From 9673217c71b433d62d56cfb56c487e31af6f5574 Mon Sep 17 00:00:00 2001 From: Alexander Shiyan Date: Sun, 5 May 2013 15:59:29 +0400 Subject: [PATCH 1/9] s390: remove non existent reference to GENERIC_KERNEL_THREAD Signed-off-by: Alexander Shiyan Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/Kconfig | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 2c9789da0e24..da183c5a103c 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -98,7 +98,6 @@ config S390 select CLONE_BACKWARDS2 select GENERIC_CLOCKEVENTS select GENERIC_CPU_DEVICES if !SMP - select GENERIC_KERNEL_THREAD select GENERIC_SMP_IDLE_THREAD select GENERIC_TIME_VSYSCALL_OLD select HAVE_ALIGNED_STRUCT_PAGE if SLUB From 7678dcfb310a190aea9bee8cdeb1e14987600737 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 6 May 2013 16:26:01 +0200 Subject: [PATCH 2/9] s390/disassembler: prevent endless loop in print_fn_code() If the size of the opcode to be printed is larger than "len" we'll see an overflow of an unsigned long value, which means that the while loop within print_fn_code() will loop quite a long time until there is the next chance for an exit. So add an early exit check. Reported-by: Christian Ehrhardt Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/dis.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/s390/kernel/dis.c b/arch/s390/kernel/dis.c index 7f4a4a8c847c..be87d3e05a5b 100644 --- a/arch/s390/kernel/dis.c +++ b/arch/s390/kernel/dis.c @@ -1862,6 +1862,8 @@ void print_fn_code(unsigned char *code, unsigned long len) while (len) { ptr = buffer; opsize = insn_length(*code); + if (opsize > len) + break; ptr += sprintf(ptr, "%p: ", code); for (i = 0; i < opsize; i++) ptr += sprintf(ptr, "%02x", code[i]); From 617e164c35c94c936e3632b387d6dbb3d80c2c02 Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Mon, 6 May 2013 21:44:36 +0200 Subject: [PATCH 3/9] s390: disable pfmf for clear page instruction Wit the introduction of large pages Linux also used pfmf for page clearing. The current implementation is not ideal, though: - currently we set usage intent=0, but cleared pages are often used directly after the clearing - z/VM does not yet provide EDAT - KVM does have to intercept PFMF even for resident pages Lets just the mvcl loop in all cases until we have a well defined pattern were pfmf is besser. Signed-off-by: Christian Borntraeger Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/page.h | 20 ++++++++------------ 1 file changed, 8 insertions(+), 12 deletions(-) diff --git a/arch/s390/include/asm/page.h b/arch/s390/include/asm/page.h index 75ce9b065f9f..5d64fb7619cc 100644 --- a/arch/s390/include/asm/page.h +++ b/arch/s390/include/asm/page.h @@ -32,7 +32,7 @@ void storage_key_init_range(unsigned long start, unsigned long end); -static unsigned long pfmf(unsigned long function, unsigned long address) +static inline unsigned long pfmf(unsigned long function, unsigned long address) { asm volatile( " .insn rre,0xb9af0000,%[function],%[address]" @@ -44,17 +44,13 @@ static unsigned long pfmf(unsigned long function, unsigned long address) static inline void clear_page(void *page) { - if (MACHINE_HAS_PFMF) { - pfmf(0x10000, (unsigned long)page); - } else { - register unsigned long reg1 asm ("1") = 0; - register void *reg2 asm ("2") = page; - register unsigned long reg3 asm ("3") = 4096; - asm volatile( - " mvcl 2,0" - : "+d" (reg2), "+d" (reg3) : "d" (reg1) - : "memory", "cc"); - } + register unsigned long reg1 asm ("1") = 0; + register void *reg2 asm ("2") = page; + register unsigned long reg3 asm ("3") = 4096; + asm volatile( + " mvcl 2,0" + : "+d" (reg2), "+d" (reg3) : "d" (reg1) + : "memory", "cc"); } static inline void copy_page(void *to, void *from) From bb4b42ce0ca36af8c113587ab64b138b3cf5459c Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Wed, 8 May 2013 15:25:38 +0200 Subject: [PATCH 4/9] s390: fix gmap_ipte_notifier vs. software dirty pages On heavy paging load some guest cpus started to loop in gmap_ipte_notify. This was visible as stalled cpus inside the guest. The gmap_ipte_notifier tries to map a user page and then made sure that the pte is valid and writable. Turns out that with the software change bit tracking the pte can become read-only (and only software writable) if the page is clean. Since we loop in this code, the page would stay clean and, therefore, be never writable again. Let us just use fixup_user_fault, that guarantees to call handle_mm_fault. Signed-off-by: Christian Borntraeger Signed-off-by: Martin Schwidefsky --- arch/s390/mm/pgtable.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index 7805ddca833d..18dc417aaf79 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -677,8 +677,7 @@ int gmap_ipte_notify(struct gmap *gmap, unsigned long start, unsigned long len) break; } /* Get the page mapped */ - if (get_user_pages(current, gmap->mm, addr, 1, 1, 0, - NULL, NULL) != 1) { + if (fixup_user_fault(current, gmap->mm, addr, FAULT_FLAG_WRITE)) { rc = -EFAULT; break; } From aca91209775514b5a335e7ef11f1d636618590c7 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 13 May 2013 14:48:52 +0200 Subject: [PATCH 5/9] s390/ftrace: fix mcount adjustment Tony Jones reported that the ftrace self tests on s390 do not work: <6>Testing dynamic ftrace ops #1: (0 0 0 0 0) FAILED! <6>Testing tracer irqsoff: <3>failed to start irqsoff tracer <4>.. no entries found ..FAILED! <6>Testing tracer wakeup: <3>failed to start wakeup tracer <4>.. no entries found ..FAILED! <6>Testing tracer function_graph: <4>Failed to init function_graph tracer, init returned -19 <4>FAILED! This happens because we forgot to adjust the instruction pointer that gets passed to the ftrace trace function by MCOUNT_INSN_SIZE. In addition change MCOUNT_INSN_SIZE to the correct value on 31 bit. It only worked so far because the to be patched instruction was identical. Reported-by: Tony Jones Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/ftrace.h | 12 +++++++----- arch/s390/kernel/ftrace.c | 9 ++------- arch/s390/kernel/mcount.S | 2 ++ arch/s390/kernel/mcount64.S | 2 ++ 4 files changed, 13 insertions(+), 12 deletions(-) diff --git a/arch/s390/include/asm/ftrace.h b/arch/s390/include/asm/ftrace.h index b7931faaef6d..bf246dae1367 100644 --- a/arch/s390/include/asm/ftrace.h +++ b/arch/s390/include/asm/ftrace.h @@ -9,11 +9,6 @@ struct dyn_arch_ftrace { }; #define MCOUNT_ADDR ((long)_mcount) -#ifdef CONFIG_64BIT -#define MCOUNT_INSN_SIZE 12 -#else -#define MCOUNT_INSN_SIZE 20 -#endif static inline unsigned long ftrace_call_adjust(unsigned long addr) { @@ -21,4 +16,11 @@ static inline unsigned long ftrace_call_adjust(unsigned long addr) } #endif /* __ASSEMBLY__ */ + +#ifdef CONFIG_64BIT +#define MCOUNT_INSN_SIZE 12 +#else +#define MCOUNT_INSN_SIZE 22 +#endif + #endif /* _ASM_S390_FTRACE_H */ diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c index 78bdf0e5dff7..e3043aef87a9 100644 --- a/arch/s390/kernel/ftrace.c +++ b/arch/s390/kernel/ftrace.c @@ -16,12 +16,6 @@ #include #include -#ifdef CONFIG_64BIT -#define MCOUNT_OFFSET_RET 12 -#else -#define MCOUNT_OFFSET_RET 22 -#endif - #ifdef CONFIG_DYNAMIC_FTRACE void ftrace_disable_code(void); @@ -155,9 +149,10 @@ unsigned long __kprobes prepare_ftrace_return(unsigned long parent, if (unlikely(atomic_read(¤t->tracing_graph_pause))) goto out; + ip = (ip & PSW_ADDR_INSN) - MCOUNT_INSN_SIZE; if (ftrace_push_return_trace(parent, ip, &trace.depth, 0) == -EBUSY) goto out; - trace.func = (ip & PSW_ADDR_INSN) - MCOUNT_OFFSET_RET; + trace.func = ip; /* Only trace if the calling function expects to. */ if (!ftrace_graph_entry(&trace)) { current->curr_ret_stack--; diff --git a/arch/s390/kernel/mcount.S b/arch/s390/kernel/mcount.S index 4567ce20d900..08dcf21cb8df 100644 --- a/arch/s390/kernel/mcount.S +++ b/arch/s390/kernel/mcount.S @@ -7,6 +7,7 @@ #include #include +#include .section .kprobes.text, "ax" @@ -33,6 +34,7 @@ ENTRY(ftrace_caller) la %r2,0(%r14) st %r0,__SF_BACKCHAIN(%r15) la %r3,0(%r3) + ahi %r2,-MCOUNT_INSN_SIZE l %r14,0b-0b(%r1) l %r14,0(%r14) basr %r14,%r14 diff --git a/arch/s390/kernel/mcount64.S b/arch/s390/kernel/mcount64.S index 11332193db30..1c52eae3396a 100644 --- a/arch/s390/kernel/mcount64.S +++ b/arch/s390/kernel/mcount64.S @@ -7,6 +7,7 @@ #include #include +#include .section .kprobes.text, "ax" @@ -29,6 +30,7 @@ ENTRY(ftrace_caller) stg %r1,__SF_BACKCHAIN(%r15) lgr %r2,%r14 lg %r3,168(%r15) + aghi %r2,-MCOUNT_INSN_SIZE larl %r14,ftrace_trace_function lg %r14,0(%r14) basr %r14,%r14 From ba54229db65b161d46df449638a1a386444681b0 Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Wed, 15 May 2013 13:43:29 +0200 Subject: [PATCH 6/9] s390/cio: add channel ID sysfs attribute Add new attributes "chid" and "chid_external" to the channel-path sysfs directory. These attributes contain information related to the channel-ID of the channel-path. Reviewed-by: Sebastian Ott Signed-off-by: Peter Oberparleiter Signed-off-by: Martin Schwidefsky --- drivers/s390/cio/chp.c | 36 ++++++++++++++++++++++++++++++++++++ drivers/s390/cio/chsc.h | 4 +++- 2 files changed, 39 insertions(+), 1 deletion(-) diff --git a/drivers/s390/cio/chp.c b/drivers/s390/cio/chp.c index 21fabc6d5a9c..6c440d4349d4 100644 --- a/drivers/s390/cio/chp.c +++ b/drivers/s390/cio/chp.c @@ -352,12 +352,48 @@ static ssize_t chp_shared_show(struct device *dev, static DEVICE_ATTR(shared, 0444, chp_shared_show, NULL); +static ssize_t chp_chid_show(struct device *dev, struct device_attribute *attr, + char *buf) +{ + struct channel_path *chp = to_channelpath(dev); + ssize_t rc; + + mutex_lock(&chp->lock); + if (chp->desc_fmt1.flags & 0x10) + rc = sprintf(buf, "%04x\n", chp->desc_fmt1.chid); + else + rc = 0; + mutex_unlock(&chp->lock); + + return rc; +} +static DEVICE_ATTR(chid, 0444, chp_chid_show, NULL); + +static ssize_t chp_chid_external_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct channel_path *chp = to_channelpath(dev); + ssize_t rc; + + mutex_lock(&chp->lock); + if (chp->desc_fmt1.flags & 0x10) + rc = sprintf(buf, "%x\n", chp->desc_fmt1.flags & 0x8 ? 1 : 0); + else + rc = 0; + mutex_unlock(&chp->lock); + + return rc; +} +static DEVICE_ATTR(chid_external, 0444, chp_chid_external_show, NULL); + static struct attribute *chp_attrs[] = { &dev_attr_status.attr, &dev_attr_configure.attr, &dev_attr_type.attr, &dev_attr_cmg.attr, &dev_attr_shared.attr, + &dev_attr_chid.attr, + &dev_attr_chid_external.attr, NULL, }; static struct attribute_group chp_attr_group = { diff --git a/drivers/s390/cio/chsc.h b/drivers/s390/cio/chsc.h index 349d5fc47196..e7ef2a683b8f 100644 --- a/drivers/s390/cio/chsc.h +++ b/drivers/s390/cio/chsc.h @@ -43,7 +43,9 @@ struct channel_path_desc_fmt1 { u8 chpid; u32:24; u8 chpp; - u32 unused[3]; + u32 unused[2]; + u16 chid; + u32:16; u16 mdc; u16:13; u8 r:1; From a4eeea4e530fee26918529eb1b36ae306095eef5 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Wed, 15 May 2013 16:53:07 +0200 Subject: [PATCH 7/9] s390/smp: fix cpu re-scan vs. cpu state The cpu-info array starts with a list of cpus in configured state, followed by the cpus in standby state. The comparison to decide which state a cpu has is incorrect, this causes configured cpus appear as standby cpus. The correct comparison is the index of the new cpu in the cpu-info array vs. the number of configured cpus. Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/smp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 8074cb4b7cbf..05674b669001 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -645,7 +645,7 @@ static int __cpuinit __smp_rescan_cpus(struct sclp_cpu_info *info, continue; pcpu = pcpu_devices + cpu; pcpu->address = info->cpu[i].address; - pcpu->state = (cpu >= info->configured) ? + pcpu->state = (i >= info->configured) ? CPU_STATE_STANDBY : CPU_STATE_CONFIGURED; smp_cpu_set_polarization(cpu, POLARIZATION_UNKNOWN); set_cpu_present(cpu, true); From 00e5d35c0ffcbc59d7126387a0ca74a3abe62f59 Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Thu, 16 May 2013 15:27:13 +0200 Subject: [PATCH 8/9] s390/xpram: mark xpram as non-rotational xpram is a memory technology. Lets mark that as non-rotational. Signed-off-by: Christian Borntraeger Signed-off-by: Martin Schwidefsky --- drivers/s390/block/xpram.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/s390/block/xpram.c b/drivers/s390/block/xpram.c index 690c3338a8ae..464dd29d06c0 100644 --- a/drivers/s390/block/xpram.c +++ b/drivers/s390/block/xpram.c @@ -343,6 +343,7 @@ static int __init xpram_setup_blkdev(void) put_disk(xpram_disks[i]); goto out; } + queue_flag_set_unlocked(QUEUE_FLAG_NONROT, xpram_queues[i]); blk_queue_make_request(xpram_queues[i], xpram_make_request); blk_queue_logical_block_size(xpram_queues[i], 4096); } From 6b8224e40af147d3300136ce6d037db48f89068f Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Thu, 16 May 2013 22:16:32 +0200 Subject: [PATCH 9/9] s390/pgtable: fix ipte notify bit Dont use the same bit as user referenced. Signed-off-by: Christian Borntraeger --- arch/s390/include/asm/pgtable.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 4105b8221fdd..0f0de30e3e3f 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -306,7 +306,7 @@ extern unsigned long MODULES_END; #define RCP_HC_BIT 0x00200000UL #define RCP_GR_BIT 0x00040000UL #define RCP_GC_BIT 0x00020000UL -#define RCP_IN_BIT 0x00008000UL /* IPTE notify bit */ +#define RCP_IN_BIT 0x00002000UL /* IPTE notify bit */ /* User dirty / referenced bit for KVM's migration feature */ #define KVM_UR_BIT 0x00008000UL @@ -374,7 +374,7 @@ extern unsigned long MODULES_END; #define RCP_HC_BIT 0x0020000000000000UL #define RCP_GR_BIT 0x0004000000000000UL #define RCP_GC_BIT 0x0002000000000000UL -#define RCP_IN_BIT 0x0000800000000000UL /* IPTE notify bit */ +#define RCP_IN_BIT 0x0000200000000000UL /* IPTE notify bit */ /* User dirty / referenced bit for KVM's migration feature */ #define KVM_UR_BIT 0x0000800000000000UL