diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2015-11-06 10:38:43 +0300 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2015-11-06 10:38:43 +0300 |
commit | 2f4bf528eca5b2d9eef12b6d323c040254f8f67c (patch) | |
tree | 3642d6c1623be0ad1cc991343eb4e0845d77de7a /arch/powerpc/mm/tlb_nohash.c | |
parent | 2e3078af2c67730c479f1d183af5b367f5d95337 (diff) | |
parent | 8bdf2023e238ce2262d0cf1aca78785dc46e15db (diff) | |
download | linux-2f4bf528eca5b2d9eef12b6d323c040254f8f67c.tar.xz |
Merge tag 'powerpc-4.4-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc updates from Michael Ellerman:
- Kconfig: remove BE-only platforms from LE kernel build from Boqun
Feng
- Refresh ps3_defconfig from Geoff Levand
- Emit GNU & SysV hashes for the vdso from Michael Ellerman
- Define an enum for the bolted SLB indexes from Anshuman Khandual
- Use a local to avoid multiple calls to get_slb_shadow() from Michael
Ellerman
- Add gettimeofday() benchmark from Michael Neuling
- Avoid link stack corruption in __get_datapage() from Michael Neuling
- Add virt_to_pfn and use this instead of opencoding from Aneesh Kumar
K.V
- Add ppc64le_defconfig from Michael Ellerman
- pseries: extract of_helpers module from Andy Shevchenko
- Correct string length in pseries_of_derive_parent() from Nathan
Fontenot
- Free the MSI bitmap if it was slab allocated from Denis Kirjanov
- Shorten irq_chip name for the SIU from Christophe Leroy
- Wait 1s for secondaries to enter OPAL during kexec from Samuel
Mendoza-Jonas
- Fix _ALIGN_* errors due to type difference, from Aneesh Kumar K.V
- powerpc/pseries/hvcserver: don't memset pi_buff if it is null from
Colin Ian King
- Disable hugepd for 64K page size, from Aneesh Kumar K.V
- Differentiate between hugetlb and THP during page walk from Aneesh
Kumar K.V
- Make PCI non-optional for pseries from Michael Ellerman
- Individual System V IPC system calls from Sam bobroff
- Add selftest of unmuxed IPC calls from Michael Ellerman
- discard .exit.data at runtime from Stephen Rothwell
- Delete old orphaned PrPMC 280/2800 DTS and boot file, from Paul
Gortmaker
- Use of_get_next_parent to simplify code from Christophe Jaillet
- Paginate some xmon output from Sam bobroff
- Add some more elements to the xmon PACA dump from Michael Ellerman
- Allow the tm-syscall selftest to build with old headers from Michael
Ellerman
- Run EBB selftests only on POWER8 from Denis Kirjanov
- Drop CONFIG_TUNE_CELL in favour of CONFIG_CELL_CPU from Michael
Ellerman
- Avoid reference to potentially freed memory in prom.c from Christophe
Jaillet
- Quieten boot wrapper output with run_cmd from Geoff Levand
- EEH fixes and cleanups from Gavin Shan
- Fix recursive fenced PHB on Broadcom shiner adapter from Gavin Shan
- Use of_get_next_parent() in of_get_ibm_chip_id() from Michael
Ellerman
- Fix section mismatch warning in msi_bitmap_alloc() from Denis
Kirjanov
- Fix ps3-lpm white space from Rudhresh Kumar J
- Fix ps3-vuart null dereference from Colin King
- nvram: Add missing kfree in error path from Christophe Jaillet
- nvram: Fix function name in some errors messages, from Christophe
Jaillet
- drivers/macintosh: adb: fix misleading Kconfig help text from Aaro
Koskinen
- agp/uninorth: fix a memleak in create_gatt_table from Denis Kirjanov
- cxl: Free virtual PHB when removing from Andrew Donnellan
- scripts/kconfig/Makefile: Allow KBUILD_DEFCONFIG to be a target from
Michael Ellerman
- scripts/kconfig/Makefile: Fix KBUILD_DEFCONFIG check when building
with O= from Michael Ellerman
- Freescale updates from Scott: Highlights include 64-bit book3e
kexec/kdump support, a rework of the qoriq clock driver, device tree
changes including qoriq fman nodes, support for a new 85xx board, and
some fixes.
- MPC5xxx updates from Anatolij: Highlights include a driver for
MPC512x LocalPlus Bus FIFO with its device tree binding
documentation, mpc512x device tree updates and some minor fixes.
* tag 'powerpc-4.4-1' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux: (106 commits)
powerpc/msi: Fix section mismatch warning in msi_bitmap_alloc()
powerpc/prom: Use of_get_next_parent() in of_get_ibm_chip_id()
powerpc/pseries: Correct string length in pseries_of_derive_parent()
powerpc/e6500: hw tablewalk: make sure we invalidate and write to the same tlb entry
powerpc/mpc85xx: Add FSL QorIQ DPAA FMan support to the SoC device tree(s)
powerpc/mpc85xx: Create dts components for the FSL QorIQ DPAA FMan
powerpc/fsl: Add #clock-cells and clockgen label to clockgen nodes
powerpc: handle error case in cpm_muram_alloc()
powerpc: mpic: use IRQCHIP_SKIP_SET_WAKE instead of redundant mpic_irq_set_wake
powerpc/book3e-64: Enable kexec
powerpc/book3e-64/kexec: Set "r4 = 0" when entering spinloop
powerpc/booke: Only use VIRT_PHYS_OFFSET on booke32
powerpc/book3e-64/kexec: Enable SMP release
powerpc/book3e-64/kexec: create an identity TLB mapping
powerpc/book3e-64: Don't limit paca to 256 MiB
powerpc/book3e/kdump: Enable crash_kexec_wait_realmode
powerpc/book3e: support CONFIG_RELOCATABLE
powerpc/booke64: Fix args to copy_and_flush
powerpc/book3e-64: rename interrupt_end_book3e with __end_interrupts
powerpc/e6500: kexec: Handle hardware threads
...
Diffstat (limited to 'arch/powerpc/mm/tlb_nohash.c')
-rw-r--r-- | arch/powerpc/mm/tlb_nohash.c | 41 |
1 files changed, 34 insertions, 7 deletions
diff --git a/arch/powerpc/mm/tlb_nohash.c b/arch/powerpc/mm/tlb_nohash.c index 723a099f6be3..bb04e4df3100 100644 --- a/arch/powerpc/mm/tlb_nohash.c +++ b/arch/powerpc/mm/tlb_nohash.c @@ -42,6 +42,7 @@ #include <asm/tlbflush.h> #include <asm/tlb.h> #include <asm/code-patching.h> +#include <asm/cputhreads.h> #include <asm/hugetlb.h> #include <asm/paca.h> @@ -628,10 +629,26 @@ static void early_init_this_mmu(void) #ifdef CONFIG_PPC_FSL_BOOK3E if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { unsigned int num_cams; + int __maybe_unused cpu = smp_processor_id(); + bool map = true; /* use a quarter of the TLBCAM for bolted linear map */ num_cams = (mfspr(SPRN_TLB1CFG) & TLBnCFG_N_ENTRY) / 4; - linear_map_top = map_mem_in_cams(linear_map_top, num_cams); + + /* + * Only do the mapping once per core, or else the + * transient mapping would cause problems. + */ +#ifdef CONFIG_SMP + if (cpu != boot_cpuid && + (cpu != cpu_first_thread_sibling(cpu) || + cpu == cpu_first_thread_sibling(boot_cpuid))) + map = false; +#endif + + if (map) + linear_map_top = map_mem_in_cams(linear_map_top, + num_cams, false); } #endif @@ -729,10 +746,14 @@ void setup_initial_memory_limit(phys_addr_t first_memblock_base, * entries are supported though that may eventually * change. * - * on FSL Embedded 64-bit, we adjust the RMA size to match the - * first bolted TLB entry size. We still limit max to 1G even if - * the TLB could cover more. This is due to what the early init - * code is setup to do. + * on FSL Embedded 64-bit, usually all RAM is bolted, but with + * unusual memory sizes it's possible for some RAM to not be mapped + * (such RAM is not used at all by Linux, since we don't support + * highmem on 64-bit). We limit ppc64_rma_size to what would be + * mappable if this memblock is the only one. Additional memblocks + * can only increase, not decrease, the amount that ends up getting + * mapped. We still limit max to 1G even if we'll eventually map + * more. This is due to what the early init code is set up to do. * * We crop it to the size of the first MEMBLOCK to * avoid going over total available memory just in case... @@ -740,8 +761,14 @@ void setup_initial_memory_limit(phys_addr_t first_memblock_base, #ifdef CONFIG_PPC_FSL_BOOK3E if (mmu_has_feature(MMU_FTR_TYPE_FSL_E)) { unsigned long linear_sz; - linear_sz = calc_cam_sz(first_memblock_size, PAGE_OFFSET, - first_memblock_base); + unsigned int num_cams; + + /* use a quarter of the TLBCAM for bolted linear map */ + num_cams = (mfspr(SPRN_TLB1CFG) & TLBnCFG_N_ENTRY) / 4; + + linear_sz = map_mem_in_cams(first_memblock_size, num_cams, + true); + ppc64_rma_size = min_t(u64, linear_sz, 0x40000000); } else #endif |