- 21 Jun, 2015 40 commits
-
-
Paul Burton authored
Migrate the JZ4740 & the qi_lb60 board to use common clock framework via the new Ingenic SoC CGU driver. Note that the JZ4740-specific debugfs code is removed since common clock framework provides its own debug capabilities. Signed-off-by:
Paul Burton <paul.burton@imgtec.com> Co-authored-by:
Paul Cercueil <paul@crapouillou.net> Cc: Ian Campbell <ijc+devicetree@hellion.org.uk> Cc: Kumar Gala <galak@codeaurora.org> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mike Turquette <mturquette@linaro.org> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Rob Herring <robh+dt@kernel.org> Cc: Stephen Boyd <sboyd@codeaurora.org> Cc: devicetree@vger.kernel.org Cc: linux-clk@vger.kernel.org Cc: linux-mips@linux-mips.org Cc: Fabian Frederick <fabf@skynet.be> Cc: Deng-Cheng Zhu <dengcheng.zhu@imgtec.com> Cc: Linus Walleij <linus.walleij@linaro.org> Cc: Stephen Warren <swarren@wwwdotorg.org> Cc: linux-kernel@vger.kernel.org Cc: Brian Norris <computersforpeace@gmail.com> Patchwork: https://patchwork.linux-mips.org/patch/10151/Signed-off-by:
Ralf Baechle <ralf@linux-mips.org>
-
Paul Burton authored
This driver supports the CGU clocks for Ingenic SoCs. It is generic enough to be usable across at least the JZ4740 to the JZ4780, and will be made use of on such devices in subsequent commits. This patch by itself only adds the SoC-agnostic infrastructure that forms the bulk of the CGU driver for the aforementioned further commits to make use of. Signed-off-by:
-
Paul Burton authored
Document the devicetree binding for Ingenic SoC CGUs, and add headers defining the clock specifiers for clocks provided by the JZ4740 & JZ4780 CGU blocks. Signed-off-by:
-
Paul Burton authored
Replace uses of the jz4740_clock_bdata struct with calls to clk_get_rate for the appropriate clock. This is in preparation for migrating the clocks towards common clock framework & devicetree. Signed-off-by:
-
Paul Burton authored
Call jz4740_clock_init before any uses of jz4740_clock_bdata occur. This is in preparation for replacing uses of that struct with calls to clk_get_rate, which will allow the clocks to be migrated towards common clock framework & devicetree. Signed-off-by:
-
Paul Burton authored
Move the driver for Ingenic SoC interrupt controllers into drivers/irqchip where it belongs. Signed-off-by:
-
Paul Burton authored
Allow the interrupt controllers of the JZ4770, JZ4775 & JZ4780 SoCs to be probed via devicetree, supporting the 64 interrupts they provide. Signed-off-by:
-
Paul Burton authored
Rename the functions including jz4740 in their names to be more generic in preparation for supporting further SoCs, and for moving this interrupt controller code to drivers/irqchip. Signed-off-by:
-
Paul Burton authored
Read the base address of the SoC interrupt controller from the device tree rather than relying upon the JZ4740_INTC_BASE_ADDR macro, in order to remove the dependency on the asm/mach-jz4740/base.h header. Signed-off-by:
-
Paul Burton authored
For interrupts numbered after those of the interrupt controller, define their numbers based upon the number of interrupts provided by the SoC interrupt controller. This is in preparation for supporting newer Ingenic SoCs which provide more interrupts. Signed-off-by:
-
Paul Burton authored
On newer Ingenic SoCs the interrupt controller supports more than 32 interrupts, which it does by duplicating the registers at intervals of 0x20 bytes within its address space. Add support for an arbitrary number of interrupts using multiple generic chips, and provide the number of chips to register from the interrupt controller probe function. Signed-off-by:
-
Paul Burton authored
Avoid the need for the global variable jz_intc_base by introducing a struct ingenic_intc_data and passing it around as the IRQ handler data. Signed-off-by:
-
Paul Burton authored
The debugfs code becomes a nuisance when attempting to avoid globals, since the interrupt controller probe function run too early for it to be safe to create the debugfs files. Drop it. Signed-off-by:
Lars-Peter Clausen <lars@metafoo.de> Signed-off-by:
-
Paul Burton authored
When probing the interrupt controller, register an IRQ domain such that the interrupts can be translated by devicetree code & thus used from devicetree. Signed-off-by:
-
Paul Burton authored
Rather than hardcoding the IRQ number used to cascade interrupts from the SoC interrupt controller to the CPU interrupt controller, read that IRQ number from the DT describing the system. Signed-off-by:
-
Paul Burton authored
Declare the JZ4740 interrupt controller for probe via DT using the standard irqchip_init function, and make use of that function to probe the controller by adding the appropriate node to the JZ4740 dtsi. Signed-off-by:
-
Paul Burton authored
Add binding documentation for Ingenic SoC interrupt controllers. Signed-off-by:
Rob Herring <robh@kernel.org> Cc: Ian Campbell <ijc+devicetree@hellion.org.uk> Cc: Jason Cooper <jason@lakedaemon.net> Cc: Kumar Gala <galak@codeaurora.org> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: devicetree@vger.kernel.org Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/10134/Signed-off-by:
-
Paul Burton authored
In preparation for moving the JZ4740 interrupt controller driver to drivers/irqchip, move arch_init_irq into setup.c such that everything remaining in irq.c is related to said JZ4740 interrupt controller. Signed-off-by:
-
Paul Burton authored
Make use of the generic plat_irq_dispatch function introduced by commit 85f7cdac "MIPS: Provide a generic plat_irq_dispatch", in order to reduce unnecessary code duplication. Signed-off-by:
-
Paul Burton authored
Use the generic irqchip_init function to probe irqchip drivers using DT, and add the appropriate node to the JZ4740 devicetree in place of the call to mips_cpu_irq_init. Signed-off-by:
-
Paul Burton authored
Allow the MIPS CPU interrupt controller to be probed from DT using the generic __irqchip_of_table for platforms which use irqchip_init. This will avoid such platforms needing to duplicate the compatible string & init function pointer. [ralf@linux-mips.org: Resolved conflict due the preceeding commit that moves irq-cpu.c. Signed-off-by:
-
Ralf Baechle authored
While at it, rename it because in drivers/irqchip no longer every CPU is a MIPS. Signed-off-by: -
Paul Burton authored
Require a DT for JZ4740 based systems, and add a stub one for the qi_lb60 (Ben NanoNote) board. Devices will be migrated to being probed via this DT over time. Signed-off-by:
-
Paul Burton authored
Ingenic have actually varied the vendor/company ID of the XBurst cores across their range of SoCs, whilst keeping the product ID & revision constant... Add definitions for vendor IDs known to be used in some of Ingenic's newer SoCs, and handle them in the same way as the existing Ingenic vendor ID from the JZ4740. Signed-off-by:
-
Paul Burton authored
In preparation for supporting Ingenic SoCs other than the JZ4740, introduce MACH_INGENIC to Kconfig & move MACH_JZ4740 to a separate entry selected by the board when appropriate. This allows MACH_INGENIC to be used to enable things generic across Ingenic SoCs. Signed-off-by:
-
Paul Burton authored
Define a vendor prefix for Qi Hardware, creators of the Ben Nanonote (qi_lb60) among other open devices. Signed-off-by:
-
Paul Burton authored
Define a vendor prefix for Ingenic Semiconductor, a vendor of MIPS-based SoCs. Simply use 'ingenic'. Signed-off-by:
-
Maciej W. Rozycki authored
Update CPU overrides for the DEC port with the recent additions, shaving off some effectively dead code: text data bss dec hex filename 5586952 233132 5990368 11810452 b43694 vmlinux.32-old 5581248 233140 5990368 11804756 b42054 vmlinux.32-new text data bss dec hex filename 6036936 356648 10756544 17150128 105b0b0 vmlinux.64-old 6029896 360752 10756544 17147192 105a538 vmlinux.64-new The data size increase is due to the special alignment requirement of `init_thread_union' aka `.data..init_task' moving it up to the nearest page boundary and making the amount of padding at its front rely on how far within a page text ends. Signed-off-by:
Maciej W. Rozycki <macro@linux-mips.org> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/10197/Signed-off-by:
-
Brian Norris authored
The cmdlinepart parser is already supported in the default probe. Signed-off-by:
Brian Norris <computersforpeace@gmail.com> Cc: linux-mtd@lists.infradead.org Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/10071/Signed-off-by:
-
Maciej W. Rozycki authored
Replace an explicit barrier with a useful processor instruction in TLB invalidation, following several other such cases elsewhere in `tlb-r3k.c'. Signed-off-by:
-
Maciej W. Rozycki authored
Move the initialisation of the CP0.Wired register implemented by Toshiba TX3922 and TX3927 processors from `tx39_cache_init' to `tlb_init' where it belongs, correcting code structure and making sure initialisation does not rely on `tx39_cache_init' being called before `tlb_init' to work correctly. Make `r3k_have_wired_reg' static as it's no longer externally referred to; remove a stale declaration too. Signed-off-by:
-
Maciej W. Rozycki authored
Most R3k processor implementations have their 8 first TLB entries fixed as wired, so we always skip them in TLB invalidation. That however means any leftover entries present there at boot will stay throughout the life of the kernel, unless replaced with new ones. So rename `local_flush_tlb_all' to `local_flush_tlb_from' and make it accept the TLB entry to start from. Then use 0 initially at bootstrap, and the first regular entry later on, bypassing any wired entries. Wrap the latter arrangement into a new `local_flush_tlb_all' entry point. There is no need to disable interrupts in the call made from `tlb_init' because it's made before the interrupt subsystem has been initialised; this is also true for secondary processors, should we ever support R3k SMP. So move this piece of code to new `local_flush_tlb_all'. Signed-off-by:
-
James Hogan authored
XPA extends the physical addresses on MIPS32, including the EntryLo registers. Update dump_tlb() to concatenate the PFNX field from the high end of the EntryLo registers (as read by mfhc0). The width of physical and virtual addresses are also separated to show only 8 nibbles of virtual but 11 nibbles of physical with XPA. Signed-off-by:
James Hogan <james.hogan@imgtec.com> Cc: Steven J. Hill <Steven.Hill@imgtec.com> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/10077/Signed-off-by:
-
James Hogan authored
The RI/XI bits when present are above the PFN field in the EntryLo registers, at bits 63,62 when read with dmfc0, and bits 31,30 when read with mfc0. This makes them appear as part of the physical address, since the other bits are masked with PAGE_MASK, for example: Index: 253 pgmask=16kb va=77b18000 asid=75 [pa=1000744000 c=5 d=1 v=1 g=0] [pa=100134c000 c=5 d=1 v=1 g=0] The physical addresses have bit 36 set, which corresponds to bit 30 of EntryLo1, the XI bit. Explicitly mask off the RI and XI bits from the printed physical address, and print the RI and XI bits separately if they exist, giving output more like this: Index: 226 pgmask=16kb va=77be0000 asid=79 [ri=0 xi=1 pa=01288000 c=5 d=1 v=1 g=0] [ri=0 xi=0 pa=010e4000 c=5 d=0 v=1 g=0] Cc: linux-mips@linux-mips.org Cc: James Hogan <james.hogan@imgtec.com> Cc: David Daney <ddaney@caviumnetworks.com> Patchwork: https://patchwork.linux-mips.org/patch/10080/Signed-off-by: -
James Hogan authored
The EHINV bit in EntryHi allows a TLB entry to be properly marked invalid so that EntryHi doesn't have to be set to a unique value to avoid machine check exceptions due to multiple matching entries. Unfortunately dump_tlb() doesn't take this into account so it will print all the uninteresting invalid TLB entries if the current ASID happens to be 00. Therefore add a condition to skip entries which are marked invalid with the EHINV bit. Signed-off-by:
-
James Hogan authored
The TLB only matches the ASID when the global bit isn't set, so dump_tlb() shouldn't really be skipping global entries just because the ASID doesn't match. Fix the condition to read the TLB entry's global bit from EntryLo0. Note that after a TLB read the global bits in both EntryLo registers reflect the same global bit in the TLB entry. Signed-off-by:
-
James Hogan authored
Make use of recently added EntryLo bit definitions in mipsregs.h when dumping TLB contents. Signed-off-by:
-
James Hogan authored
Refactor the TLB matching code in dump_tlb() slightly so that the conditions which can cause a TLB entry to be skipped can be more easily extended. This should prevent the match condition getting unwieldy once it is updated to take further conditions into account. Signed-off-by:
-
James Hogan authored
Use the new tlb read hazard macros from <asm/hazards.h> rather than the local BARRIER() macro which uses 7 ops regardless of the kernel configuration. We use mtc0_tlbr_hazard for the hazard between mtc0 to the index register and the tlbr, and tlb_read_hazard for the hazard between the tlbr and the mfc0 of the TLB registers written by tlbr. Signed-off-by:
-
James Hogan authored
Add definitions for EntryLo register bits in mipsregs.h. The R4000 compatible ones are prefixed MIPS_ENTRYLO_ and the R3000 compatible ones are prefixed R3K_ENTRYLO_. These will be used in later patches. Signed-off-by:
-
