From bb8c29caffdda55d3df0315ff2485f7df7290b1d Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 7 Jul 2009 16:37:05 +0200 Subject: [S390] udelay: disable lockdep to avoid false positives Our udelay implementation enables interrupts to receive a special timer interrupt regardless of the context it is called from. This might lead to false positive lockdep reports. Since lockdep isn't aware of the fact that only a single interrupt source is enabled it warns about possible deadlocks that in reality won't happen, like the one below. To fix this disable lockdep before enabling interrupts. [ 254.040888] ================================= [ 254.040904] [ INFO: inconsistent lock state ] [ 254.040910] 2.6.30 #9 [ 254.040914] --------------------------------- [ 254.040920] inconsistent {IN-HARDIRQ-W} -> {HARDIRQ-ON-W} usage. [ 254.040927] swapper/0 [HC0[0]:SC1[1]:HE1:SE0] takes: [ 254.040934] (sch->lock){?.-...}, at: [<00000000002e4778>] ccw_device_timeout+0x48/0x2f0 [ 254.040961] {IN-HARDIRQ-W} state was registered at: [ 254.040969] [<0000000000096f74>] __lock_acquire+0x9d4/0x188c [ 254.040985] [<0000000000097f68>] lock_acquire+0x13c/0x16c [ 254.040998] [<00000000004527e0>] _spin_lock+0x74/0xb8 [ 254.041016] [<0000000000457eb2>] do_IRQ+0xde/0x208 [ 254.041031] [<000000000002d190>] io_return+0x0/0x8 [ 254.041049] [<0000000000029faa>] vtime_stop_cpu+0xbe/0x114 [ 254.041066] irq event stamp: 259629 [ 254.041076] hardirqs last enabled at (259628): [<000000000045238e>] _spin_unlock_irq+0x5e/0x9c [ 254.041095] hardirqs last disabled at (259629): [<000000000045292e>] _spin_lock_irq+0x4a/0xc4 [ 254.041126] softirqs last enabled at (259614): [<000000000006500e>] __do_softirq+0x296/0x2b0 [ 254.041137] softirqs last disabled at (259619): [<0000000000024cf6>] do_softirq+0x102/0x108 [ 254.041147] [ 254.041148] other info that might help us debug this: [ 254.041153] 2 locks held by swapper/0: [ 254.041157] #0: (&priv->timer){+.-...}, at: [<000000000006bf9a>] run_timer_softirq+0x19a/0x340 [ 254.041170] #1: (sch->lock){?.-...}, at: [<00000000002e4778>] ccw_device_timeout+0x48/0x2f0 [ 254.041182] [ 254.041310] Call Trace: [ 254.041313] ([<00000000000174fc>] show_trace+0x16c/0x170) [ 254.041321] [<0000000000017578>] show_stack+0x78/0x104 [ 254.041327] [<000000000044d0ca>] dump_stack+0xc6/0xd4 [ 254.041342] [<00000000000949b4>] print_usage_bug+0x1c8/0x1fc [ 254.041353] [<0000000000094e8a>] mark_lock+0x4a2/0x670 [ 254.041364] [<00000000000950e2>] mark_held_locks+0x8a/0xb4 [ 254.041375] [<0000000000095398>] trace_hardirqs_on_caller+0x74/0x1ac [ 254.041388] [<00000000000954fa>] trace_hardirqs_on+0x2a/0x38 [ 254.041402] [<000000000025f1ec>] __udelay_disabled+0xac/0xfc [ 254.041419] [<000000000025f432>] __udelay+0x12a/0x148 [ 254.041433] [<00000000002d64d8>] cio_commit_config+0x170/0x290 [ 254.041451] [<00000000002d6978>] cio_disable_subchannel+0x120/0x1cc [ 254.041468] [<00000000002e32a4>] ccw_device_recog_done+0x54/0x2f4 [ 254.041485] [<00000000002e3638>] ccw_device_sense_id_done+0x50/0x90 [ 254.041508] [<00000000002e615a>] snsid_callback+0xfa/0x3a8 [ 254.041515] [<00000000002dd96c>] ccwreq_stop+0x80/0x90 [ 254.041523] [<00000000002dda8e>] ccw_request_timeout+0xc2/0xd0 [ 254.041530] [<00000000002e2f70>] ccw_device_request_event+0x58/0x90 [ 254.041537] [<00000000002e47ae>] ccw_device_timeout+0x7e/0x2f0 [ 254.041555] [<000000000006c02a>] run_timer_softirq+0x22a/0x340 [ 254.041566] [<0000000000064eb0>] __do_softirq+0x138/0x2b0 [ 254.041578] [<0000000000024cf6>] do_softirq+0x102/0x108 [ 254.041590] [<00000000000647ce>] irq_exit+0xee/0x114 [ 254.041603] [<0000000000457d88>] do_extint+0x130/0x17c [ 254.041617] [<000000000002d41e>] ext_no_vtime+0x1e/0x22 [ 254.041631] [<0000000000029faa>] vtime_stop_cpu+0xbe/0x114 [ 254.041646] ([<0000000000029f58>] vtime_stop_cpu+0x6c/0x114) [ 254.041662] [<000000000001d842>] cpu_idle+0x122/0x1c0 [ 254.041679] [<00000000004482c6>] start_secondary+0xce/0xe0 [ 254.041696] [<0000000000000000>] 0x0 [ 254.041715] [<0000000000000000>] 0x0 [ 254.041745] INFO: lockdep is turned off. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/lib/delay.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/s390') diff --git a/arch/s390/lib/delay.c b/arch/s390/lib/delay.c index 3f5f680726e..97c1eca83cc 100644 --- a/arch/s390/lib/delay.c +++ b/arch/s390/lib/delay.c @@ -36,9 +36,11 @@ static void __udelay_disabled(unsigned long usecs) cr0 = (cr0_saved & 0xffff00e0) | 0x00000800; __ctl_load(cr0 , 0, 0); mask = psw_kernel_bits | PSW_MASK_WAIT | PSW_MASK_EXT; + lockdep_off(); trace_hardirqs_on(); __load_psw_mask(mask); local_irq_disable(); + lockdep_on(); __ctl_load(cr0_saved, 0, 0); local_tick_enable(clock_saved); set_clock_comparator(S390_lowcore.clock_comparator); -- cgit From 81088819d588ba4d2dce3bf083fafc91974b14e0 Mon Sep 17 00:00:00 2001 From: Frank Munzert Date: Tue, 7 Jul 2009 16:37:07 +0200 Subject: [S390] shutdown actions: save/return rc from init function We always returned -EINVAL when setting of a shutdown action failed. This was misleading, if for example the hardware did not support the shutdown action. Now we save each shutdown action's init return code and return it when the action is being set. Signed-off-by: Frank Munzert Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/ipl.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index b8bf4b14006..371a2d88f4a 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -70,6 +70,7 @@ struct shutdown_action { char *name; void (*fn) (struct shutdown_trigger *trigger); int (*init) (void); + int init_rc; }; static char *ipl_type_str(enum ipl_type type) @@ -1486,11 +1487,13 @@ static int set_trigger(const char *buf, struct shutdown_trigger *trigger, int i; for (i = 0; i < SHUTDOWN_ACTIONS_COUNT; i++) { - if (!shutdown_actions_list[i]) - continue; if (sysfs_streq(buf, shutdown_actions_list[i]->name)) { - trigger->action = shutdown_actions_list[i]; - return len; + if (shutdown_actions_list[i]->init_rc) { + return shutdown_actions_list[i]->init_rc; + } else { + trigger->action = shutdown_actions_list[i]; + return len; + } } } return -EINVAL; @@ -1640,8 +1643,8 @@ static void __init shutdown_actions_init(void) for (i = 0; i < SHUTDOWN_ACTIONS_COUNT; i++) { if (!shutdown_actions_list[i]->init) continue; - if (shutdown_actions_list[i]->init()) - shutdown_actions_list[i] = NULL; + shutdown_actions_list[i]->init_rc = + shutdown_actions_list[i]->init(); } } -- cgit From 2651fa2bcb1a7b06793441d50dcbf98efda592c6 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 7 Jul 2009 16:37:08 +0200 Subject: [S390] perf_counter build fix Add PERF_COUNTER_INDEX_OFFSET define to fix this build bug: kernel/perf_counter.c: In function 'perf_counter_index': kernel/perf_counter.c:1889: error: 'PERF_COUNTER_INDEX_OFFSET' undeclared Same fix as for FRV since s390 doesn't support hw counters. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/perf_counter.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/s390') diff --git a/arch/s390/include/asm/perf_counter.h b/arch/s390/include/asm/perf_counter.h index a7205a3828c..7015188c2cc 100644 --- a/arch/s390/include/asm/perf_counter.h +++ b/arch/s390/include/asm/perf_counter.h @@ -6,3 +6,5 @@ static inline void set_perf_counter_pending(void) {} static inline void clear_perf_counter_pending(void) {} + +#define PERF_COUNTER_INDEX_OFFSET 0 -- cgit From 5075baca2e95840d76210c3cfda56134007462c9 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 7 Jul 2009 16:37:09 +0200 Subject: [S390] add __ucmpdi2() helper function Provide __ucmpdi2() helper function on 31 bit so we don't run again and again in compile errors like this one: kernel/built-in.o: In function `T.689': perf_counter.c:(.text+0x56c86): undefined reference to `__ucmpdi2' Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/lib/Makefile | 2 +- arch/s390/lib/ucmpdi2.c | 26 ++++++++++++++++++++++++++ 2 files changed, 27 insertions(+), 1 deletion(-) create mode 100644 arch/s390/lib/ucmpdi2.c (limited to 'arch/s390') diff --git a/arch/s390/lib/Makefile b/arch/s390/lib/Makefile index ab6735df2d2..97975ec7a27 100644 --- a/arch/s390/lib/Makefile +++ b/arch/s390/lib/Makefile @@ -3,6 +3,6 @@ # lib-y += delay.o string.o uaccess_std.o uaccess_pt.o -obj-$(CONFIG_32BIT) += div64.o qrnnd.o +obj-$(CONFIG_32BIT) += div64.o qrnnd.o ucmpdi2.o lib-$(CONFIG_64BIT) += uaccess_mvcos.o lib-$(CONFIG_SMP) += spinlock.o diff --git a/arch/s390/lib/ucmpdi2.c b/arch/s390/lib/ucmpdi2.c new file mode 100644 index 00000000000..3e05ff53258 --- /dev/null +++ b/arch/s390/lib/ucmpdi2.c @@ -0,0 +1,26 @@ +#include + +union ull_union { + unsigned long long ull; + struct { + unsigned int high; + unsigned int low; + } ui; +}; + +int __ucmpdi2(unsigned long long a, unsigned long long b) +{ + union ull_union au = {.ull = a}; + union ull_union bu = {.ull = b}; + + if (au.ui.high < bu.ui.high) + return 0; + else if (au.ui.high > bu.ui.high) + return 2; + if (au.ui.low < bu.ui.low) + return 0; + else if (au.ui.low > bu.ui.low) + return 2; + return 1; +} +EXPORT_SYMBOL(__ucmpdi2); -- cgit From fca3e357d5de31e484c3e0ec7f356e4fc488eb7f Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Tue, 7 Jul 2009 16:37:10 +0200 Subject: [S390] set SCHED_OMIT_FRAME_POINTER for s390 The frame pointer is useless for s390 in the sched.c code. Signed-off-by: Martin Schwidefsky --- arch/s390/Kconfig | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch/s390') diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index e577839f307..4cd12b5202a 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -96,6 +96,10 @@ config S390 select INIT_ALL_POSSIBLE select HAVE_PERF_COUNTERS +config SCHED_OMIT_FRAME_POINTER + bool + default y + source "init/Kconfig" source "kernel/Kconfig.freezer" -- cgit From 25ca1251dc55673da1f3c6ddc1bf93fbb7fc83fc Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 7 Jul 2009 16:37:12 +0200 Subject: [S390] add generic atomic64 support for 31 bit Performance counters need 64 bit atomic operations. To keep the patch small we use the simple generic atomic64_t implementation. The native implementation follows with the next kernel. Fixes this build bug: In file included from kernel/sched.c:42: include/linux/perf_counter.h:427: error: expected specifier-qualifier-list before 'atomic64_t' Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/Kconfig | 1 + arch/s390/include/asm/atomic.h | 7 ++++++- 2 files changed, 7 insertions(+), 1 deletion(-) (limited to 'arch/s390') diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 4cd12b5202a..f46a1b50dde 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -95,6 +95,7 @@ config S390 select HAVE_ARCH_TRACEHOOK select INIT_ALL_POSSIBLE select HAVE_PERF_COUNTERS + select GENERIC_ATOMIC64 if !64BIT config SCHED_OMIT_FRAME_POINTER bool diff --git a/arch/s390/include/asm/atomic.h b/arch/s390/include/asm/atomic.h index fca9dffcc66..c7d0abfb0f0 100644 --- a/arch/s390/include/asm/atomic.h +++ b/arch/s390/include/asm/atomic.h @@ -268,7 +268,12 @@ static __inline__ int atomic64_add_unless(atomic64_t *v, #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) #undef __CSG_LOOP -#endif + +#else /* __s390x__ */ + +#include + +#endif /* __s390x__ */ #define smp_mb__before_atomic_dec() smp_mb() #define smp_mb__after_atomic_dec() smp_mb() -- cgit From 07606309ff5a9136e340b9ca87c831650c1ec47a Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Tue, 7 Jul 2009 16:37:13 +0200 Subject: [S390] define KTIME_SCALAR for 32-bit s390 32-bit s390 has efficient support for 64/32-bit conversions, define KTIME_SCALAR to enable the use of the plain scalar nanosecond based representation of ktime. Signed-off-by: Martin Schwidefsky --- arch/s390/Kconfig | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch/s390') diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index f46a1b50dde..2ae5d72f47e 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -121,6 +121,9 @@ config 32BIT bool default y if !64BIT +config KTIME_SCALAR + def_bool 32BIT + config SMP bool "Symmetric multi-processing support" ---help--- -- cgit From c99e6efe1ba04561e7d93a81f0be07e37427e835 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 10 Jul 2009 14:57:56 +0200 Subject: sched: INIT_PREEMPT_COUNT Pull the initial preempt_count value into a single definition site. Maintainers for: alpha, ia64 and m68k, please have a look, your arch code is funny. The header magic is a bit odd, but similar to the KERNEL_DS one, CPP waits with expanding these macros until the INIT_THREAD_INFO macro itself is expanded, which is in arch/*/kernel/init_task.c where we've already included sched.h so we're good. Cc: tony.luck@intel.com Cc: rth@twiddle.net Cc: geert@linux-m68k.org Signed-off-by: Peter Zijlstra Acked-by: Matt Mackall Signed-off-by: Linus Torvalds --- arch/s390/include/asm/thread_info.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/s390') diff --git a/arch/s390/include/asm/thread_info.h b/arch/s390/include/asm/thread_info.h index 925bcc64903..ba1cab9fc1f 100644 --- a/arch/s390/include/asm/thread_info.h +++ b/arch/s390/include/asm/thread_info.h @@ -61,7 +61,7 @@ struct thread_info { .exec_domain = &default_exec_domain, \ .flags = 0, \ .cpu = 0, \ - .preempt_count = 1, \ + .preempt_count = INIT_PREEMPT_COUNT, \ .restart_block = { \ .fn = do_no_restart_syscall, \ }, \ -- cgit From 405f55712dfe464b3240d7816cc4fe4174831be2 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Sat, 11 Jul 2009 22:08:37 +0400 Subject: headers: smp_lock.h redux * Remove smp_lock.h from files which don't need it (including some headers!) * Add smp_lock.h to files which do need it * Make smp_lock.h include conditional in hardirq.h It's needed only for one kernel_locked() usage which is under CONFIG_PREEMPT This will make hardirq.h inclusion cheaper for every PREEMPT=n config (which includes allmodconfig/allyesconfig, BTW) Signed-off-by: Alexey Dobriyan Signed-off-by: Linus Torvalds --- arch/s390/kernel/dis.c | 1 - arch/s390/kernel/ptrace.c | 1 - arch/s390/mm/fault.c | 1 - 3 files changed, 3 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/kernel/dis.c b/arch/s390/kernel/dis.c index d2f270c995d..db943a7ec51 100644 --- a/arch/s390/kernel/dis.c +++ b/arch/s390/kernel/dis.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 490b39934d6..43acd73105b 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 74eb26bf197..e5e119fe03b 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include -- cgit From 5f954c3426190f7ae432a09abd62164d5d14c709 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 24 Jul 2009 12:39:49 +0200 Subject: [S390] hibernation: fix lowcore handling Our swsusp_arch_suspend() backend implementation disables prefixing by setting the contents of the prefix register to 0. However afterwards common code functions are called which might access percpu data structures. Since the lowcore contains e.g. the percpu base pointer this isn't a good idea. So fix this by copying the hibernating cpu's lowcore to absolute address zero. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/early.c | 4 ++-- arch/s390/power/swsusp_asm64.S | 13 ++++++++----- 2 files changed, 10 insertions(+), 7 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index f9b144049dc..8d15314381e 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -210,7 +210,7 @@ static noinline __init void detect_machine_type(void) machine_flags |= MACHINE_FLAG_VM; } -static void early_pgm_check_handler(void) +static __init void early_pgm_check_handler(void) { unsigned long addr; const struct exception_table_entry *fixup; @@ -222,7 +222,7 @@ static void early_pgm_check_handler(void) S390_lowcore.program_old_psw.addr = fixup->fixup | PSW_ADDR_AMODE; } -void setup_lowcore_early(void) +static noinline __init void setup_lowcore_early(void) { psw_t psw; diff --git a/arch/s390/power/swsusp_asm64.S b/arch/s390/power/swsusp_asm64.S index 76d688da32f..e27bd316489 100644 --- a/arch/s390/power/swsusp_asm64.S +++ b/arch/s390/power/swsusp_asm64.S @@ -40,11 +40,11 @@ swsusp_arch_suspend: /* Store prefix register on stack */ stpx __SF_EMPTY(%r15) - /* Setup base register for lowcore (absolute 0) */ - llgf %r1,__SF_EMPTY(%r15) + /* Save prefix register contents for lowcore */ + llgf %r4,__SF_EMPTY(%r15) /* Get pointer to save area */ - aghi %r1,0x1000 + lghi %r1,0x1000 /* Store registers */ mvc 0x318(4,%r1),__SF_EMPTY(%r15) /* move prefix to lowcore */ @@ -79,8 +79,11 @@ swsusp_arch_suspend: xc __SF_EMPTY(4,%r15),__SF_EMPTY(%r15) spx __SF_EMPTY(%r15) - /* Setup lowcore */ - brasl %r14,setup_lowcore_early + lghi %r2,0 + lghi %r3,2*PAGE_SIZE + lghi %r5,2*PAGE_SIZE +1: mvcle %r2,%r4,0 + jo 1b /* Save image */ brasl %r14,swsusp_save -- cgit From c63b196afcf22405527abe4c2c57926a5bbd6fc9 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 24 Jul 2009 12:39:50 +0200 Subject: [S390] hibernation: fix register corruption on machine checks swsusp_arch_suspend() actually saves all cpu register contents on hibernation. Machine checks must be disabled since swsusp_arch_suspend() stores register contents to their lowcore save areas. That's the same place where register contents on machine checks would be saved. To avoid register corruption disable machine checks. We must also disable machine checks in the new psw mask for program checks, since swsusp_arch_suspend() may generate program checks. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/power/swsusp.c | 36 ++++++++++++++++++++++++------------ arch/s390/power/swsusp_asm64.S | 22 +--------------------- 2 files changed, 25 insertions(+), 33 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/power/swsusp.c b/arch/s390/power/swsusp.c index e6a4fe9f5f2..bd1f5c6b0b8 100644 --- a/arch/s390/power/swsusp.c +++ b/arch/s390/power/swsusp.c @@ -7,24 +7,36 @@ * */ +#include -/* - * save CPU registers before creating a hibernation image and before - * restoring the memory state from it - */ void save_processor_state(void) { - /* implentation contained in the - * swsusp_arch_suspend function + /* swsusp_arch_suspend() actually saves all cpu register contents. + * Machine checks must be disabled since swsusp_arch_suspend() stores + * register contents to their lowcore save areas. That's the same + * place where register contents on machine checks would be saved. + * To avoid register corruption disable machine checks. + * We must also disable machine checks in the new psw mask for + * program checks, since swsusp_arch_suspend() may generate program + * checks. Disabling machine checks for all other new psw masks is + * just paranoia. */ + local_mcck_disable(); + /* Disable lowcore protection */ + __ctl_clear_bit(0,28); + S390_lowcore.external_new_psw.mask &= ~PSW_MASK_MCHECK; + S390_lowcore.svc_new_psw.mask &= ~PSW_MASK_MCHECK; + S390_lowcore.io_new_psw.mask &= ~PSW_MASK_MCHECK; + S390_lowcore.program_new_psw.mask &= ~PSW_MASK_MCHECK; } -/* - * restore the contents of CPU registers - */ void restore_processor_state(void) { - /* implentation contained in the - * swsusp_arch_resume function - */ + S390_lowcore.external_new_psw.mask |= PSW_MASK_MCHECK; + S390_lowcore.svc_new_psw.mask |= PSW_MASK_MCHECK; + S390_lowcore.io_new_psw.mask |= PSW_MASK_MCHECK; + S390_lowcore.program_new_psw.mask |= PSW_MASK_MCHECK; + /* Enable lowcore protection */ + __ctl_set_bit(0,28); + local_mcck_enable(); } diff --git a/arch/s390/power/swsusp_asm64.S b/arch/s390/power/swsusp_asm64.S index e27bd316489..b26df5c5933 100644 --- a/arch/s390/power/swsusp_asm64.S +++ b/arch/s390/power/swsusp_asm64.S @@ -32,11 +32,6 @@ swsusp_arch_suspend: /* Deactivate DAT */ stnsm __SF_EMPTY(%r15),0xfb - /* Switch off lowcore protection */ - stctg %c0,%c0,__SF_EMPTY(%r15) - ni __SF_EMPTY+4(%r15),0xef - lctlg %c0,%c0,__SF_EMPTY(%r15) - /* Store prefix register on stack */ stpx __SF_EMPTY(%r15) @@ -88,11 +83,6 @@ swsusp_arch_suspend: /* Save image */ brasl %r14,swsusp_save - /* Switch on lowcore protection */ - stctg %c0,%c0,__SF_EMPTY(%r15) - oi __SF_EMPTY+4(%r15),0x10 - lctlg %c0,%c0,__SF_EMPTY(%r15) - /* Restore prefix register and return */ lghi %r1,0x1000 spx 0x318(%r1) @@ -120,11 +110,6 @@ swsusp_arch_resume: /* Deactivate DAT */ stnsm __SF_EMPTY(%r15),0xfb - /* Switch off lowcore protection */ - stctg %c0,%c0,__SF_EMPTY(%r15) - ni __SF_EMPTY+4(%r15),0xef - lctlg %c0,%c0,__SF_EMPTY(%r15) - /* Set prefix page to zero */ xc __SF_EMPTY(4,%r15),__SF_EMPTY(%r15) spx __SF_EMPTY(%r15) @@ -178,7 +163,7 @@ swsusp_arch_resume: /* Load old stack */ lg %r15,0x2f8(%r13) - /* Pointer to save arae */ + /* Pointer to save area */ lghi %r13,0x1000 #ifdef CONFIG_SMP @@ -190,11 +175,6 @@ swsusp_arch_resume: /* Restore prefix register */ spx 0x318(%r13) - /* Switch on lowcore protection */ - stctg %c0,%c0,__SF_EMPTY(%r15) - oi __SF_EMPTY+4(%r15),0x10 - lctlg %c0,%c0,__SF_EMPTY(%r15) - /* Activate DAT */ stosm __SF_EMPTY(%r15),0x04 -- cgit From 3a6ba4600d6fb913ddb0dd08843ad75405795883 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 24 Jul 2009 12:39:51 +0200 Subject: [S390] vdso: fix per cpu area allocation vdso per cpu area allocation in smp_prepare_cpus() happens with GFP_KERNEL but irqs disabled. Triggers this one: Badness at kernel/lockdep.c:2280 Modules linked in: CPU: 0 Not tainted 2.6.30 #2 Process swapper (pid: 1, task: 000000003fe88000, ksp: 000000003fe87eb8) Krnl PSW : 0400c00180000000 0000000000083360 (lockdep_trace_alloc+0xec/0xf8) [...] Call Trace: ([<00000000000832b6>] lockdep_trace_alloc+0x42/0xf8) [<00000000000b1880>] __alloc_pages_internal+0x3e8/0x5c4 [<00000000000b1b4a>] __get_free_pages+0x3a/0xb0 [<0000000000026546>] vdso_alloc_per_cpu+0x6a/0x18c [<00000000005eff82>] smp_prepare_cpus+0x322/0x594 [<00000000005e8232>] kernel_init+0x76/0x398 [<000000000001bb1e>] kernel_thread_starter+0x6/0xc [<000000000001bb18>] kernel_thread_starter+0x0/0xc Fix this by moving the allocation out of the irqs disabled section. Reported-by: Christian Borntraeger Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/smp.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 2270730f535..be2cae08340 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -687,13 +687,14 @@ void __init smp_prepare_cpus(unsigned int max_cpus) #ifndef CONFIG_64BIT if (MACHINE_HAS_IEEE) lowcore->extended_save_area_addr = (u32) save_area; -#else - if (vdso_alloc_per_cpu(smp_processor_id(), lowcore)) - BUG(); #endif set_prefix((u32)(unsigned long) lowcore); local_mcck_enable(); local_irq_enable(); +#ifdef CONFIG_64BIT + if (vdso_alloc_per_cpu(smp_processor_id(), &S390_lowcore)) + BUG(); +#endif for_each_possible_cpu(cpu) if (cpu != smp_processor_id()) smp_create_idle(cpu); -- cgit From 1277580fe5dfb5aef84854bdb7983657df00b920 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 24 Jul 2009 12:39:52 +0200 Subject: [S390] vdso: clock_gettime of CLOCK_THREAD_CPUTIME_ID with noexec=on The combination of noexec=on and a clock_gettime call with clock id CLOCK_THREAD_CPUTIME_ID is broken. The vdso code switches to the access register mode to get access to the per-cpu data structure to execute the magic ectg instruction. After the ectg instruction the code always switches back to the primary mode but for noexec=on the correct mode is the secondary mode. The effect of the bug is that the user space program looses the access to all mappings without PROT_EXEC, e.g. the stack. The problem is fixed by restoring the mode that has been active before the switch to the access register mode. Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/vdso64/clock_gettime.S | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/kernel/vdso64/clock_gettime.S b/arch/s390/kernel/vdso64/clock_gettime.S index 79dbfee831e..49106c6e6f8 100644 --- a/arch/s390/kernel/vdso64/clock_gettime.S +++ b/arch/s390/kernel/vdso64/clock_gettime.S @@ -88,10 +88,17 @@ __kernel_clock_gettime: llilh %r4,0x0100 sar %a4,%r4 lghi %r4,0 + epsw %r5,0 sacf 512 /* Magic ectg instruction */ .insn ssf,0xc80100000000,__VDSO_ECTG_BASE(4),__VDSO_ECTG_USER(4),4 - sacf 0 - sar %a4,%r2 + tml %r5,0x4000 + jo 11f + tml %r5,0x8000 + jno 10f + sacf 256 + j 11f +10: sacf 0 +11: sar %a4,%r2 algr %r1,%r0 /* r1 = cputime as TOD value */ mghi %r1,1000 /* convert to nanoseconds */ srlg %r1,%r1,12 /* r1 = cputime in nanosec */ -- cgit From 9e1b32caa525cb236e80e9c671e179bcecccc657 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 22 Jul 2009 15:44:28 +1000 Subject: mm: Pass virtual address to [__]p{te,ud,md}_free_tlb() mm: Pass virtual address to [__]p{te,ud,md}_free_tlb() Upcoming paches to support the new 64-bit "BookE" powerpc architecture will need to have the virtual address corresponding to PTE page when freeing it, due to the way the HW table walker works. Basically, the TLB can be loaded with "large" pages that cover the whole virtual space (well, sort-of, half of it actually) represented by a PTE page, and which contain an "indirect" bit indicating that this TLB entry RPN points to an array of PTEs from which the TLB can then create direct entries. Thus, in order to invalidate those when PTE pages are deleted, we need the virtual address to pass to tlbilx or tlbivax instructions. The old trick of sticking it somewhere in the PTE page struct page sucks too much, the address is almost readily available in all call sites and almost everybody implemets these as macros, so we may as well add the argument everywhere. I added it to the pmd and pud variants for consistency. Signed-off-by: Benjamin Herrenschmidt Acked-by: David Howells [MN10300 & FRV] Acked-by: Nick Piggin Acked-by: Martin Schwidefsky [s390] Signed-off-by: Linus Torvalds --- arch/s390/include/asm/tlb.h | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h index 3d8a96d39d9..81150b05368 100644 --- a/arch/s390/include/asm/tlb.h +++ b/arch/s390/include/asm/tlb.h @@ -96,7 +96,8 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page) * pte_free_tlb frees a pte table and clears the CRSTE for the * page table from the tlb. */ -static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte) +static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte, + unsigned long address) { if (!tlb->fullmm) { tlb->array[tlb->nr_ptes++] = pte; @@ -113,7 +114,8 @@ static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte) * as the pgd. pmd_free_tlb checks the asce_limit against 2GB * to avoid the double free of the pmd in this case. */ -static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd) +static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd, + unsigned long address) { #ifdef __s390x__ if (tlb->mm->context.asce_limit <= (1UL << 31)) @@ -134,7 +136,8 @@ static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd) * as the pgd. pud_free_tlb checks the asce_limit against 4TB * to avoid the double free of the pud in this case. */ -static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud) +static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud, + unsigned long address) { #ifdef __s390x__ if (tlb->mm->context.asce_limit <= (1UL << 42)) -- cgit From d3bc2f91b4761a8d9f96bea167fef2f8c00dea54 Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Thu, 16 Jul 2009 17:17:37 +0200 Subject: KVM: s390: fix wait_queue handling There are two waitqueues in kvm for wait handling: vcpu->wq for virt/kvm/kvm_main.c and vpcu->arch.local_int.wq for the s390 specific wait code. the wait handling in kvm_s390_handle_wait was broken by using different wait_queues for add_wait queue and remove_wait_queue. There are two options to fix the problem: o move all the s390 specific code to vcpu->wq and remove vcpu->arch.local_int.wq o move all the s390 specific code to vcpu->arch.local_int.wq This patch chooses the 2nd variant for two reasons: o s390 does not use kvm_vcpu_block but implements its own enabled wait handling. Having a separate wait_queue make it clear, that our wait mechanism is different o the patch is much smaller Report-by: Julia Lawall Signed-off-by: Christian Borntraeger Signed-off-by: Marcelo Tosatti Signed-off-by: Avi Kivity --- arch/s390/kvm/interrupt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/s390') diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c index f04f5301b1b..4d613415c43 100644 --- a/arch/s390/kvm/interrupt.c +++ b/arch/s390/kvm/interrupt.c @@ -386,7 +386,7 @@ no_timer: } __unset_cpu_idle(vcpu); __set_current_state(TASK_RUNNING); - remove_wait_queue(&vcpu->wq, &wait); + remove_wait_queue(&vcpu->arch.local_int.wq, &wait); spin_unlock_bh(&vcpu->arch.local_int.lock); spin_unlock(&vcpu->arch.local_int.float_int->lock); hrtimer_try_to_cancel(&vcpu->arch.ckc_timer); -- cgit From 677c1dd706d9cc384730cbd52baf821923d8be9b Mon Sep 17 00:00:00 2001 From: Hendrik Brueckner Date: Fri, 7 Aug 2009 10:39:24 +0200 Subject: [S390] kernel: Storing machine flags early in lowcore Currently, the machine_flags are stored late in the startup initialization which results in failing machine type checks (e.g. for MACHINE_IS_VM). To allow these checks, store the machine flags in the lowcore when the machine type has been detected. Moving the machine_flags to the lowcore has been introduced with git commit 25097bf153391f7be4c591d47061b3dc4990dac2 Signed-off-by: Hendrik Brueckner Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/early.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch/s390') diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 8d15314381e..cae14c49951 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -208,6 +208,9 @@ static noinline __init void detect_machine_type(void) machine_flags |= MACHINE_FLAG_KVM; else machine_flags |= MACHINE_FLAG_VM; + + /* Store machine flags for setting up lowcore early */ + S390_lowcore.machine_flags = machine_flags; } static __init void early_pgm_check_handler(void) -- cgit From 53cb780adbe69df90c8dc23e992ce40455e687c3 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Fri, 7 Aug 2009 10:39:25 +0200 Subject: [S390] KVM: Read buffer overflow Check whether index is within bounds before testing the element. Signed-off-by: Roel Kluin Signed-off-by: Martin Schwidefsky --- arch/s390/kvm/sigp.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c index 36678835034..0ef81d6776e 100644 --- a/arch/s390/kvm/sigp.c +++ b/arch/s390/kvm/sigp.c @@ -169,7 +169,7 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address, unsigned long *reg) { struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int; - struct kvm_s390_local_interrupt *li; + struct kvm_s390_local_interrupt *li = NULL; struct kvm_s390_interrupt_info *inti; int rc; u8 tmp; @@ -189,9 +189,10 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address, return 2; /* busy */ spin_lock(&fi->lock); - li = fi->local_int[cpu_addr]; + if (cpu_addr < KVM_MAX_VCPUS) + li = fi->local_int[cpu_addr]; - if ((cpu_addr >= KVM_MAX_VCPUS) || (li == NULL)) { + if (li == NULL) { rc = 1; /* incorrect state */ *reg &= SIGP_STAT_INCORRECT_STATE; kfree(inti); -- cgit