Merge branch 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...

Pull x86 mm changes from Ingo Molnar:
 "The main changes in this cycle were:

   - reduce the x86/32 PAE per task PGD allocation overhead from 4K to
     0.032k (Fenghua Yu)

   - early_ioremap/memunmap() usage cleanups (Juergen Gross)

   - gbpages support cleanups (Luis R Rodriguez)

   - improve AMD Bulldozer (family 0x15) ASLR I$ aliasing workaround to
     increase randomization by 3 bits (per bootup) (Hector
     Marco-Gisbert)

   - misc fixlets"

* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mm: Improve AMD Bulldozer ASLR workaround
  x86/mm/pat: Initialize __cachemode2pte_tbl[] and __pte2cachemode_tbl[] in a bit more readable fashion
  init.h: Clean up the __setup()/early_param() macros
  x86/mm: Simplify probe_page_size_mask()
  x86/mm: Further simplify 1 GB kernel linear mappings handling
  x86/mm: Use early_param_on_off() for direct_gbpages
  init.h: Add early_param_on_off()
  x86/mm: Simplify enabling direct_gbpages
  x86/mm: Use IS_ENABLED() for direct_gbpages
  x86/mm: Unexport set_memory_ro() and set_memory_rw()
  x86/mm, efi: Use early_ioremap() in arch/x86/platform/efi/efi-bgrt.c
  x86/mm: Use early_memunmap() instead of early_iounmap()
  x86/mm/pat: Ensure different messages in STRICT_DEVMEM and PAT cases
  x86/mm: Reduce PAE-mode per task pgd allocation overhead from 4K to 32 bytes

diff --cc arch/x86/Kconfig
Simple merge

diff --cc arch/x86/include/asm/elf.h
Simple merge

diff --cc arch/x86/kernel/cpu/amd.c
Simple merge

diff --cc arch/x86/kernel/setup.c
Simple merge

diff --cc arch/x86/mm/init.c
index 52417e771af9f66029f42812d02cbce58953b76a,4469563f8c3b3ece258e2f5ad59d80c76bd1573f..1d553186c4345c02be5c0152764d988cfe365ae1
--- 1/arch/x86/mm/init.c
--- 2/arch/x86/mm/init.c
+++ b/arch/x86/mm/init.c
@@@ -179,8 -165,15 +165,16 @@@ static void __init probe_page_size_mask
        if (cpu_has_pge) {
                cr4_set_bits_and_update_boot(X86_CR4_PGE);
                __supported_pte_mask |= _PAGE_GLOBAL;
 -      }
 +      } else
 +              __supported_pte_mask &= ~_PAGE_GLOBAL;
+ 
+       /* Enable 1 GB linear kernel mappings if available: */
+       if (direct_gbpages && cpu_has_gbpages) {
+               printk(KERN_INFO "Using GB pages for direct mapping\n");
+               page_size_mask |= 1 << PG_LEVEL_1G;
+       } else {
+               direct_gbpages = 0;
+       }
  }
  
  #ifdef CONFIG_X86_32