misc/fsl_law.c | 20 + > drivers/net/tsec.c | 23 + > drivers/pci/fsl_pci_init.c | 12 +- > drivers/usb/host/ehci-pci.c | 1 + > include/common.h | 4 + > include/configs/MPC8536DS.h | 1 + > include/configs/MPC8569MDS.h | 4 +- > include/configs/MPC8572DS.h | 17 +- > include/configs/MPC8610HPCD.h | 2 +- > include/configs/P1022DS.h | 48 +- > include/configs/P1_P2_RDB.h | 83 ++- > include/configs/P2020DS.h | 49 +- > include/configs/corenet_ds.h | 1 + > include/configs/xpedite550x.h | 2 +- > include/ddr_spd.h | 28 +- > 97 files changed, 2233 insertions(+), 1241 deletions(-) > create mode 100644 arch/powerpc/cpu/mpc85xx/p1010_serdes.c > create mode 100644 arch/powerpc/cpu/mpc85xx/p1023_serdes.c > create mode 100644 arch/powerpc/cpu/mpc8xxx/fsl_ifc.c > create mode 100644 arch/powerpc/include/asm/fsl_ifc.h > delete mode 100644 board/freescale/mpc8536ds/config.mk > delete mode 100644 board/freescale/mpc8569mds/config.mk > delete mode 100644 board/freescale/mpc8572ds/config.mk > delete mode 100644 board/freescale/p1_p2_rdb/config.mk Applied, thanks. Best regards, Wolfgang Denk
On 04/04/2011 03:42 PM, Wolfgang Denk wrote: > Dear Kumar Gala, > > In message<16CF6CAD-477C-4E04-93E0-1DADCF05C724@kernel.crashing.org> you wrote: [snip] > Timur: does this fix any real problem? > >>> powerpc/85xx: introduce 'fdt verify' command >> >> I'm wondering if its better to just set the dtb props instead of verifying. > > Unfortunately Timur's posting broke the mail thread because he failed > to add keep reference headers in his patch postings. I'm not happy > with this implementation either. > > In any case this needs the ACK from the responsible custodian (FDT => > gvb). I don't see any problem with the patch other than a minor typo On the last line, s/patch/match./ >> +/* >> + * Verify the addresses for all of the PCI controllers >> + * >> + * PCI is complicated because there is no correlation between the numbering >> + * of the controllers by U-Boot and the numbering the device tree. So we need >> + * to search all of the aliases until we find a patch >> + */ It appears to provide a valuable service. Acked-by: Gerald Van Baren <vanbaren@cideas.com> > Best regards, > > Wolfgang Denk >