[git pull] x86 updates

From: Ingo Molnar
Date: Fri Feb 29 2008 - 13:08:50 EST



Linus, please pull the latest x86 fixes git tree from:

git://git.kernel.org/pub/scm/linux/kernel/git/x86/linux-2.6-x86.git

[ Sidenote: i believe "x86: fix pmd_bad and pud_bad to support huge
pages" might be solvable in mm/memory.c as well - but this is the
minimal (and tested) fix that solves it within the scope of arch/x86. ]

Thanks,

Ingo

------------------>
Dave Anderson (1):
x86 ptrace: fix ptrace_bts_config structure declaration

Hans Rosenfeld (1):
x86: fix pmd_bad and pud_bad to support huge pages

Ingo Molnar (3):
x86: fix leak un ioremap_page_range() failure
x86: delay the export removal of init_mm
x86: disable BTS ptrace extensions for now

Jeremy Fitzhardinge (1):
xen: mask out SEP from CPUID

Roland McGrath (2):
x86: tls prevent_tail_call
x86 vdso: fix build locale dependency

Thomas Gleixner (2):
x86: restore vsyscall64 prochandler
x86: CPA: avoid split of alias mappings

Documentation/feature-removal-schedule.txt | 12 ++++++++++++
arch/x86/kernel/init_task.c | 1 +
arch/x86/kernel/process_32.c | 2 ++
arch/x86/kernel/process_64.c | 2 ++
arch/x86/kernel/ptrace.c | 12 ++++++++++++
arch/x86/kernel/tls.c | 8 ++++++--
arch/x86/kernel/vsyscall_64.c | 11 ++++++++++-
arch/x86/mm/ioremap.c | 2 +-
arch/x86/mm/pageattr.c | 21 +++++++++++----------
arch/x86/vdso/Makefile | 6 ++++--
arch/x86/xen/enlighten.c | 1 +
include/asm-x86/pgtable_32.h | 4 +++-
include/asm-x86/pgtable_64.h | 6 ++++--
include/asm-x86/ptrace-abi.h | 8 ++++----
14 files changed, 73 insertions(+), 23 deletions(-)

diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt
index ba899ff..c1d1fd0 100644
--- a/Documentation/feature-removal-schedule.txt
+++ b/Documentation/feature-removal-schedule.txt
@@ -316,3 +316,15 @@ Why: Largely unmaintained and almost entirely unused. File system
is largely pointless as without a lot of work only the most
trivial of Solaris binaries can work with the emulation code.
Who: David S. Miller <davem@xxxxxxxxxxxxx>
+
+---------------------------
+
+What: init_mm export
+When: 2.6.26
+Why: Not used in-tree. The current out-of-tree users used it to
+ work around problems in the CPA code which should be resolved
+ by now. One usecase was described to provide verification code
+ of the CPA operation. That's a good idea in general, but such
+ code / infrastructure should be in the kernel and not in some
+ out-of-tree driver.
+Who: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
diff --git a/arch/x86/kernel/init_task.c b/arch/x86/kernel/init_task.c
index 5b3ce79..3d01e47 100644
--- a/arch/x86/kernel/init_task.c
+++ b/arch/x86/kernel/init_task.c
@@ -15,6 +15,7 @@ static struct files_struct init_files = INIT_FILES;
static struct signal_struct init_signals = INIT_SIGNALS(init_signals);
static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand);
struct mm_struct init_mm = INIT_MM(init_mm);
+EXPORT_UNUSED_SYMBOL(init_mm); /* will be removed in 2.6.26 */

/*
* Initial thread structure.
diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c
index a7d50a5..be3c7a2 100644
--- a/arch/x86/kernel/process_32.c
+++ b/arch/x86/kernel/process_32.c
@@ -603,11 +603,13 @@ __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
}
#endif

+#ifdef X86_BTS
if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS))
ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS);

if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS))
ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES);
+#endif


if (!test_tsk_thread_flag(next_p, TIF_IO_BITMAP)) {
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index 43f2877..3baf9b9 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -604,11 +604,13 @@ static inline void __switch_to_xtra(struct task_struct *prev_p,
memset(tss->io_bitmap, 0xff, prev->io_bitmap_max);
}

+#ifdef X86_BTS
if (test_tsk_thread_flag(prev_p, TIF_BTS_TRACE_TS))
ptrace_bts_take_timestamp(prev_p, BTS_TASK_DEPARTS);

if (test_tsk_thread_flag(next_p, TIF_BTS_TRACE_TS))
ptrace_bts_take_timestamp(next_p, BTS_TASK_ARRIVES);
+#endif
}

/*
diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c
index d862e39..f41fdc9 100644
--- a/arch/x86/kernel/ptrace.c
+++ b/arch/x86/kernel/ptrace.c
@@ -544,6 +544,8 @@ static int ptrace_set_debugreg(struct task_struct *child,
return 0;
}

+#ifdef X86_BTS
+
static int ptrace_bts_get_size(struct task_struct *child)
{
if (!child->thread.ds_area_msr)
@@ -826,6 +828,7 @@ void ptrace_bts_take_timestamp(struct task_struct *tsk,

ptrace_bts_write_record(tsk, &rec);
}
+#endif /* X86_BTS */

/*
* Called by kernel/ptrace.c when detaching..
@@ -839,7 +842,9 @@ void ptrace_disable(struct task_struct *child)
clear_tsk_thread_flag(child, TIF_SYSCALL_EMU);
#endif
if (child->thread.ds_area_msr) {
+#ifdef X86_BTS
ptrace_bts_realloc(child, 0, 0);
+#endif
child->thread.debugctlmsr &= ~ds_debugctl_mask();
if (!child->thread.debugctlmsr)
clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR);
@@ -961,6 +966,10 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data)
break;
#endif

+ /*
+ * These bits need more cooking - not enabled yet:
+ */
+#ifdef X86_BTS
case PTRACE_BTS_CONFIG:
ret = ptrace_bts_config
(child, data, (struct ptrace_bts_config __user *)addr);
@@ -988,6 +997,7 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data)
ret = ptrace_bts_drain
(child, data, (struct bts_struct __user *) addr);
break;
+#endif

default:
ret = ptrace_request(child, request, addr, data);
@@ -1226,12 +1236,14 @@ asmlinkage long sys32_ptrace(long request, u32 pid, u32 addr, u32 data)
case PTRACE_SETOPTIONS:
case PTRACE_SET_THREAD_AREA:
case PTRACE_GET_THREAD_AREA:
+#ifdef X86_BTS
case PTRACE_BTS_CONFIG:
case PTRACE_BTS_STATUS:
case PTRACE_BTS_SIZE:
case PTRACE_BTS_GET:
case PTRACE_BTS_CLEAR:
case PTRACE_BTS_DRAIN:
+#endif
return sys_ptrace(request, pid, addr, data);

default:
diff --git a/arch/x86/kernel/tls.c b/arch/x86/kernel/tls.c
index 6dfd4e7..022bcaa 100644
--- a/arch/x86/kernel/tls.c
+++ b/arch/x86/kernel/tls.c
@@ -91,7 +91,9 @@ int do_set_thread_area(struct task_struct *p, int idx,

asmlinkage int sys_set_thread_area(struct user_desc __user *u_info)
{
- return do_set_thread_area(current, -1, u_info, 1);
+ int ret = do_set_thread_area(current, -1, u_info, 1);
+ prevent_tail_call(ret);
+ return ret;
}


@@ -139,7 +141,9 @@ int do_get_thread_area(struct task_struct *p, int idx,

asmlinkage int sys_get_thread_area(struct user_desc __user *u_info)
{
- return do_get_thread_area(current, -1, u_info);
+ int ret = do_get_thread_area(current, -1, u_info);
+ prevent_tail_call(ret);
+ return ret;
}

int regset_tls_active(struct task_struct *target,
diff --git a/arch/x86/kernel/vsyscall_64.c b/arch/x86/kernel/vsyscall_64.c
index b6be812..edff4c9 100644
--- a/arch/x86/kernel/vsyscall_64.c
+++ b/arch/x86/kernel/vsyscall_64.c
@@ -222,10 +222,19 @@ long __vsyscall(3) venosys_1(void)
}

#ifdef CONFIG_SYSCTL
+
+static int
+vsyscall_sysctl_change(ctl_table *ctl, int write, struct file * filp,
+ void __user *buffer, size_t *lenp, loff_t *ppos)
+{
+ return proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
+}
+
static ctl_table kernel_table2[] = {
{ .procname = "vsyscall64",
.data = &vsyscall_gtod_data.sysctl_enabled, .maxlen = sizeof(int),
- .mode = 0644 },
+ .mode = 0644,
+ .proc_handler = vsyscall_sysctl_change },
{}
};

diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c
index 882328e..ac3c959 100644
--- a/arch/x86/mm/ioremap.c
+++ b/arch/x86/mm/ioremap.c
@@ -162,7 +162,7 @@ static void __iomem *__ioremap(unsigned long phys_addr, unsigned long size,
area->phys_addr = phys_addr;
vaddr = (unsigned long) area->addr;
if (ioremap_page_range(vaddr, vaddr + size, phys_addr, prot)) {
- remove_vm_area((void *)(vaddr & PAGE_MASK));
+ free_vm_area(area);
return NULL;
}

diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c
index 14e48b5..7049294 100644
--- a/arch/x86/mm/pageattr.c
+++ b/arch/x86/mm/pageattr.c
@@ -26,6 +26,7 @@ struct cpa_data {
pgprot_t mask_set;
pgprot_t mask_clr;
int numpages;
+ int processed;
int flushtlb;
unsigned long pfn;
};
@@ -290,8 +291,8 @@ try_preserve_large_page(pte_t *kpte, unsigned long address,
*/
nextpage_addr = (address + psize) & pmask;
numpages = (nextpage_addr - address) >> PAGE_SHIFT;
- if (numpages < cpa->numpages)
- cpa->numpages = numpages;
+ if (numpages < cpa->processed)
+ cpa->processed = numpages;

/*
* We are safe now. Check whether the new pgprot is the same:
@@ -318,7 +319,7 @@ try_preserve_large_page(pte_t *kpte, unsigned long address,
*/
addr = address + PAGE_SIZE;
pfn++;
- for (i = 1; i < cpa->numpages; i++, addr += PAGE_SIZE, pfn++) {
+ for (i = 1; i < cpa->processed; i++, addr += PAGE_SIZE, pfn++) {
pgprot_t chk_prot = static_protections(new_prot, addr, pfn);

if (pgprot_val(chk_prot) != pgprot_val(new_prot))
@@ -342,7 +343,7 @@ try_preserve_large_page(pte_t *kpte, unsigned long address,
* that we limited the number of possible pages already to
* the number of pages in the large page.
*/
- if (address == (nextpage_addr - psize) && cpa->numpages == numpages) {
+ if (address == (nextpage_addr - psize) && cpa->processed == numpages) {
/*
* The address is aligned and the number of pages
* covers the full page.
@@ -572,7 +573,7 @@ repeat:
set_pte_atomic(kpte, new_pte);
cpa->flushtlb = 1;
}
- cpa->numpages = 1;
+ cpa->processed = 1;
return 0;
}

@@ -583,7 +584,7 @@ repeat:
do_split = try_preserve_large_page(kpte, address, cpa);
/*
* When the range fits into the existing large page,
- * return. cp->numpages and cpa->tlbflush have been updated in
+ * return. cp->processed and cpa->tlbflush have been updated in
* try_large_page:
*/
if (do_split <= 0)
@@ -662,7 +663,7 @@ static int __change_page_attr_set_clr(struct cpa_data *cpa, int checkalias)
* Store the remaining nr of pages for the large page
* preservation check.
*/
- cpa->numpages = numpages;
+ cpa->numpages = cpa->processed = numpages;

ret = __change_page_attr(cpa, checkalias);
if (ret)
@@ -679,9 +680,9 @@ static int __change_page_attr_set_clr(struct cpa_data *cpa, int checkalias)
* CPA operation. Either a large page has been
* preserved or a single page update happened.
*/
- BUG_ON(cpa->numpages > numpages);
- numpages -= cpa->numpages;
- cpa->vaddr += cpa->numpages * PAGE_SIZE;
+ BUG_ON(cpa->processed > numpages);
+ numpages -= cpa->processed;
+ cpa->vaddr += cpa->processed * PAGE_SIZE;
}
return 0;
}
diff --git a/arch/x86/vdso/Makefile b/arch/x86/vdso/Makefile
index b8bd0c4..0a8f474 100644
--- a/arch/x86/vdso/Makefile
+++ b/arch/x86/vdso/Makefile
@@ -48,9 +48,11 @@ obj-$(VDSO64-y) += vdso-syms.lds
# Match symbols in the DSO that look like VDSO*; produce a file of constants.
#
sed-vdsosym := -e 's/^00*/0/' \
- -e 's/^\([[:xdigit:]]*\) . \(VDSO[[:alnum:]_]*\)$$/\2 = 0x\1;/p'
+ -e 's/^\([0-9a-fA-F]*\) . \(VDSO[a-zA-Z0-9_]*\)$$/\2 = 0x\1;/p'
quiet_cmd_vdsosym = VDSOSYM $@
- cmd_vdsosym = $(NM) $< | sed -n $(sed-vdsosym) | LC_ALL=C sort > $@
+define cmd_vdsosym
+ $(NM) $< | LC_ALL=C sed -n $(sed-vdsosym) | LC_ALL=C sort > $@
+endef

$(obj)/%-syms.lds: $(obj)/%.so.dbg FORCE
$(call if_changed,vdsosym)
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 49e5358..8b9ee27 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -153,6 +153,7 @@ static void xen_cpuid(unsigned int *ax, unsigned int *bx,
if (*ax == 1)
maskedx = ~((1 << X86_FEATURE_APIC) | /* disable APIC */
(1 << X86_FEATURE_ACPI) | /* disable ACPI */
+ (1 << X86_FEATURE_SEP) | /* disable SEP */
(1 << X86_FEATURE_ACC)); /* thermal monitoring */

asm(XEN_EMULATE_PREFIX "cpuid"
diff --git a/include/asm-x86/pgtable_32.h b/include/asm-x86/pgtable_32.h
index a842c72..b478efa 100644
--- a/include/asm-x86/pgtable_32.h
+++ b/include/asm-x86/pgtable_32.h
@@ -91,7 +91,9 @@ extern unsigned long pg0[];
/* To avoid harmful races, pmd_none(x) should check only the lower when PAE */
#define pmd_none(x) (!(unsigned long)pmd_val(x))
#define pmd_present(x) (pmd_val(x) & _PAGE_PRESENT)
-#define pmd_bad(x) ((pmd_val(x) & (~PAGE_MASK & ~_PAGE_USER)) != _KERNPG_TABLE)
+#define pmd_bad(x) ((pmd_val(x) \
+ & ~(PAGE_MASK | _PAGE_USER | _PAGE_PSE | _PAGE_NX)) \
+ != _KERNPG_TABLE)


#define pages_to_mb(x) ((x) >> (20-PAGE_SHIFT))
diff --git a/include/asm-x86/pgtable_64.h b/include/asm-x86/pgtable_64.h
index 0a0b77b..0a92583 100644
--- a/include/asm-x86/pgtable_64.h
+++ b/include/asm-x86/pgtable_64.h
@@ -153,12 +153,14 @@ static inline unsigned long pgd_bad(pgd_t pgd)

static inline unsigned long pud_bad(pud_t pud)
{
- return pud_val(pud) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER);
+ return pud_val(pud) &
+ ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER | _PAGE_PSE | _PAGE_NX);
}

static inline unsigned long pmd_bad(pmd_t pmd)
{
- return pmd_val(pmd) & ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER);
+ return pmd_val(pmd) &
+ ~(PTE_MASK | _KERNPG_TABLE | _PAGE_USER | _PAGE_PSE | _PAGE_NX);
}

#define pte_none(x) (!pte_val(x))
diff --git a/include/asm-x86/ptrace-abi.h b/include/asm-x86/ptrace-abi.h
index 81a8ee4..f224eb3 100644
--- a/include/asm-x86/ptrace-abi.h
+++ b/include/asm-x86/ptrace-abi.h
@@ -89,13 +89,13 @@
*/
struct ptrace_bts_config {
/* requested or actual size of BTS buffer in bytes */
- u32 size;
+ __u32 size;
/* bitmask of below flags */
- u32 flags;
+ __u32 flags;
/* buffer overflow signal */
- u32 signal;
+ __u32 signal;
/* actual size of bts_struct in bytes */
- u32 bts_size;
+ __u32 bts_size;
};
#endif

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/