From 24bdb0b62cc82120924762ae6bc85afc8c3f2b26 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Tue, 12 Apr 2011 12:19:52 +0100 Subject: xen: do not create the extra e820 region at an addr lower than 4G Do not add the extra e820 region at a physical address lower than 4G because it breaks e820_end_of_low_ram_pfn(). It is OK for us to move the xen_extra_mem_start up and down because this is the index of the memory that can be ballooned in/out - it is memory not available to the kernel during bootup. Signed-off-by: Stefano Stabellini Signed-off-by: Konrad Rzeszutek Wilk --- arch/x86/xen/setup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/xen') diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c index fa0269a99377..90bac0aac3a5 100644 --- a/arch/x86/xen/setup.c +++ b/arch/x86/xen/setup.c @@ -227,7 +227,7 @@ char * __init xen_memory_setup(void) memcpy(map_raw, map, sizeof(map)); e820.nr_map = 0; - xen_extra_mem_start = mem_end; + xen_extra_mem_start = max((1ULL << 32), mem_end); for (i = 0; i < memmap.nr_entries; i++) { unsigned long long end; -- cgit v1.2.1 From ee176455e28469e2420032aab3db11ac2ae3eaa8 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Tue, 19 Apr 2011 14:47:31 +0100 Subject: xen: mask_rw_pte: do not apply the early_ioremap checks on x86_32 The two "is_early_ioremap_ptep" checks in mask_rw_pte are only used on x86_64, in fact early_ioremap is not used at all to setup the initial pagetable on x86_32. Moreover on x86_32 the two checks are wrong because the range pgt_buf_start..pgt_buf_end initially should be mapped RW because the pages in the range are not pagetable pages yet and haven't been cleared yet. Afterwards considering the pgt_buf_start..pgt_buf_end is part of the initial mapping, xen_alloc_pte is capable of turning the ptes RO when they become pagetable pages. Fix the issue and improve the readability of the code providing two different implementation of mask_rw_pte for x86_32 and x86_64. Signed-off-by: Stefano Stabellini Signed-off-by: Konrad Rzeszutek Wilk --- arch/x86/xen/mmu.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'arch/x86/xen') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index a991b57f91fe..aef7af92b28b 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1473,16 +1473,20 @@ static void xen_pgd_free(struct mm_struct *mm, pgd_t *pgd) #endif } +#ifdef CONFIG_X86_32 static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) { - unsigned long pfn = pte_pfn(pte); - -#ifdef CONFIG_X86_32 /* If there's an existing pte, then don't allow _PAGE_RW to be set */ if (pte_val_ma(*ptep) & _PAGE_PRESENT) pte = __pte_ma(((pte_val_ma(*ptep) & _PAGE_RW) | ~_PAGE_RW) & pte_val_ma(pte)); -#endif + + return pte; +} +#else /* CONFIG_X86_64 */ +static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) +{ + unsigned long pfn = pte_pfn(pte); /* * If the new pfn is within the range of the newly allocated @@ -1497,6 +1501,7 @@ static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) return pte; } +#endif /* CONFIG_X86_64 */ /* Init-time set_pte while constructing initial pagetables, which doesn't allow RO pagetable pages to be remapped RW */ -- cgit v1.2.1 From a38647837a411f7df79623128421eef2118b5884 Mon Sep 17 00:00:00 2001 From: Konrad Rzeszutek Wilk Date: Fri, 29 Apr 2011 11:34:00 -0400 Subject: xen/mmu: Add workaround "x86-64, mm: Put early page table high" As a consequence of the commit: commit 4b239f458c229de044d6905c2b0f9fe16ed9e01e Author: Yinghai Lu Date: Fri Dec 17 16:58:28 2010 -0800 x86-64, mm: Put early page table high it causes the Linux kernel to crash under Xen: mapping kernel into physical memory Xen: setup ISA identity maps about to get started... (XEN) mm.c:2466:d0 Bad type (saw 7400000000000001 != exp 1000000000000000) for mfn b1d89 (pfn bacf7) (XEN) mm.c:3027:d0 Error while pinning mfn b1d89 (XEN) traps.c:481:d0 Unhandled invalid opcode fault/trap [#6] on VCPU 0 [ec=0000] (XEN) domain_crash_sync called from entry.S (XEN) Domain 0 (vcpu#0) crashed on cpu#0: ... The reason is that at some point init_memory_mapping is going to reach the pagetable pages area and map those pages too (mapping them as normal memory that falls in the range of addresses passed to init_memory_mapping as argument). Some of those pages are already pagetable pages (they are in the range pgt_buf_start-pgt_buf_end) therefore they are going to be mapped RO and everything is fine. Some of these pages are not pagetable pages yet (they fall in the range pgt_buf_end-pgt_buf_top; for example the page at pgt_buf_end) so they are going to be mapped RW. When these pages become pagetable pages and are hooked into the pagetable, xen will find that the guest has already a RW mapping of them somewhere and fail the operation. The reason Xen requires pagetables to be RO is that the hypervisor needs to verify that the pagetables are valid before using them. The validation operations are called "pinning" (more details in arch/x86/xen/mmu.c). In order to fix the issue we mark all the pages in the entire range pgt_buf_start-pgt_buf_top as RO, however when the pagetable allocation is completed only the range pgt_buf_start-pgt_buf_end is reserved by init_memory_mapping. Hence the kernel is going to crash as soon as one of the pages in the range pgt_buf_end-pgt_buf_top is reused (b/c those ranges are RO). For this reason, this function is introduced which is called _after_ the init_memory_mapping has completed (in a perfect world we would call this function from init_memory_mapping, but lets ignore that). Because we are called _after_ init_memory_mapping the pgt_buf_[start, end,top] have all changed to new values (b/c another init_memory_mapping is called). Hence, the first time we enter this function, we save away the pgt_buf_start value and update the pgt_buf_[end,top]. When we detect that the "old" pgt_buf_start through pgt_buf_end PFNs have been reserved (so memblock_x86_reserve_range has been called), we immediately set out to RW the "old" pgt_buf_end through pgt_buf_top. And then we update those "old" pgt_buf_[end|top] with the new ones so that we can redo this on the next pagetable. Acked-by: "H. Peter Anvin" Reviewed-by: Jeremy Fitzhardinge [v1: Updated with Jeremy's comments] [v2: Added the crash output] Signed-off-by: Konrad Rzeszutek Wilk --- arch/x86/xen/mmu.c | 123 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 123 insertions(+) (limited to 'arch/x86/xen') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index aef7af92b28b..1bca25f60ff2 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1463,6 +1463,119 @@ static int xen_pgd_alloc(struct mm_struct *mm) return ret; } +#ifdef CONFIG_X86_64 +static __initdata u64 __last_pgt_set_rw = 0; +static __initdata u64 __pgt_buf_start = 0; +static __initdata u64 __pgt_buf_end = 0; +static __initdata u64 __pgt_buf_top = 0; +/* + * As a consequence of the commit: + * + * commit 4b239f458c229de044d6905c2b0f9fe16ed9e01e + * Author: Yinghai Lu + * Date: Fri Dec 17 16:58:28 2010 -0800 + * + * x86-64, mm: Put early page table high + * + * at some point init_memory_mapping is going to reach the pagetable pages + * area and map those pages too (mapping them as normal memory that falls + * in the range of addresses passed to init_memory_mapping as argument). + * Some of those pages are already pagetable pages (they are in the range + * pgt_buf_start-pgt_buf_end) therefore they are going to be mapped RO and + * everything is fine. + * Some of these pages are not pagetable pages yet (they fall in the range + * pgt_buf_end-pgt_buf_top; for example the page at pgt_buf_end) so they + * are going to be mapped RW. When these pages become pagetable pages and + * are hooked into the pagetable, xen will find that the guest has already + * a RW mapping of them somewhere and fail the operation. + * The reason Xen requires pagetables to be RO is that the hypervisor needs + * to verify that the pagetables are valid before using them. The validation + * operations are called "pinning". + * + * In order to fix the issue we mark all the pages in the entire range + * pgt_buf_start-pgt_buf_top as RO, however when the pagetable allocation + * is completed only the range pgt_buf_start-pgt_buf_end is reserved by + * init_memory_mapping. Hence the kernel is going to crash as soon as one + * of the pages in the range pgt_buf_end-pgt_buf_top is reused (b/c those + * ranges are RO). + * + * For this reason, 'mark_rw_past_pgt' is introduced which is called _after_ + * the init_memory_mapping has completed (in a perfect world we would + * call this function from init_memory_mapping, but lets ignore that). + * + * Because we are called _after_ init_memory_mapping the pgt_buf_[start, + * end,top] have all changed to new values (b/c init_memory_mapping + * is called and setting up another new page-table). Hence, the first time + * we enter this function, we save away the pgt_buf_start value and update + * the pgt_buf_[end,top]. + * + * When we detect that the "old" pgt_buf_start through pgt_buf_end + * PFNs have been reserved (so memblock_x86_reserve_range has been called), + * we immediately set out to RW the "old" pgt_buf_end through pgt_buf_top. + * + * And then we update those "old" pgt_buf_[end|top] with the new ones + * so that we can redo this on the next pagetable. + */ +static __init void mark_rw_past_pgt(void) { + + if (pgt_buf_end > pgt_buf_start) { + u64 addr, size; + + /* Save it away. */ + if (!__pgt_buf_start) { + __pgt_buf_start = pgt_buf_start; + __pgt_buf_end = pgt_buf_end; + __pgt_buf_top = pgt_buf_top; + return; + } + /* If we get the range that starts at __pgt_buf_end that means + * the range is reserved, and that in 'init_memory_mapping' + * the 'memblock_x86_reserve_range' has been called with the + * outdated __pgt_buf_start, __pgt_buf_end (the "new" + * pgt_buf_[start|end|top] refer now to a new pagetable. + * Note: we are called _after_ the pgt_buf_[..] have been + * updated.*/ + + addr = memblock_x86_find_in_range_size(PFN_PHYS(__pgt_buf_start), + &size, PAGE_SIZE); + + /* Still not reserved, meaning 'memblock_x86_reserve_range' + * hasn't been called yet. Update the _end and _top.*/ + if (addr == PFN_PHYS(__pgt_buf_start)) { + __pgt_buf_end = pgt_buf_end; + __pgt_buf_top = pgt_buf_top; + return; + } + + /* OK, the area is reserved, meaning it is time for us to + * set RW for the old end->top PFNs. */ + + /* ..unless we had already done this. */ + if (__pgt_buf_end == __last_pgt_set_rw) + return; + + addr = PFN_PHYS(__pgt_buf_end); + + /* set as RW the rest */ + printk(KERN_DEBUG "xen: setting RW the range %llx - %llx\n", + PFN_PHYS(__pgt_buf_end), PFN_PHYS(__pgt_buf_top)); + + while (addr < PFN_PHYS(__pgt_buf_top)) { + make_lowmem_page_readwrite(__va(addr)); + addr += PAGE_SIZE; + } + /* And update everything so that we are ready for the next + * pagetable (the one created for regions past 4GB) */ + __last_pgt_set_rw = __pgt_buf_end; + __pgt_buf_start = pgt_buf_start; + __pgt_buf_end = pgt_buf_end; + __pgt_buf_top = pgt_buf_top; + } + return; +} +#else +static __init void mark_rw_past_pgt(void) { } +#endif static void xen_pgd_free(struct mm_struct *mm, pgd_t *pgd) { #ifdef CONFIG_X86_64 @@ -1488,6 +1601,14 @@ static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) { unsigned long pfn = pte_pfn(pte); + /* + * A bit of optimization. We do not need to call the workaround + * when xen_set_pte_init is called with a PTE with 0 as PFN. + * That is b/c the pagetable at that point are just being populated + * with empty values and we can save some cycles by not calling + * the 'memblock' code.*/ + if (pfn) + mark_rw_past_pgt(); /* * If the new pfn is within the range of the newly allocated * kernel pagetable, and it isn't being mapped into an @@ -1997,6 +2118,8 @@ __init void xen_ident_map_ISA(void) static __init void xen_post_allocator_init(void) { + mark_rw_past_pgt(); + #ifdef CONFIG_XEN_DEBUG pv_mmu_ops.make_pte = PV_CALLEE_SAVE(xen_make_pte_debug); #endif -- cgit v1.2.1 From b9269dc7bfdf8c985971c09f2dcb2aa04ad7986d Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Tue, 12 Apr 2011 12:19:49 +0100 Subject: xen: mask_rw_pte mark RO all pagetable pages up to pgt_buf_top mask_rw_pte is currently checking if a pfn is a pagetable page if it falls in the range pgt_buf_start - pgt_buf_end but that is incorrect because pgt_buf_end is a moving target: pgt_buf_top is the real boundary. Acked-by: "H. Peter Anvin" Signed-off-by: Stefano Stabellini Signed-off-by: Konrad Rzeszutek Wilk --- arch/x86/xen/mmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/xen') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index 1bca25f60ff2..55c965b38c27 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1616,7 +1616,7 @@ static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) * it is RO. */ if (((!is_early_ioremap_ptep(ptep) && - pfn >= pgt_buf_start && pfn < pgt_buf_end)) || + pfn >= pgt_buf_start && pfn < pgt_buf_top)) || (is_early_ioremap_ptep(ptep) && pfn != (pgt_buf_end - 1))) pte = pte_wrprotect(pte); -- cgit v1.2.1