aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
-rw-r--r--arch/mips/Kconfig7
-rw-r--r--arch/mips/alchemy/common/time.c2
-rw-r--r--arch/mips/cavium-octeon/setup.c2
-rw-r--r--arch/mips/configs/ip27_defconfig2
-rw-r--r--arch/mips/include/asm/atomic.h52
-rw-r--r--arch/mips/include/asm/mach-rc32434/gpio.h5
-rw-r--r--arch/mips/include/asm/mach-rc32434/irq.h3
-rw-r--r--arch/mips/include/asm/mach-rc32434/rb.h3
-rw-r--r--arch/mips/include/asm/ptrace.h2
-rw-r--r--arch/mips/include/asm/termios.h100
-rw-r--r--arch/mips/include/asm/txx9/tx4939.h1
-rw-r--r--arch/mips/kernel/genex.S6
-rw-r--r--arch/mips/kernel/mips-mt-fpaff.c3
-rw-r--r--arch/mips/kernel/traps.c16
-rw-r--r--arch/mips/lib/memcpy-inatomic.S2
-rw-r--r--arch/mips/lib/memcpy.S2
-rw-r--r--arch/mips/mm/c-r4k.c22
-rw-r--r--arch/mips/mm/fault.c21
-rw-r--r--arch/mips/pci/pci-rc32434.c11
-rw-r--r--arch/mips/rb532/devices.c57
-rw-r--r--arch/mips/rb532/gpio.c90
-rw-r--r--arch/mips/rb532/irq.c27
-rw-r--r--arch/mips/rb532/serial.c2
-rw-r--r--arch/mips/txx9/generic/setup_tx4939.c22
-rw-r--r--arch/mips/txx9/rbtx4939/setup.c1
-rw-r--r--drivers/char/selection.c2
-rw-r--r--drivers/gpu/drm/drm_gem.c2
-rw-r--r--drivers/net/igb/e1000_82575.c11
-rw-r--r--drivers/net/igb/igb.h9
-rw-r--r--drivers/net/igb/igb_main.c24
-rw-r--r--drivers/net/sfc/efx.c59
-rw-r--r--drivers/net/sfc/efx.h9
-rw-r--r--drivers/net/sfc/ethtool.c3
-rw-r--r--drivers/net/sfc/falcon.c32
-rw-r--r--drivers/net/sfc/mdio_10g.c191
-rw-r--r--drivers/net/sfc/mdio_10g.h3
-rw-r--r--drivers/net/sfc/net_driver.h9
-rw-r--r--drivers/net/sfc/phy.h1
-rw-r--r--drivers/net/sfc/selftest.c7
-rw-r--r--drivers/net/sfc/sfe4001.c42
-rw-r--r--drivers/net/sfc/tenxpress.c213
-rw-r--r--drivers/net/sfc/workarounds.h12
-rw-r--r--drivers/net/sky2.c6
-rw-r--r--ipc/shm.c4
-rw-r--r--mm/mmap.c48
-rw-r--r--mm/shmem.c2
-rw-r--r--net/ipv6/ip6mr.c1
-rw-r--r--net/packet/af_packet.c9
48 files changed, 630 insertions, 530 deletions
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index 52c80c2a57f2..600eef3f3ac7 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -351,7 +351,7 @@ config SGI_IP27
351 select ARC64 351 select ARC64
352 select BOOT_ELF64 352 select BOOT_ELF64
353 select DEFAULT_SGI_PARTITION 353 select DEFAULT_SGI_PARTITION
354 select DMA_IP27 354 select DMA_COHERENT
355 select SYS_HAS_EARLY_PRINTK 355 select SYS_HAS_EARLY_PRINTK
356 select HW_HAS_PCI 356 select HW_HAS_PCI
357 select NR_CPUS_DEFAULT_64 357 select NR_CPUS_DEFAULT_64
@@ -761,9 +761,6 @@ config CFE
761config DMA_COHERENT 761config DMA_COHERENT
762 bool 762 bool
763 763
764config DMA_IP27
765 bool
766
767config DMA_NONCOHERENT 764config DMA_NONCOHERENT
768 bool 765 bool
769 select DMA_NEED_PCI_MAP_STATE 766 select DMA_NEED_PCI_MAP_STATE
@@ -1368,7 +1365,7 @@ config CPU_SUPPORTS_64BIT_KERNEL
1368# 1365#
1369config HARDWARE_WATCHPOINTS 1366config HARDWARE_WATCHPOINTS
1370 bool 1367 bool
1371 default y if CPU_MIPS32 || CPU_MIPS64 1368 default y if CPU_MIPSR1 || CPU_MIPSR2
1372 1369
1373menu "Kernel type" 1370menu "Kernel type"
1374 1371
diff --git a/arch/mips/alchemy/common/time.c b/arch/mips/alchemy/common/time.c
index 32880146cbc1..6fd441d16af5 100644
--- a/arch/mips/alchemy/common/time.c
+++ b/arch/mips/alchemy/common/time.c
@@ -89,7 +89,7 @@ static struct clock_event_device au1x_rtcmatch2_clockdev = {
89 .irq = AU1000_RTC_MATCH2_INT, 89 .irq = AU1000_RTC_MATCH2_INT,
90 .set_next_event = au1x_rtcmatch2_set_next_event, 90 .set_next_event = au1x_rtcmatch2_set_next_event,
91 .set_mode = au1x_rtcmatch2_set_mode, 91 .set_mode = au1x_rtcmatch2_set_mode,
92 .cpumask = CPU_MASK_ALL, 92 .cpumask = CPU_MASK_ALL_PTR,
93}; 93};
94 94
95static struct irqaction au1x_rtcmatch2_irqaction = { 95static struct irqaction au1x_rtcmatch2_irqaction = {
diff --git a/arch/mips/cavium-octeon/setup.c b/arch/mips/cavium-octeon/setup.c
index e085feddb4a4..5f4e49ba4713 100644
--- a/arch/mips/cavium-octeon/setup.c
+++ b/arch/mips/cavium-octeon/setup.c
@@ -15,13 +15,11 @@
15#include <linux/serial.h> 15#include <linux/serial.h>
16#include <linux/types.h> 16#include <linux/types.h>
17#include <linux/string.h> /* for memset */ 17#include <linux/string.h> /* for memset */
18#include <linux/serial.h>
19#include <linux/tty.h> 18#include <linux/tty.h>
20#include <linux/time.h> 19#include <linux/time.h>
21#include <linux/platform_device.h> 20#include <linux/platform_device.h>
22#include <linux/serial_core.h> 21#include <linux/serial_core.h>
23#include <linux/serial_8250.h> 22#include <linux/serial_8250.h>
24#include <linux/string.h>
25 23
26#include <asm/processor.h> 24#include <asm/processor.h>
27#include <asm/reboot.h> 25#include <asm/reboot.h>
diff --git a/arch/mips/configs/ip27_defconfig b/arch/mips/configs/ip27_defconfig
index 34ea319be94c..f2baea3039bb 100644
--- a/arch/mips/configs/ip27_defconfig
+++ b/arch/mips/configs/ip27_defconfig
@@ -53,7 +53,7 @@ CONFIG_GENERIC_TIME=y
53CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y 53CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER=y
54CONFIG_GENERIC_HARDIRQS_NO__DO_IRQ=y 54CONFIG_GENERIC_HARDIRQS_NO__DO_IRQ=y
55CONFIG_ARC=y 55CONFIG_ARC=y
56CONFIG_DMA_IP27=y 56CONFIG_DMA_COHERENT=y
57CONFIG_EARLY_PRINTK=y 57CONFIG_EARLY_PRINTK=y
58CONFIG_SYS_HAS_EARLY_PRINTK=y 58CONFIG_SYS_HAS_EARLY_PRINTK=y
59# CONFIG_NO_IOPORT is not set 59# CONFIG_NO_IOPORT is not set
diff --git a/arch/mips/include/asm/atomic.h b/arch/mips/include/asm/atomic.h
index c996c3b4d074..1b332e15ab52 100644
--- a/arch/mips/include/asm/atomic.h
+++ b/arch/mips/include/asm/atomic.h
@@ -50,7 +50,7 @@
50static __inline__ void atomic_add(int i, atomic_t * v) 50static __inline__ void atomic_add(int i, atomic_t * v)
51{ 51{
52 if (cpu_has_llsc && R10000_LLSC_WAR) { 52 if (cpu_has_llsc && R10000_LLSC_WAR) {
53 unsigned long temp; 53 int temp;
54 54
55 __asm__ __volatile__( 55 __asm__ __volatile__(
56 " .set mips3 \n" 56 " .set mips3 \n"
@@ -62,7 +62,7 @@ static __inline__ void atomic_add(int i, atomic_t * v)
62 : "=&r" (temp), "=m" (v->counter) 62 : "=&r" (temp), "=m" (v->counter)
63 : "Ir" (i), "m" (v->counter)); 63 : "Ir" (i), "m" (v->counter));
64 } else if (cpu_has_llsc) { 64 } else if (cpu_has_llsc) {
65 unsigned long temp; 65 int temp;
66 66
67 __asm__ __volatile__( 67 __asm__ __volatile__(
68 " .set mips3 \n" 68 " .set mips3 \n"
@@ -95,7 +95,7 @@ static __inline__ void atomic_add(int i, atomic_t * v)
95static __inline__ void atomic_sub(int i, atomic_t * v) 95static __inline__ void atomic_sub(int i, atomic_t * v)
96{ 96{
97 if (cpu_has_llsc && R10000_LLSC_WAR) { 97 if (cpu_has_llsc && R10000_LLSC_WAR) {
98 unsigned long temp; 98 int temp;
99 99
100 __asm__ __volatile__( 100 __asm__ __volatile__(
101 " .set mips3 \n" 101 " .set mips3 \n"
@@ -107,7 +107,7 @@ static __inline__ void atomic_sub(int i, atomic_t * v)
107 : "=&r" (temp), "=m" (v->counter) 107 : "=&r" (temp), "=m" (v->counter)
108 : "Ir" (i), "m" (v->counter)); 108 : "Ir" (i), "m" (v->counter));
109 } else if (cpu_has_llsc) { 109 } else if (cpu_has_llsc) {
110 unsigned long temp; 110 int temp;
111 111
112 __asm__ __volatile__( 112 __asm__ __volatile__(
113 " .set mips3 \n" 113 " .set mips3 \n"
@@ -135,12 +135,12 @@ static __inline__ void atomic_sub(int i, atomic_t * v)
135 */ 135 */
136static __inline__ int atomic_add_return(int i, atomic_t * v) 136static __inline__ int atomic_add_return(int i, atomic_t * v)
137{ 137{
138 unsigned long result; 138 int result;
139 139
140 smp_llsc_mb(); 140 smp_llsc_mb();
141 141
142 if (cpu_has_llsc && R10000_LLSC_WAR) { 142 if (cpu_has_llsc && R10000_LLSC_WAR) {
143 unsigned long temp; 143 int temp;
144 144
145 __asm__ __volatile__( 145 __asm__ __volatile__(
146 " .set mips3 \n" 146 " .set mips3 \n"
@@ -154,7 +154,7 @@ static __inline__ int atomic_add_return(int i, atomic_t * v)
154 : "Ir" (i), "m" (v->counter) 154 : "Ir" (i), "m" (v->counter)
155 : "memory"); 155 : "memory");
156 } else if (cpu_has_llsc) { 156 } else if (cpu_has_llsc) {
157 unsigned long temp; 157 int temp;
158 158
159 __asm__ __volatile__( 159 __asm__ __volatile__(
160 " .set mips3 \n" 160 " .set mips3 \n"
@@ -187,12 +187,12 @@ static __inline__ int atomic_add_return(int i, atomic_t * v)
187 187
188static __inline__ int atomic_sub_return(int i, atomic_t * v) 188static __inline__ int atomic_sub_return(int i, atomic_t * v)
189{ 189{
190 unsigned long result; 190 int result;
191 191
192 smp_llsc_mb(); 192 smp_llsc_mb();
193 193
194 if (cpu_has_llsc && R10000_LLSC_WAR) { 194 if (cpu_has_llsc && R10000_LLSC_WAR) {
195 unsigned long temp; 195 int temp;
196 196
197 __asm__ __volatile__( 197 __asm__ __volatile__(
198 " .set mips3 \n" 198 " .set mips3 \n"
@@ -206,7 +206,7 @@ static __inline__ int atomic_sub_return(int i, atomic_t * v)
206 : "Ir" (i), "m" (v->counter) 206 : "Ir" (i), "m" (v->counter)
207 : "memory"); 207 : "memory");
208 } else if (cpu_has_llsc) { 208 } else if (cpu_has_llsc) {
209 unsigned long temp; 209 int temp;
210 210
211 __asm__ __volatile__( 211 __asm__ __volatile__(
212 " .set mips3 \n" 212 " .set mips3 \n"
@@ -247,12 +247,12 @@ static __inline__ int atomic_sub_return(int i, atomic_t * v)
247 */ 247 */
248static __inline__ int atomic_sub_if_positive(int i, atomic_t * v) 248static __inline__ int atomic_sub_if_positive(int i, atomic_t * v)
249{ 249{
250 unsigned long result; 250 int result;
251 251
252 smp_llsc_mb(); 252 smp_llsc_mb();
253 253
254 if (cpu_has_llsc && R10000_LLSC_WAR) { 254 if (cpu_has_llsc && R10000_LLSC_WAR) {
255 unsigned long temp; 255 int temp;
256 256
257 __asm__ __volatile__( 257 __asm__ __volatile__(
258 " .set mips3 \n" 258 " .set mips3 \n"
@@ -270,7 +270,7 @@ static __inline__ int atomic_sub_if_positive(int i, atomic_t * v)
270 : "Ir" (i), "m" (v->counter) 270 : "Ir" (i), "m" (v->counter)
271 : "memory"); 271 : "memory");
272 } else if (cpu_has_llsc) { 272 } else if (cpu_has_llsc) {
273 unsigned long temp; 273 int temp;
274 274
275 __asm__ __volatile__( 275 __asm__ __volatile__(
276 " .set mips3 \n" 276 " .set mips3 \n"
@@ -429,7 +429,7 @@ static __inline__ int atomic_add_unless(atomic_t *v, int a, int u)
429static __inline__ void atomic64_add(long i, atomic64_t * v) 429static __inline__ void atomic64_add(long i, atomic64_t * v)
430{ 430{
431 if (cpu_has_llsc && R10000_LLSC_WAR) { 431 if (cpu_has_llsc && R10000_LLSC_WAR) {
432 unsigned long temp; 432 long temp;
433 433
434 __asm__ __volatile__( 434 __asm__ __volatile__(
435 " .set mips3 \n" 435 " .set mips3 \n"
@@ -441,7 +441,7 @@ static __inline__ void atomic64_add(long i, atomic64_t * v)
441 : "=&r" (temp), "=m" (v->counter) 441 : "=&r" (temp), "=m" (v->counter)
442 : "Ir" (i), "m" (v->counter)); 442 : "Ir" (i), "m" (v->counter));
443 } else if (cpu_has_llsc) { 443 } else if (cpu_has_llsc) {
444 unsigned long temp; 444 long temp;
445 445
446 __asm__ __volatile__( 446 __asm__ __volatile__(
447 " .set mips3 \n" 447 " .set mips3 \n"
@@ -474,7 +474,7 @@ static __inline__ void atomic64_add(long i, atomic64_t * v)
474static __inline__ void atomic64_sub(long i, atomic64_t * v) 474static __inline__ void atomic64_sub(long i, atomic64_t * v)
475{ 475{
476 if (cpu_has_llsc && R10000_LLSC_WAR) { 476 if (cpu_has_llsc && R10000_LLSC_WAR) {
477 unsigned long temp; 477 long temp;
478 478
479 __asm__ __volatile__( 479 __asm__ __volatile__(
480 " .set mips3 \n" 480 " .set mips3 \n"
@@ -486,7 +486,7 @@ static __inline__ void atomic64_sub(long i, atomic64_t * v)
486 : "=&r" (temp), "=m" (v->counter) 486 : "=&r" (temp), "=m" (v->counter)
487 : "Ir" (i), "m" (v->counter)); 487 : "Ir" (i), "m" (v->counter));
488 } else if (cpu_has_llsc) { 488 } else if (cpu_has_llsc) {
489 unsigned long temp; 489 long temp;
490 490
491 __asm__ __volatile__( 491 __asm__ __volatile__(
492 " .set mips3 \n" 492 " .set mips3 \n"
@@ -514,12 +514,12 @@ static __inline__ void atomic64_sub(long i, atomic64_t * v)
514 */ 514 */
515static __inline__ long atomic64_add_return(long i, atomic64_t * v) 515static __inline__ long atomic64_add_return(long i, atomic64_t * v)
516{ 516{
517 unsigned long result; 517 long result;
518 518
519 smp_llsc_mb(); 519 smp_llsc_mb();
520 520
521 if (cpu_has_llsc && R10000_LLSC_WAR) { 521 if (cpu_has_llsc && R10000_LLSC_WAR) {
522 unsigned long temp; 522 long temp;
523 523
524 __asm__ __volatile__( 524 __asm__ __volatile__(
525 " .set mips3 \n" 525 " .set mips3 \n"
@@ -533,7 +533,7 @@ static __inline__ long atomic64_add_return(long i, atomic64_t * v)
533 : "Ir" (i), "m" (v->counter) 533 : "Ir" (i), "m" (v->counter)
534 : "memory"); 534 : "memory");
535 } else if (cpu_has_llsc) { 535 } else if (cpu_has_llsc) {
536 unsigned long temp; 536 long temp;
537 537
538 __asm__ __volatile__( 538 __asm__ __volatile__(
539 " .set mips3 \n" 539 " .set mips3 \n"
@@ -566,12 +566,12 @@ static __inline__ long atomic64_add_return(long i, atomic64_t * v)
566 566
567static __inline__ long atomic64_sub_return(long i, atomic64_t * v) 567static __inline__ long atomic64_sub_return(long i, atomic64_t * v)
568{ 568{
569 unsigned long result; 569 long result;
570 570
571 smp_llsc_mb(); 571 smp_llsc_mb();
572 572
573 if (cpu_has_llsc && R10000_LLSC_WAR) { 573 if (cpu_has_llsc && R10000_LLSC_WAR) {
574 unsigned long temp; 574 long temp;
575 575
576 __asm__ __volatile__( 576 __asm__ __volatile__(
577 " .set mips3 \n" 577 " .set mips3 \n"
@@ -585,7 +585,7 @@ static __inline__ long atomic64_sub_return(long i, atomic64_t * v)
585 : "Ir" (i), "m" (v->counter) 585 : "Ir" (i), "m" (v->counter)
586 : "memory"); 586 : "memory");
587 } else if (cpu_has_llsc) { 587 } else if (cpu_has_llsc) {
588 unsigned long temp; 588 long temp;
589 589
590 __asm__ __volatile__( 590 __asm__ __volatile__(
591 " .set mips3 \n" 591 " .set mips3 \n"
@@ -626,12 +626,12 @@ static __inline__ long atomic64_sub_return(long i, atomic64_t * v)
626 */ 626 */
627static __inline__ long atomic64_sub_if_positive(long i, atomic64_t * v) 627static __inline__ long atomic64_sub_if_positive(long i, atomic64_t * v)
628{ 628{
629 unsigned long result; 629 long result;
630 630
631 smp_llsc_mb(); 631 smp_llsc_mb();
632 632
633 if (cpu_has_llsc && R10000_LLSC_WAR) { 633 if (cpu_has_llsc && R10000_LLSC_WAR) {
634 unsigned long temp; 634 long temp;
635 635
636 __asm__ __volatile__( 636 __asm__ __volatile__(
637 " .set mips3 \n" 637 " .set mips3 \n"
@@ -649,7 +649,7 @@ static __inline__ long atomic64_sub_if_positive(long i, atomic64_t * v)
649 : "Ir" (i), "m" (v->counter) 649 : "Ir" (i), "m" (v->counter)
650 : "memory"); 650 : "memory");
651 } else if (cpu_has_llsc) { 651 } else if (cpu_has_llsc) {
652 unsigned long temp; 652 long temp;
653 653
654 __asm__ __volatile__( 654 __asm__ __volatile__(
655 " .set mips3 \n" 655 " .set mips3 \n"
diff --git a/arch/mips/include/asm/mach-rc32434/gpio.h b/arch/mips/include/asm/mach-rc32434/gpio.h
index b5cf6457305a..3cb50d17b62d 100644
--- a/arch/mips/include/asm/mach-rc32434/gpio.h
+++ b/arch/mips/include/asm/mach-rc32434/gpio.h
@@ -80,11 +80,8 @@ struct rb532_gpio_reg {
80/* Compact Flash GPIO pin */ 80/* Compact Flash GPIO pin */
81#define CF_GPIO_NUM 13 81#define CF_GPIO_NUM 13
82 82
83extern void set_434_reg(unsigned reg_offs, unsigned bit, unsigned len, unsigned val);
84extern unsigned get_434_reg(unsigned reg_offs);
85extern void set_latch_u5(unsigned char or_mask, unsigned char nand_mask);
86extern unsigned char get_latch_u5(void);
87extern void rb532_gpio_set_ilevel(int bit, unsigned gpio); 83extern void rb532_gpio_set_ilevel(int bit, unsigned gpio);
88extern void rb532_gpio_set_istat(int bit, unsigned gpio); 84extern void rb532_gpio_set_istat(int bit, unsigned gpio);
85extern void rb532_gpio_set_func(unsigned gpio);
89 86
90#endif /* _RC32434_GPIO_H_ */ 87#endif /* _RC32434_GPIO_H_ */
diff --git a/arch/mips/include/asm/mach-rc32434/irq.h b/arch/mips/include/asm/mach-rc32434/irq.h
index 56738d8ec4e2..023a5b100ed0 100644
--- a/arch/mips/include/asm/mach-rc32434/irq.h
+++ b/arch/mips/include/asm/mach-rc32434/irq.h
@@ -30,4 +30,7 @@
30#define ETH0_RX_OVR_IRQ (GROUP3_IRQ_BASE + 9) 30#define ETH0_RX_OVR_IRQ (GROUP3_IRQ_BASE + 9)
31#define ETH0_TX_UND_IRQ (GROUP3_IRQ_BASE + 10) 31#define ETH0_TX_UND_IRQ (GROUP3_IRQ_BASE + 10)
32 32
33#define GPIO_MAPPED_IRQ_BASE GROUP4_IRQ_BASE
34#define GPIO_MAPPED_IRQ_GROUP 4
35
33#endif /* __ASM_RC32434_IRQ_H */ 36#endif /* __ASM_RC32434_IRQ_H */
diff --git a/arch/mips/include/asm/mach-rc32434/rb.h b/arch/mips/include/asm/mach-rc32434/rb.h
index f25a84916703..6dc5f8df1f3e 100644
--- a/arch/mips/include/asm/mach-rc32434/rb.h
+++ b/arch/mips/include/asm/mach-rc32434/rb.h
@@ -83,4 +83,7 @@ struct mpmc_device {
83 void __iomem *base; 83 void __iomem *base;
84}; 84};
85 85
86extern void set_latch_u5(unsigned char or_mask, unsigned char nand_mask);
87extern unsigned char get_latch_u5(void);
88
86#endif /* __ASM_RC32434_RB_H */ 89#endif /* __ASM_RC32434_RB_H */
diff --git a/arch/mips/include/asm/ptrace.h b/arch/mips/include/asm/ptrace.h
index 1f30d16d4669..ce47118e52b7 100644
--- a/arch/mips/include/asm/ptrace.h
+++ b/arch/mips/include/asm/ptrace.h
@@ -105,7 +105,7 @@ struct pt_watch_regs {
105 enum pt_watch_style style; 105 enum pt_watch_style style;
106 union { 106 union {
107 struct mips32_watch_regs mips32; 107 struct mips32_watch_regs mips32;
108 struct mips32_watch_regs mips64; 108 struct mips64_watch_regs mips64;
109 }; 109 };
110}; 110};
111 111
diff --git a/arch/mips/include/asm/termios.h b/arch/mips/include/asm/termios.h
index a275661fa7e1..8f77f774a2a0 100644
--- a/arch/mips/include/asm/termios.h
+++ b/arch/mips/include/asm/termios.h
@@ -9,6 +9,7 @@
9#ifndef _ASM_TERMIOS_H 9#ifndef _ASM_TERMIOS_H
10#define _ASM_TERMIOS_H 10#define _ASM_TERMIOS_H
11 11
12#include <linux/errno.h>
12#include <asm/termbits.h> 13#include <asm/termbits.h>
13#include <asm/ioctls.h> 14#include <asm/ioctls.h>
14 15
@@ -94,38 +95,81 @@ struct termio {
94/* 95/*
95 * Translate a "termio" structure into a "termios". Ugh. 96 * Translate a "termio" structure into a "termios". Ugh.
96 */ 97 */
97#define user_termio_to_kernel_termios(termios, termio) \ 98static inline int user_termio_to_kernel_termios(struct ktermios *termios,
98({ \ 99 struct termio __user *termio)
99 unsigned short tmp; \ 100{
100 get_user(tmp, &(termio)->c_iflag); \ 101 unsigned short iflag, oflag, cflag, lflag;
101 (termios)->c_iflag = (0xffff0000 & ((termios)->c_iflag)) | tmp; \ 102 unsigned int err;
102 get_user(tmp, &(termio)->c_oflag); \ 103
103 (termios)->c_oflag = (0xffff0000 & ((termios)->c_oflag)) | tmp; \ 104 if (!access_ok(VERIFY_READ, termio, sizeof(struct termio)))
104 get_user(tmp, &(termio)->c_cflag); \ 105 return -EFAULT;
105 (termios)->c_cflag = (0xffff0000 & ((termios)->c_cflag)) | tmp; \ 106
106 get_user(tmp, &(termio)->c_lflag); \ 107 err = __get_user(iflag, &termio->c_iflag);
107 (termios)->c_lflag = (0xffff0000 & ((termios)->c_lflag)) | tmp; \ 108 termios->c_iflag = (termios->c_iflag & 0xffff0000) | iflag;
108 get_user((termios)->c_line, &(termio)->c_line); \ 109 err |=__get_user(oflag, &termio->c_oflag);
109 copy_from_user((termios)->c_cc, (termio)->c_cc, NCC); \ 110 termios->c_oflag = (termios->c_oflag & 0xffff0000) | oflag;
110}) 111 err |=__get_user(cflag, &termio->c_cflag);
112 termios->c_cflag = (termios->c_cflag & 0xffff0000) | cflag;
113 err |=__get_user(lflag, &termio->c_lflag);
114 termios->c_lflag = (termios->c_lflag & 0xffff0000) | lflag;
115 err |=__get_user(termios->c_line, &termio->c_line);
116 if (err)
117 return -EFAULT;
118
119 if (__copy_from_user(termios->c_cc, termio->c_cc, NCC))
120 return -EFAULT;
121
122 return 0;
123}
111 124
112/* 125/*
113 * Translate a "termios" structure into a "termio". Ugh. 126 * Translate a "termios" structure into a "termio". Ugh.
114 */ 127 */
115#define kernel_termios_to_user_termio(termio, termios) \ 128static inline int kernel_termios_to_user_termio(struct termio __user *termio,
116({ \ 129 struct ktermios *termios)
117 put_user((termios)->c_iflag, &(termio)->c_iflag); \ 130{
118 put_user((termios)->c_oflag, &(termio)->c_oflag); \ 131 int err;
119 put_user((termios)->c_cflag, &(termio)->c_cflag); \ 132
120 put_user((termios)->c_lflag, &(termio)->c_lflag); \ 133 if (!access_ok(VERIFY_WRITE, termio, sizeof(struct termio)))
121 put_user((termios)->c_line, &(termio)->c_line); \ 134 return -EFAULT;
122 copy_to_user((termio)->c_cc, (termios)->c_cc, NCC); \ 135
123}) 136 err = __put_user(termios->c_iflag, &termio->c_iflag);
124 137 err |= __put_user(termios->c_oflag, &termio->c_oflag);
125#define user_termios_to_kernel_termios(k, u) copy_from_user(k, u, sizeof(struct termios2)) 138 err |= __put_user(termios->c_cflag, &termio->c_cflag);
126#define kernel_termios_to_user_termios(u, k) copy_to_user(u, k, sizeof(struct termios2)) 139 err |= __put_user(termios->c_lflag, &termio->c_lflag);
127#define user_termios_to_kernel_termios_1(k, u) copy_from_user(k, u, sizeof(struct termios)) 140 err |= __put_user(termios->c_line, &termio->c_line);
128#define kernel_termios_to_user_termios_1(u, k) copy_to_user(u, k, sizeof(struct termios)) 141 if (err)
142 return -EFAULT;
143
144 if (__copy_to_user(termio->c_cc, termios->c_cc, NCC))
145 return -EFAULT;
146
147 return 0;
148}
149
150static inline int user_termios_to_kernel_termios(struct ktermios __user *k,
151 struct termios2 *u)
152{
153 return copy_from_user(k, u, sizeof(struct termios2)) ? -EFAULT : 0;
154}
155
156static inline int kernel_termios_to_user_termios(struct termios2 __user *u,
157 struct ktermios *k)
158{
159 return copy_to_user(u, k, sizeof(struct termios2)) ? -EFAULT : 0;
160}
161
162static inline int user_termios_to_kernel_termios_1(struct ktermios *k,
163 struct termios __user *u)
164{
165 return copy_from_user(k, u, sizeof(struct termios)) ? -EFAULT : 0;
166}
167
168static inline int kernel_termios_to_user_termios_1(struct termios __user *u,
169 struct ktermios *k)
170{
171 return copy_to_user(u, k, sizeof(struct termios)) ? -EFAULT : 0;
172}
129 173
130#endif /* defined(__KERNEL__) */ 174#endif /* defined(__KERNEL__) */
131 175
diff --git a/arch/mips/include/asm/txx9/tx4939.h b/arch/mips/include/asm/txx9/tx4939.h
index 88badb423010..964ef7ede268 100644
--- a/arch/mips/include/asm/txx9/tx4939.h
+++ b/arch/mips/include/asm/txx9/tx4939.h
@@ -541,5 +541,6 @@ void tx4939_irq_init(void);
541int tx4939_irq(void); 541int tx4939_irq(void);
542void tx4939_mtd_init(int ch); 542void tx4939_mtd_init(int ch);
543void tx4939_ata_init(void); 543void tx4939_ata_init(void);
544void tx4939_rtc_init(void);
544 545
545#endif /* __ASM_TXX9_TX4939_H */ 546#endif /* __ASM_TXX9_TX4939_H */
diff --git a/arch/mips/kernel/genex.S b/arch/mips/kernel/genex.S
index fb6f73148df2..8882e5766f27 100644
--- a/arch/mips/kernel/genex.S
+++ b/arch/mips/kernel/genex.S
@@ -458,7 +458,11 @@ NESTED(nmi_handler, PT_SIZE, sp)
458 BUILD_HANDLER fpe fpe fpe silent /* #15 */ 458 BUILD_HANDLER fpe fpe fpe silent /* #15 */
459 BUILD_HANDLER mdmx mdmx sti silent /* #22 */ 459 BUILD_HANDLER mdmx mdmx sti silent /* #22 */
460#ifdef CONFIG_HARDWARE_WATCHPOINTS 460#ifdef CONFIG_HARDWARE_WATCHPOINTS
461 BUILD_HANDLER watch watch sti silent /* #23 */ 461 /*
462 * For watch, interrupts will be enabled after the watch
463 * registers are read.
464 */
465 BUILD_HANDLER watch watch cli silent /* #23 */
462#else 466#else
463 BUILD_HANDLER watch watch sti verbose /* #23 */ 467 BUILD_HANDLER watch watch sti verbose /* #23 */
464#endif 468#endif
diff --git a/arch/mips/kernel/mips-mt-fpaff.c b/arch/mips/kernel/mips-mt-fpaff.c
index 5e77a3a21f98..42461310b185 100644
--- a/arch/mips/kernel/mips-mt-fpaff.c
+++ b/arch/mips/kernel/mips-mt-fpaff.c
@@ -79,7 +79,8 @@ asmlinkage long mipsmt_sys_sched_setaffinity(pid_t pid, unsigned int len,
79 79
80 euid = current_euid(); 80 euid = current_euid();
81 retval = -EPERM; 81 retval = -EPERM;
82 if (euid != p->euid && euid != p->uid && !capable(CAP_SYS_NICE)) { 82 if (euid != p->cred->euid && euid != p->cred->uid &&
83 !capable(CAP_SYS_NICE)) {
83 read_unlock(&tasklist_lock); 84 read_unlock(&tasklist_lock);
84 goto out_unlock; 85 goto out_unlock;
85 } 86 }
diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c
index f6083c6bfaa4..b2d7041341b8 100644
--- a/arch/mips/kernel/traps.c
+++ b/arch/mips/kernel/traps.c
@@ -944,6 +944,9 @@ asmlinkage void do_mdmx(struct pt_regs *regs)
944 force_sig(SIGILL, current); 944 force_sig(SIGILL, current);
945} 945}
946 946
947/*
948 * Called with interrupts disabled.
949 */
947asmlinkage void do_watch(struct pt_regs *regs) 950asmlinkage void do_watch(struct pt_regs *regs)
948{ 951{
949 u32 cause; 952 u32 cause;
@@ -963,9 +966,12 @@ asmlinkage void do_watch(struct pt_regs *regs)
963 */ 966 */
964 if (test_tsk_thread_flag(current, TIF_LOAD_WATCH)) { 967 if (test_tsk_thread_flag(current, TIF_LOAD_WATCH)) {
965 mips_read_watch_registers(); 968 mips_read_watch_registers();
969 local_irq_enable();
966 force_sig(SIGTRAP, current); 970 force_sig(SIGTRAP, current);
967 } else 971 } else {
968 mips_clear_watch_registers(); 972 mips_clear_watch_registers();
973 local_irq_enable();
974 }
969} 975}
970 976
971asmlinkage void do_mcheck(struct pt_regs *regs) 977asmlinkage void do_mcheck(struct pt_regs *regs)
@@ -1582,7 +1588,11 @@ void __init set_handler(unsigned long offset, void *addr, unsigned long size)
1582static char panic_null_cerr[] __cpuinitdata = 1588static char panic_null_cerr[] __cpuinitdata =
1583 "Trying to set NULL cache error exception handler"; 1589 "Trying to set NULL cache error exception handler";
1584 1590
1585/* Install uncached CPU exception handler */ 1591/*
1592 * Install uncached CPU exception handler.
1593 * This is suitable only for the cache error exception which is the only
1594 * exception handler that is being run uncached.
1595 */
1586void __cpuinit set_uncached_handler(unsigned long offset, void *addr, 1596void __cpuinit set_uncached_handler(unsigned long offset, void *addr,
1587 unsigned long size) 1597 unsigned long size)
1588{ 1598{
@@ -1593,7 +1603,7 @@ void __cpuinit set_uncached_handler(unsigned long offset, void *addr,
1593 unsigned long uncached_ebase = TO_UNCAC(ebase); 1603 unsigned long uncached_ebase = TO_UNCAC(ebase);
1594#endif 1604#endif
1595 if (cpu_has_mips_r2) 1605 if (cpu_has_mips_r2)
1596 ebase += (read_c0_ebase() & 0x3ffff000); 1606 uncached_ebase += (read_c0_ebase() & 0x3ffff000);
1597 1607
1598 if (!addr) 1608 if (!addr)
1599 panic(panic_null_cerr); 1609 panic(panic_null_cerr);
diff --git a/arch/mips/lib/memcpy-inatomic.S b/arch/mips/lib/memcpy-inatomic.S
index 736d0fb56a94..68853a038d3f 100644
--- a/arch/mips/lib/memcpy-inatomic.S
+++ b/arch/mips/lib/memcpy-inatomic.S
@@ -21,7 +21,7 @@
21 * end of memory on some systems. It's also a seriously bad idea on non 21 * end of memory on some systems. It's also a seriously bad idea on non
22 * dma-coherent systems. 22 * dma-coherent systems.
23 */ 23 */
24#if !defined(CONFIG_DMA_COHERENT) || !defined(CONFIG_DMA_IP27) 24#ifdef CONFIG_DMA_NONCOHERENT
25#undef CONFIG_CPU_HAS_PREFETCH 25#undef CONFIG_CPU_HAS_PREFETCH
26#endif 26#endif
27#ifdef CONFIG_MIPS_MALTA 27#ifdef CONFIG_MIPS_MALTA
diff --git a/arch/mips/lib/memcpy.S b/arch/mips/lib/memcpy.S
index c06cccf60bec..56a1f85a1ce8 100644
--- a/arch/mips/lib/memcpy.S
+++ b/arch/mips/lib/memcpy.S
@@ -21,7 +21,7 @@
21 * end of memory on some systems. It's also a seriously bad idea on non 21 * end of memory on some systems. It's also a seriously bad idea on non
22 * dma-coherent systems. 22 * dma-coherent systems.
23 */ 23 */
24#if !defined(CONFIG_DMA_COHERENT) || !defined(CONFIG_DMA_IP27) 24#ifdef CONFIG_DMA_NONCOHERENT
25#undef CONFIG_CPU_HAS_PREFETCH 25#undef CONFIG_CPU_HAS_PREFETCH
26#endif 26#endif
27#ifdef CONFIG_MIPS_MALTA 27#ifdef CONFIG_MIPS_MALTA
diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c
index 6e99665ae860..c43f4b26a690 100644
--- a/arch/mips/mm/c-r4k.c
+++ b/arch/mips/mm/c-r4k.c
@@ -618,15 +618,35 @@ static void r4k_dma_cache_inv(unsigned long addr, unsigned long size)
618 if (cpu_has_inclusive_pcaches) { 618 if (cpu_has_inclusive_pcaches) {
619 if (size >= scache_size) 619 if (size >= scache_size)
620 r4k_blast_scache(); 620 r4k_blast_scache();
621 else 621 else {
622 unsigned long lsize = cpu_scache_line_size();
623 unsigned long almask = ~(lsize - 1);
624
625 /*
626 * There is no clearly documented alignment requirement
627 * for the cache instruction on MIPS processors and
628 * some processors, among them the RM5200 and RM7000
629 * QED processors will throw an address error for cache
630 * hit ops with insufficient alignment. Solved by
631 * aligning the address to cache line size.
632 */
633 cache_op(Hit_Writeback_Inv_SD, addr & almask);
634 cache_op(Hit_Writeback_Inv_SD,
635 (addr + size - 1) & almask);
622 blast_inv_scache_range(addr, addr + size); 636 blast_inv_scache_range(addr, addr + size);
637 }
623 return; 638 return;
624 } 639 }
625 640
626 if (cpu_has_safe_index_cacheops && size >= dcache_size) { 641 if (cpu_has_safe_index_cacheops && size >= dcache_size) {
627 r4k_blast_dcache(); 642 r4k_blast_dcache();
628 } else { 643 } else {
644 unsigned long lsize = cpu_dcache_line_size();
645 unsigned long almask = ~(lsize - 1);
646
629 R4600_HIT_CACHEOP_WAR_IMPL; 647 R4600_HIT_CACHEOP_WAR_IMPL;
648 cache_op(Hit_Writeback_Inv_D, addr & almask);
649 cache_op(Hit_Writeback_Inv_D, (addr + size - 1) & almask);
630 blast_inv_dcache_range(addr, addr + size); 650 blast_inv_dcache_range(addr, addr + size);
631 } 651 }
632 652
diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c
index fa636fc6b7b9..55767ad9f00e 100644
--- a/arch/mips/mm/fault.c
+++ b/arch/mips/mm/fault.c
@@ -97,7 +97,6 @@ good_area:
97 goto bad_area; 97 goto bad_area;
98 } 98 }
99 99
100survive:
101 /* 100 /*
102 * If for any reason at all we couldn't handle the fault, 101 * If for any reason at all we couldn't handle the fault,
103 * make sure we exit gracefully rather than endlessly redo 102 * make sure we exit gracefully rather than endlessly redo
@@ -167,21 +166,13 @@ no_context:
167 field, regs->regs[31]); 166 field, regs->regs[31]);
168 die("Oops", regs); 167 die("Oops", regs);
169 168
170/*
171 * We ran out of memory, or some other thing happened to us that made
172 * us unable to handle the page fault gracefully.
173 */
174out_of_memory: 169out_of_memory:
175 up_read(&mm->mmap_sem); 170 /*
176 if (is_global_init(tsk)) { 171 * We ran out of memory, call the OOM killer, and return the userspace
177 yield(); 172 * (which will retry the fault, or kill us if we got oom-killed).
178 down_read(&mm->mmap_sem); 173 */
179 goto survive; 174 pagefault_out_of_memory();
180 } 175 return;
181 printk("VM: killing process %s\n", tsk->comm);
182 if (user_mode(regs))
183 do_group_exit(SIGKILL);
184 goto no_context;
185 176
186do_sigbus: 177do_sigbus:
187 up_read(&mm->mmap_sem); 178 up_read(&mm->mmap_sem);
diff --git a/arch/mips/pci/pci-rc32434.c b/arch/mips/pci/pci-rc32434.c
index 1c2821e2f494..71f7d27b0d4c 100644
--- a/arch/mips/pci/pci-rc32434.c
+++ b/arch/mips/pci/pci-rc32434.c
@@ -205,6 +205,8 @@ static int __init rc32434_pcibridge_init(void)
205 205
206static int __init rc32434_pci_init(void) 206static int __init rc32434_pci_init(void)
207{ 207{
208 void __iomem *io_map_base;
209
208 pr_info("PCI: Initializing PCI\n"); 210 pr_info("PCI: Initializing PCI\n");
209 211
210 ioport_resource.start = rc32434_res_pci_io1.start; 212 ioport_resource.start = rc32434_res_pci_io1.start;
@@ -212,6 +214,15 @@ static int __init rc32434_pci_init(void)
212 214
213 rc32434_pcibridge_init(); 215 rc32434_pcibridge_init();
214 216
217 io_map_base = ioremap(rc32434_res_pci_io1.start,
218 rc32434_res_pci_io1.end - rc32434_res_pci_io1.start + 1);
219
220 if (!io_map_base)
221 return -ENOMEM;
222
223 rc32434_controller.io_map_base =
224 (unsigned long)io_map_base - rc32434_res_pci_io1.start;
225
215 register_pci_controller(&rc32434_controller); 226 register_pci_controller(&rc32434_controller);
216 rc32434_sync(); 227 rc32434_sync();
217 228
diff --git a/arch/mips/rb532/devices.c b/arch/mips/rb532/devices.c
index c1c29181bd46..4a5f05b662ae 100644
--- a/arch/mips/rb532/devices.c
+++ b/arch/mips/rb532/devices.c
@@ -24,6 +24,7 @@
24#include <linux/mtd/partitions.h> 24#include <linux/mtd/partitions.h>
25#include <linux/gpio_keys.h> 25#include <linux/gpio_keys.h>
26#include <linux/input.h> 26#include <linux/input.h>
27#include <linux/serial_8250.h>
27 28
28#include <asm/bootinfo.h> 29#include <asm/bootinfo.h>
29 30
@@ -39,6 +40,29 @@
39#define ETH0_RX_DMA_ADDR (DMA0_BASE_ADDR + 0 * DMA_CHAN_OFFSET) 40#define ETH0_RX_DMA_ADDR (DMA0_BASE_ADDR + 0 * DMA_CHAN_OFFSET)
40#define ETH0_TX_DMA_ADDR (DMA0_BASE_ADDR + 1 * DMA_CHAN_OFFSET) 41#define ETH0_TX_DMA_ADDR (DMA0_BASE_ADDR + 1 * DMA_CHAN_OFFSET)
41 42
43extern unsigned int idt_cpu_freq;
44
45static struct mpmc_device dev3;
46
47void set_latch_u5(unsigned char or_mask, unsigned char nand_mask)
48{
49 unsigned long flags;
50
51 spin_lock_irqsave(&dev3.lock, flags);
52
53 dev3.state = (dev3.state | or_mask) & ~nand_mask;
54 writeb(dev3.state, dev3.base);
55
56 spin_unlock_irqrestore(&dev3.lock, flags);
57}
58EXPORT_SYMBOL(set_latch_u5);
59
60unsigned char get_latch_u5(void)
61{
62 return dev3.state;
63}
64EXPORT_SYMBOL(get_latch_u5);
65
42static struct resource korina_dev0_res[] = { 66static struct resource korina_dev0_res[] = {
43 { 67 {
44 .name = "korina_regs", 68 .name = "korina_regs",
@@ -86,7 +110,7 @@ static struct korina_device korina_dev0_data = {
86static struct platform_device korina_dev0 = { 110static struct platform_device korina_dev0 = {
87 .id = -1, 111 .id = -1,
88 .name = "korina", 112 .name = "korina",
89 .dev.platform_data = &korina_dev0_data, 113 .dev.driver_data = &korina_dev0_data,
90 .resource = korina_dev0_res, 114 .resource = korina_dev0_res,
91 .num_resources = ARRAY_SIZE(korina_dev0_res), 115 .num_resources = ARRAY_SIZE(korina_dev0_res),
92}; 116};
@@ -214,12 +238,32 @@ static struct platform_device rb532_wdt = {
214 .num_resources = ARRAY_SIZE(rb532_wdt_res), 238 .num_resources = ARRAY_SIZE(rb532_wdt_res),
215}; 239};
216 240
241static struct plat_serial8250_port rb532_uart_res[] = {
242 {
243 .membase = (char *)KSEG1ADDR(REGBASE + UART0BASE),
244 .irq = UART0_IRQ,
245 .regshift = 2,
246 .iotype = UPIO_MEM,
247 .flags = UPF_BOOT_AUTOCONF,
248 },
249 {
250 .flags = 0,
251 }
252};
253
254static struct platform_device rb532_uart = {
255 .name = "serial8250",
256 .id = PLAT8250_DEV_PLATFORM,
257 .dev.platform_data = &rb532_uart_res,
258};
259
217static struct platform_device *rb532_devs[] = { 260static struct platform_device *rb532_devs[] = {
218 &korina_dev0, 261 &korina_dev0,
219 &nand_slot0, 262 &nand_slot0,
220 &cf_slot0, 263 &cf_slot0,
221 &rb532_led, 264 &rb532_led,
222 &rb532_button, 265 &rb532_button,
266 &rb532_uart,
223 &rb532_wdt 267 &rb532_wdt
224}; 268};
225 269
@@ -291,9 +335,20 @@ static int __init plat_setup_devices(void)
291 nand_slot0_res[0].start = readl(IDT434_REG_BASE + DEV2BASE); 335 nand_slot0_res[0].start = readl(IDT434_REG_BASE + DEV2BASE);
292 nand_slot0_res[0].end = nand_slot0_res[0].start + 0x1000; 336 nand_slot0_res[0].end = nand_slot0_res[0].start + 0x1000;
293 337
338 /* Read and map device controller 3 */
339 dev3.base = ioremap_nocache(readl(IDT434_REG_BASE + DEV3BASE), 1);
340
341 if (!dev3.base) {
342 printk(KERN_ERR "rb532: cannot remap device controller 3\n");
343 return -ENXIO;
344 }
345
294 /* Initialise the NAND device */ 346 /* Initialise the NAND device */
295 rb532_nand_setup(); 347 rb532_nand_setup();
296 348
349 /* set the uart clock to the current cpu frequency */
350 rb532_uart_res[0].uartclk = idt_cpu_freq;
351
297 return platform_add_devices(rb532_devs, ARRAY_SIZE(rb532_devs)); 352 return platform_add_devices(rb532_devs, ARRAY_SIZE(rb532_devs));
298} 353}
299 354
diff --git a/arch/mips/rb532/gpio.c b/arch/mips/rb532/gpio.c
index 0e84c8ab6a39..37de05d595e7 100644
--- a/arch/mips/rb532/gpio.c
+++ b/arch/mips/rb532/gpio.c
@@ -41,8 +41,6 @@ struct rb532_gpio_chip {
41 void __iomem *regbase; 41 void __iomem *regbase;
42}; 42};
43 43
44struct mpmc_device dev3;
45
46static struct resource rb532_gpio_reg0_res[] = { 44static struct resource rb532_gpio_reg0_res[] = {
47 { 45 {
48 .name = "gpio_reg0", 46 .name = "gpio_reg0",
@@ -52,61 +50,6 @@ static struct resource rb532_gpio_reg0_res[] = {
52 } 50 }
53}; 51};
54 52
55static struct resource rb532_dev3_ctl_res[] = {
56 {
57 .name = "dev3_ctl",
58 .start = REGBASE + DEV3BASE,
59 .end = REGBASE + DEV3BASE + sizeof(struct dev_reg) - 1,
60 .flags = IORESOURCE_MEM,
61 }
62};
63
64void set_434_reg(unsigned reg_offs, unsigned bit, unsigned len, unsigned val)
65{
66 unsigned long flags;
67 unsigned data;
68 unsigned i = 0;
69
70 spin_lock_irqsave(&dev3.lock, flags);
71
72 data = readl(IDT434_REG_BASE + reg_offs);
73 for (i = 0; i != len; ++i) {
74 if (val & (1 << i))
75 data |= (1 << (i + bit));
76 else
77 data &= ~(1 << (i + bit));
78 }
79 writel(data, (IDT434_REG_BASE + reg_offs));
80
81 spin_unlock_irqrestore(&dev3.lock, flags);
82}
83EXPORT_SYMBOL(set_434_reg);
84
85unsigned get_434_reg(unsigned reg_offs)
86{
87 return readl(IDT434_REG_BASE + reg_offs);
88}
89EXPORT_SYMBOL(get_434_reg);
90
91void set_latch_u5(unsigned char or_mask, unsigned char nand_mask)
92{
93 unsigned long flags;
94
95 spin_lock_irqsave(&dev3.lock, flags);
96
97 dev3.state = (dev3.state | or_mask) & ~nand_mask;
98 writel(dev3.state, &dev3.base);
99
100 spin_unlock_irqrestore(&dev3.lock, flags);
101}
102EXPORT_SYMBOL(set_latch_u5);
103
104unsigned char get_latch_u5(void)
105{
106 return dev3.state;
107}
108EXPORT_SYMBOL(get_latch_u5);
109
110/* rb532_set_bit - sanely set a bit 53/* rb532_set_bit - sanely set a bit
111 * 54 *
112 * bitval: new value for the bit 55 * bitval: new value for the bit
@@ -119,13 +62,11 @@ static inline void rb532_set_bit(unsigned bitval,
119 unsigned long flags; 62 unsigned long flags;
120 u32 val; 63 u32 val;
121 64
122 bitval = !!bitval; /* map parameter to {0,1} */
123
124 local_irq_save(flags); 65 local_irq_save(flags);
125 66
126 val = readl(ioaddr); 67 val = readl(ioaddr);
127 val &= ~( ~bitval << offset ); /* unset bit if bitval == 0 */ 68 val &= ~(!bitval << offset); /* unset bit if bitval == 0 */
128 val |= ( bitval << offset ); /* set bit if bitval == 1 */ 69 val |= (!!bitval << offset); /* set bit if bitval == 1 */
129 writel(val, ioaddr); 70 writel(val, ioaddr);
130 71
131 local_irq_restore(flags); 72 local_irq_restore(flags);
@@ -171,8 +112,8 @@ static int rb532_gpio_direction_input(struct gpio_chip *chip, unsigned offset)
171 112
172 gpch = container_of(chip, struct rb532_gpio_chip, chip); 113 gpch = container_of(chip, struct rb532_gpio_chip, chip);
173 114
174 if (rb532_get_bit(offset, gpch->regbase + GPIOFUNC)) 115 /* disable alternate function in case it's set */
175 return 1; /* alternate function, GPIOCFG is ignored */ 116 rb532_set_bit(0, offset, gpch->regbase + GPIOFUNC);
176 117
177 rb532_set_bit(0, offset, gpch->regbase + GPIOCFG); 118 rb532_set_bit(0, offset, gpch->regbase + GPIOCFG);
178 return 0; 119 return 0;
@@ -188,8 +129,8 @@ static int rb532_gpio_direction_output(struct gpio_chip *chip,
188 129
189 gpch = container_of(chip, struct rb532_gpio_chip, chip); 130 gpch = container_of(chip, struct rb532_gpio_chip, chip);
190 131
191 if (rb532_get_bit(offset, gpch->regbase + GPIOFUNC)) 132 /* disable alternate function in case it's set */
192 return 1; /* alternate function, GPIOCFG is ignored */ 133 rb532_set_bit(0, offset, gpch->regbase + GPIOFUNC);
193 134
194 /* set the initial output value */ 135 /* set the initial output value */
195 rb532_set_bit(value, offset, gpch->regbase + GPIOD); 136 rb532_set_bit(value, offset, gpch->regbase + GPIOD);
@@ -233,10 +174,11 @@ EXPORT_SYMBOL(rb532_gpio_set_istat);
233/* 174/*
234 * Configure GPIO alternate function 175 * Configure GPIO alternate function
235 */ 176 */
236static void rb532_gpio_set_func(int bit, unsigned gpio) 177void rb532_gpio_set_func(unsigned gpio)
237{ 178{
238 rb532_set_bit(bit, gpio, rb532_gpio_chip->regbase + GPIOFUNC); 179 rb532_set_bit(1, gpio, rb532_gpio_chip->regbase + GPIOFUNC);
239} 180}
181EXPORT_SYMBOL(rb532_gpio_set_func);
240 182
241int __init rb532_gpio_init(void) 183int __init rb532_gpio_init(void)
242{ 184{
@@ -253,20 +195,6 @@ int __init rb532_gpio_init(void)
253 /* Register our GPIO chip */ 195 /* Register our GPIO chip */
254 gpiochip_add(&rb532_gpio_chip->chip); 196 gpiochip_add(&rb532_gpio_chip->chip);
255 197
256 r = rb532_dev3_ctl_res;
257 dev3.base = ioremap_nocache(r->start, r->end - r->start);
258
259 if (!dev3.base) {
260 printk(KERN_ERR "rb532: cannot remap device controller 3\n");
261 return -ENXIO;
262 }
263
264 /* configure CF_GPIO_NUM as CFRDY IRQ source */
265 rb532_gpio_set_func(0, CF_GPIO_NUM);
266 rb532_gpio_direction_input(&rb532_gpio_chip->chip, CF_GPIO_NUM);
267 rb532_gpio_set_ilevel(1, CF_GPIO_NUM);
268 rb532_gpio_set_istat(0, CF_GPIO_NUM);
269
270 return 0; 198 return 0;
271} 199}
272arch_initcall(rb532_gpio_init); 200arch_initcall(rb532_gpio_init);
diff --git a/arch/mips/rb532/irq.c b/arch/mips/rb532/irq.c
index 549b46d2fcee..53eeb5e7bc5b 100644
--- a/arch/mips/rb532/irq.c
+++ b/arch/mips/rb532/irq.c
@@ -46,6 +46,7 @@
46#include <asm/system.h> 46#include <asm/system.h>
47 47
48#include <asm/mach-rc32434/irq.h> 48#include <asm/mach-rc32434/irq.h>
49#include <asm/mach-rc32434/gpio.h>
49 50
50struct intr_group { 51struct intr_group {
51 u32 mask; /* mask of valid bits in pending/mask registers */ 52 u32 mask; /* mask of valid bits in pending/mask registers */
@@ -150,6 +151,9 @@ static void rb532_disable_irq(unsigned int irq_nr)
150 mask |= intr_bit; 151 mask |= intr_bit;
151 WRITE_MASK(addr, mask); 152 WRITE_MASK(addr, mask);
152 153
154 if (group == GPIO_MAPPED_IRQ_GROUP)
155 rb532_gpio_set_istat(0, irq_nr - GPIO_MAPPED_IRQ_BASE);
156
153 /* 157 /*
154 * if there are no more interrupts enabled in this 158 * if there are no more interrupts enabled in this
155 * group, disable corresponding IP 159 * group, disable corresponding IP
@@ -165,12 +169,35 @@ static void rb532_mask_and_ack_irq(unsigned int irq_nr)
165 ack_local_irq(group_to_ip(irq_to_group(irq_nr))); 169 ack_local_irq(group_to_ip(irq_to_group(irq_nr)));
166} 170}
167 171
172static int rb532_set_type(unsigned int irq_nr, unsigned type)
173{
174 int gpio = irq_nr - GPIO_MAPPED_IRQ_BASE;
175 int group = irq_to_group(irq_nr);
176
177 if (group != GPIO_MAPPED_IRQ_GROUP)
178 return (type == IRQ_TYPE_LEVEL_HIGH) ? 0 : -EINVAL;
179
180 switch (type) {
181 case IRQ_TYPE_LEVEL_HIGH:
182 rb532_gpio_set_ilevel(1, gpio);
183 break;
184 case IRQ_TYPE_LEVEL_LOW:
185 rb532_gpio_set_ilevel(0, gpio);
186 break;
187 default:
188 return -EINVAL;
189 }
190
191 return 0;
192}
193
168static struct irq_chip rc32434_irq_type = { 194static struct irq_chip rc32434_irq_type = {
169 .name = "RB532", 195 .name = "RB532",
170 .ack = rb532_disable_irq, 196 .ack = rb532_disable_irq,
171 .mask = rb532_disable_irq, 197 .mask = rb532_disable_irq,
172 .mask_ack = rb532_mask_and_ack_irq, 198 .mask_ack = rb532_mask_and_ack_irq,
173 .unmask = rb532_enable_irq, 199 .unmask = rb532_enable_irq,
200 .set_type = rb532_set_type,
174}; 201};
175 202
176void __init arch_init_irq(void) 203void __init arch_init_irq(void)
diff --git a/arch/mips/rb532/serial.c b/arch/mips/rb532/serial.c
index 3e0d7ec3a579..00ed19f0bdb5 100644
--- a/arch/mips/rb532/serial.c
+++ b/arch/mips/rb532/serial.c
@@ -36,7 +36,7 @@
36extern unsigned int idt_cpu_freq; 36extern unsigned int idt_cpu_freq;
37 37
38static struct uart_port rb532_uart = { 38static struct uart_port rb532_uart = {
39 .type = PORT_16550A, 39 .flags = UPF_BOOT_AUTOCONF,
40 .line = 0, 40 .line = 0,
41 .irq = UART0_IRQ, 41 .irq = UART0_IRQ,
42 .iotype = UPIO_MEM, 42 .iotype = UPIO_MEM,
diff --git a/arch/mips/txx9/generic/setup_tx4939.c b/arch/mips/txx9/generic/setup_tx4939.c
index 6c0049a5bbc1..55440967b3a8 100644
--- a/arch/mips/txx9/generic/setup_tx4939.c
+++ b/arch/mips/txx9/generic/setup_tx4939.c
@@ -435,6 +435,28 @@ void __init tx4939_ata_init(void)
435 platform_device_register(&ata1_dev); 435 platform_device_register(&ata1_dev);
436} 436}
437 437
438void __init tx4939_rtc_init(void)
439{
440 static struct resource res[] = {
441 {
442 .start = TX4939_RTC_REG & 0xfffffffffULL,
443 .end = (TX4939_RTC_REG & 0xfffffffffULL) + 0x100 - 1,
444 .flags = IORESOURCE_MEM,
445 }, {
446 .start = TXX9_IRQ_BASE + TX4939_IR_RTC,
447 .flags = IORESOURCE_IRQ,
448 },
449 };
450 static struct platform_device rtc_dev = {
451 .name = "tx4939rtc",
452 .id = -1,
453 .num_resources = ARRAY_SIZE(res),
454 .resource = res,
455 };
456
457 platform_device_register(&rtc_dev);
458}
459
438static void __init tx4939_stop_unused_modules(void) 460static void __init tx4939_stop_unused_modules(void)
439{ 461{
440 __u64 pcfg, rst = 0, ckd = 0; 462 __u64 pcfg, rst = 0, ckd = 0;
diff --git a/arch/mips/txx9/rbtx4939/setup.c b/arch/mips/txx9/rbtx4939/setup.c
index 98fbd9391bf8..656603b85b71 100644
--- a/arch/mips/txx9/rbtx4939/setup.c
+++ b/arch/mips/txx9/rbtx4939/setup.c
@@ -336,6 +336,7 @@ static void __init rbtx4939_device_init(void)
336 rbtx4939_led_setup(); 336 rbtx4939_led_setup();
337 tx4939_wdt_init(); 337 tx4939_wdt_init();
338 tx4939_ata_init(); 338 tx4939_ata_init();
339 tx4939_rtc_init();
339} 340}
340 341
341static void __init rbtx4939_setup(void) 342static void __init rbtx4939_setup(void)
diff --git a/drivers/char/selection.c b/drivers/char/selection.c
index f29fbe9b8ed7..cb8ca5698963 100644
--- a/drivers/char/selection.c
+++ b/drivers/char/selection.c
@@ -268,7 +268,7 @@ int set_selection(const struct tiocl_selection __user *sel, struct tty_struct *t
268 268
269 /* Allocate a new buffer before freeing the old one ... */ 269 /* Allocate a new buffer before freeing the old one ... */
270 multiplier = use_unicode ? 3 : 1; /* chars can take up to 3 bytes */ 270 multiplier = use_unicode ? 3 : 1; /* chars can take up to 3 bytes */
271 bp = kmalloc((sel_end-sel_start)/2*multiplier+1, GFP_KERNEL); 271 bp = kmalloc(((sel_end-sel_start)/2+1)*multiplier, GFP_KERNEL);
272 if (!bp) { 272 if (!bp) {
273 printk(KERN_WARNING "selection: kmalloc() failed\n"); 273 printk(KERN_WARNING "selection: kmalloc() failed\n");
274 clear_selection(); 274 clear_selection();
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index 9da581452874..6915fb82d0b0 100644
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -136,7 +136,7 @@ drm_gem_object_alloc(struct drm_device *dev, size_t size)
136 obj = kcalloc(1, sizeof(*obj), GFP_KERNEL); 136 obj = kcalloc(1, sizeof(*obj), GFP_KERNEL);
137 137
138 obj->dev = dev; 138 obj->dev = dev;
139 obj->filp = shmem_file_setup("drm mm object", size, 0); 139 obj->filp = shmem_file_setup("drm mm object", size, VM_NORESERVE);
140 if (IS_ERR(obj->filp)) { 140 if (IS_ERR(obj->filp)) {
141 kfree(obj); 141 kfree(obj);
142 return NULL; 142 return NULL;
diff --git a/drivers/net/igb/e1000_82575.c b/drivers/net/igb/e1000_82575.c
index f5e2e7235fcb..13ca73f96ec6 100644
--- a/drivers/net/igb/e1000_82575.c
+++ b/drivers/net/igb/e1000_82575.c
@@ -699,11 +699,18 @@ static s32 igb_check_for_link_82575(struct e1000_hw *hw)
699 699
700 /* SGMII link check is done through the PCS register. */ 700 /* SGMII link check is done through the PCS register. */
701 if ((hw->phy.media_type != e1000_media_type_copper) || 701 if ((hw->phy.media_type != e1000_media_type_copper) ||
702 (igb_sgmii_active_82575(hw))) 702 (igb_sgmii_active_82575(hw))) {
703 ret_val = igb_get_pcs_speed_and_duplex_82575(hw, &speed, 703 ret_val = igb_get_pcs_speed_and_duplex_82575(hw, &speed,
704 &duplex); 704 &duplex);
705 else 705 /*
706 * Use this flag to determine if link needs to be checked or
707 * not. If we have link clear the flag so that we do not
708 * continue to check for link.
709 */
710 hw->mac.get_link_status = !hw->mac.serdes_has_link;
711 } else {
706 ret_val = igb_check_for_copper_link(hw); 712 ret_val = igb_check_for_copper_link(hw);
713 }
707 714
708 return ret_val; 715 return ret_val;
709} 716}
diff --git a/drivers/net/igb/igb.h b/drivers/net/igb/igb.h
index 5a27825cc48a..aebef8e48e76 100644
--- a/drivers/net/igb/igb.h
+++ b/drivers/net/igb/igb.h
@@ -300,11 +300,10 @@ struct igb_adapter {
300 300
301#define IGB_FLAG_HAS_MSI (1 << 0) 301#define IGB_FLAG_HAS_MSI (1 << 0)
302#define IGB_FLAG_MSI_ENABLE (1 << 1) 302#define IGB_FLAG_MSI_ENABLE (1 << 1)
303#define IGB_FLAG_HAS_DCA (1 << 2) 303#define IGB_FLAG_DCA_ENABLED (1 << 2)
304#define IGB_FLAG_DCA_ENABLED (1 << 3) 304#define IGB_FLAG_IN_NETPOLL (1 << 3)
305#define IGB_FLAG_IN_NETPOLL (1 << 5) 305#define IGB_FLAG_QUAD_PORT_A (1 << 4)
306#define IGB_FLAG_QUAD_PORT_A (1 << 6) 306#define IGB_FLAG_NEED_CTX_IDX (1 << 5)
307#define IGB_FLAG_NEED_CTX_IDX (1 << 7)
308 307
309enum e1000_state_t { 308enum e1000_state_t {
310 __IGB_TESTING, 309 __IGB_TESTING,
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index b82b0fb2056c..a50db5398fa5 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -206,10 +206,11 @@ static int __init igb_init_module(void)
206 206
207 global_quad_port_a = 0; 207 global_quad_port_a = 0;
208 208
209 ret = pci_register_driver(&igb_driver);
210#ifdef CONFIG_IGB_DCA 209#ifdef CONFIG_IGB_DCA
211 dca_register_notify(&dca_notifier); 210 dca_register_notify(&dca_notifier);
212#endif 211#endif
212
213 ret = pci_register_driver(&igb_driver);
213 return ret; 214 return ret;
214} 215}
215 216
@@ -1156,11 +1157,10 @@ static int __devinit igb_probe(struct pci_dev *pdev,
1156 1157
1157 /* set flags */ 1158 /* set flags */
1158 switch (hw->mac.type) { 1159 switch (hw->mac.type) {
1159 case e1000_82576:
1160 case e1000_82575: 1160 case e1000_82575:
1161 adapter->flags |= IGB_FLAG_HAS_DCA;
1162 adapter->flags |= IGB_FLAG_NEED_CTX_IDX; 1161 adapter->flags |= IGB_FLAG_NEED_CTX_IDX;
1163 break; 1162 break;
1163 case e1000_82576:
1164 default: 1164 default:
1165 break; 1165 break;
1166 } 1166 }
@@ -1310,8 +1310,7 @@ static int __devinit igb_probe(struct pci_dev *pdev,
1310 goto err_register; 1310 goto err_register;
1311 1311
1312#ifdef CONFIG_IGB_DCA 1312#ifdef CONFIG_IGB_DCA
1313 if ((adapter->flags & IGB_FLAG_HAS_DCA) && 1313 if (dca_add_requester(&pdev->dev) == 0) {
1314 (dca_add_requester(&pdev->dev) == 0)) {
1315 adapter->flags |= IGB_FLAG_DCA_ENABLED; 1314 adapter->flags |= IGB_FLAG_DCA_ENABLED;
1316 dev_info(&pdev->dev, "DCA enabled\n"); 1315 dev_info(&pdev->dev, "DCA enabled\n");
1317 /* Always use CB2 mode, difference is masked 1316 /* Always use CB2 mode, difference is masked
@@ -1835,11 +1834,11 @@ static void igb_setup_rctl(struct igb_adapter *adapter)
1835 rctl |= E1000_RCTL_SECRC; 1834 rctl |= E1000_RCTL_SECRC;
1836 1835
1837 /* 1836 /*
1838 * disable store bad packets, long packet enable, and clear size bits. 1837 * disable store bad packets and clear size bits.
1839 */ 1838 */
1840 rctl &= ~(E1000_RCTL_SBP | E1000_RCTL_LPE | E1000_RCTL_SZ_256); 1839 rctl &= ~(E1000_RCTL_SBP | E1000_RCTL_SZ_256);
1841 1840
1842 if (adapter->netdev->mtu > ETH_DATA_LEN) 1841 /* enable LPE when to prevent packets larger than max_frame_size */
1843 rctl |= E1000_RCTL_LPE; 1842 rctl |= E1000_RCTL_LPE;
1844 1843
1845 /* Setup buffer sizes */ 1844 /* Setup buffer sizes */
@@ -1865,7 +1864,7 @@ static void igb_setup_rctl(struct igb_adapter *adapter)
1865 */ 1864 */
1866 /* allocations using alloc_page take too long for regular MTU 1865 /* allocations using alloc_page take too long for regular MTU
1867 * so only enable packet split for jumbo frames */ 1866 * so only enable packet split for jumbo frames */
1868 if (rctl & E1000_RCTL_LPE) { 1867 if (adapter->netdev->mtu > ETH_DATA_LEN) {
1869 adapter->rx_ps_hdr_size = IGB_RXBUFFER_128; 1868 adapter->rx_ps_hdr_size = IGB_RXBUFFER_128;
1870 srrctl |= adapter->rx_ps_hdr_size << 1869 srrctl |= adapter->rx_ps_hdr_size <<
1871 E1000_SRRCTL_BSIZEHDRSIZE_SHIFT; 1870 E1000_SRRCTL_BSIZEHDRSIZE_SHIFT;
@@ -3473,19 +3472,16 @@ static int __igb_notify_dca(struct device *dev, void *data)
3473 struct e1000_hw *hw = &adapter->hw; 3472 struct e1000_hw *hw = &adapter->hw;
3474 unsigned long event = *(unsigned long *)data; 3473 unsigned long event = *(unsigned long *)data;
3475 3474
3476 if (!(adapter->flags & IGB_FLAG_HAS_DCA))
3477 goto out;
3478
3479 switch (event) { 3475 switch (event) {
3480 case DCA_PROVIDER_ADD: 3476 case DCA_PROVIDER_ADD:
3481 /* if already enabled, don't do it again */ 3477 /* if already enabled, don't do it again */
3482 if (adapter->flags & IGB_FLAG_DCA_ENABLED) 3478 if (adapter->flags & IGB_FLAG_DCA_ENABLED)
3483 break; 3479 break;
3484 adapter->flags |= IGB_FLAG_DCA_ENABLED;
3485 /* Always use CB2 mode, difference is masked 3480 /* Always use CB2 mode, difference is masked
3486 * in the CB driver. */ 3481 * in the CB driver. */
3487 wr32(E1000_DCA_CTRL, 2); 3482 wr32(E1000_DCA_CTRL, 2);
3488 if (dca_add_requester(dev) == 0) { 3483 if (dca_add_requester(dev) == 0) {
3484 adapter->flags |= IGB_FLAG_DCA_ENABLED;
3489 dev_info(&adapter->pdev->dev, "DCA enabled\n"); 3485 dev_info(&adapter->pdev->dev, "DCA enabled\n");
3490 igb_setup_dca(adapter); 3486 igb_setup_dca(adapter);
3491 break; 3487 break;
@@ -3502,7 +3498,7 @@ static int __igb_notify_dca(struct device *dev, void *data)
3502 } 3498 }
3503 break; 3499 break;
3504 } 3500 }
3505out: 3501
3506 return 0; 3502 return 0;
3507} 3503}
3508 3504
diff --git a/drivers/net/sfc/efx.c b/drivers/net/sfc/efx.c
index 7673fd92eaf5..ab0e09bf154d 100644
--- a/drivers/net/sfc/efx.c
+++ b/drivers/net/sfc/efx.c
@@ -676,9 +676,8 @@ static int efx_init_port(struct efx_nic *efx)
676 rc = efx->phy_op->init(efx); 676 rc = efx->phy_op->init(efx);
677 if (rc) 677 if (rc)
678 return rc; 678 return rc;
679 efx->phy_op->reconfigure(efx);
680
681 mutex_lock(&efx->mac_lock); 679 mutex_lock(&efx->mac_lock);
680 efx->phy_op->reconfigure(efx);
682 rc = falcon_switch_mac(efx); 681 rc = falcon_switch_mac(efx);
683 mutex_unlock(&efx->mac_lock); 682 mutex_unlock(&efx->mac_lock);
684 if (rc) 683 if (rc)
@@ -686,7 +685,7 @@ static int efx_init_port(struct efx_nic *efx)
686 efx->mac_op->reconfigure(efx); 685 efx->mac_op->reconfigure(efx);
687 686
688 efx->port_initialized = true; 687 efx->port_initialized = true;
689 efx->stats_enabled = true; 688 efx_stats_enable(efx);
690 return 0; 689 return 0;
691 690
692fail: 691fail:
@@ -735,6 +734,7 @@ static void efx_fini_port(struct efx_nic *efx)
735 if (!efx->port_initialized) 734 if (!efx->port_initialized)
736 return; 735 return;
737 736
737 efx_stats_disable(efx);
738 efx->phy_op->fini(efx); 738 efx->phy_op->fini(efx);
739 efx->port_initialized = false; 739 efx->port_initialized = false;
740 740
@@ -1361,6 +1361,20 @@ static int efx_net_stop(struct net_device *net_dev)
1361 return 0; 1361 return 0;
1362} 1362}
1363 1363
1364void efx_stats_disable(struct efx_nic *efx)
1365{
1366 spin_lock(&efx->stats_lock);
1367 ++efx->stats_disable_count;
1368 spin_unlock(&efx->stats_lock);
1369}
1370
1371void efx_stats_enable(struct efx_nic *efx)
1372{
1373 spin_lock(&efx->stats_lock);
1374 --efx->stats_disable_count;
1375 spin_unlock(&efx->stats_lock);
1376}
1377
1364/* Context: process, dev_base_lock or RTNL held, non-blocking. */ 1378/* Context: process, dev_base_lock or RTNL held, non-blocking. */
1365static struct net_device_stats *efx_net_stats(struct net_device *net_dev) 1379static struct net_device_stats *efx_net_stats(struct net_device *net_dev)
1366{ 1380{
@@ -1369,12 +1383,12 @@ static struct net_device_stats *efx_net_stats(struct net_device *net_dev)
1369 struct net_device_stats *stats = &net_dev->stats; 1383 struct net_device_stats *stats = &net_dev->stats;
1370 1384
1371 /* Update stats if possible, but do not wait if another thread 1385 /* Update stats if possible, but do not wait if another thread
1372 * is updating them (or resetting the NIC); slightly stale 1386 * is updating them or if MAC stats fetches are temporarily
1373 * stats are acceptable. 1387 * disabled; slightly stale stats are acceptable.
1374 */ 1388 */
1375 if (!spin_trylock(&efx->stats_lock)) 1389 if (!spin_trylock(&efx->stats_lock))
1376 return stats; 1390 return stats;
1377 if (efx->stats_enabled) { 1391 if (!efx->stats_disable_count) {
1378 efx->mac_op->update_stats(efx); 1392 efx->mac_op->update_stats(efx);
1379 falcon_update_nic_stats(efx); 1393 falcon_update_nic_stats(efx);
1380 } 1394 }
@@ -1622,16 +1636,12 @@ static void efx_unregister_netdev(struct efx_nic *efx)
1622 1636
1623/* Tears down the entire software state and most of the hardware state 1637/* Tears down the entire software state and most of the hardware state
1624 * before reset. */ 1638 * before reset. */
1625void efx_reset_down(struct efx_nic *efx, struct ethtool_cmd *ecmd) 1639void efx_reset_down(struct efx_nic *efx, enum reset_type method,
1640 struct ethtool_cmd *ecmd)
1626{ 1641{
1627 EFX_ASSERT_RESET_SERIALISED(efx); 1642 EFX_ASSERT_RESET_SERIALISED(efx);
1628 1643
1629 /* The net_dev->get_stats handler is quite slow, and will fail 1644 efx_stats_disable(efx);
1630 * if a fetch is pending over reset. Serialise against it. */
1631 spin_lock(&efx->stats_lock);
1632 efx->stats_enabled = false;
1633 spin_unlock(&efx->stats_lock);
1634
1635 efx_stop_all(efx); 1645 efx_stop_all(efx);
1636 mutex_lock(&efx->mac_lock); 1646 mutex_lock(&efx->mac_lock);
1637 mutex_lock(&efx->spi_lock); 1647 mutex_lock(&efx->spi_lock);
@@ -1639,6 +1649,8 @@ void efx_reset_down(struct efx_nic *efx, struct ethtool_cmd *ecmd)
1639 efx->phy_op->get_settings(efx, ecmd); 1649 efx->phy_op->get_settings(efx, ecmd);
1640 1650
1641 efx_fini_channels(efx); 1651 efx_fini_channels(efx);
1652 if (efx->port_initialized && method != RESET_TYPE_INVISIBLE)
1653 efx->phy_op->fini(efx);
1642} 1654}
1643 1655
1644/* This function will always ensure that the locks acquired in 1656/* This function will always ensure that the locks acquired in
@@ -1646,7 +1658,8 @@ void efx_reset_down(struct efx_nic *efx, struct ethtool_cmd *ecmd)
1646 * that we were unable to reinitialise the hardware, and the 1658 * that we were unable to reinitialise the hardware, and the
1647 * driver should be disabled. If ok is false, then the rx and tx 1659 * driver should be disabled. If ok is false, then the rx and tx
1648 * engines are not restarted, pending a RESET_DISABLE. */ 1660 * engines are not restarted, pending a RESET_DISABLE. */
1649int efx_reset_up(struct efx_nic *efx, struct ethtool_cmd *ecmd, bool ok) 1661int efx_reset_up(struct efx_nic *efx, enum reset_type method,
1662 struct ethtool_cmd *ecmd, bool ok)
1650{ 1663{
1651 int rc; 1664 int rc;
1652 1665
@@ -1658,6 +1671,15 @@ int efx_reset_up(struct efx_nic *efx, struct ethtool_cmd *ecmd, bool ok)
1658 ok = false; 1671 ok = false;
1659 } 1672 }
1660 1673
1674 if (efx->port_initialized && method != RESET_TYPE_INVISIBLE) {
1675 if (ok) {
1676 rc = efx->phy_op->init(efx);
1677 if (rc)
1678 ok = false;
1679 } else
1680 efx->port_initialized = false;
1681 }
1682
1661 if (ok) { 1683 if (ok) {
1662 efx_init_channels(efx); 1684 efx_init_channels(efx);
1663 1685
@@ -1670,7 +1692,7 @@ int efx_reset_up(struct efx_nic *efx, struct ethtool_cmd *ecmd, bool ok)
1670 1692
1671 if (ok) { 1693 if (ok) {
1672 efx_start_all(efx); 1694 efx_start_all(efx);
1673 efx->stats_enabled = true; 1695 efx_stats_enable(efx);
1674 } 1696 }
1675 return rc; 1697 return rc;
1676} 1698}
@@ -1702,7 +1724,7 @@ static int efx_reset(struct efx_nic *efx)
1702 1724
1703 EFX_INFO(efx, "resetting (%d)\n", method); 1725 EFX_INFO(efx, "resetting (%d)\n", method);
1704 1726
1705 efx_reset_down(efx, &ecmd); 1727 efx_reset_down(efx, method, &ecmd);
1706 1728
1707 rc = falcon_reset_hw(efx, method); 1729 rc = falcon_reset_hw(efx, method);
1708 if (rc) { 1730 if (rc) {
@@ -1721,10 +1743,10 @@ static int efx_reset(struct efx_nic *efx)
1721 1743
1722 /* Leave device stopped if necessary */ 1744 /* Leave device stopped if necessary */
1723 if (method == RESET_TYPE_DISABLE) { 1745 if (method == RESET_TYPE_DISABLE) {
1724 efx_reset_up(efx, &ecmd, false); 1746 efx_reset_up(efx, method, &ecmd, false);
1725 rc = -EIO; 1747 rc = -EIO;
1726 } else { 1748 } else {
1727 rc = efx_reset_up(efx, &ecmd, true); 1749 rc = efx_reset_up(efx, method, &ecmd, true);
1728 } 1750 }
1729 1751
1730out_disable: 1752out_disable:
@@ -1876,6 +1898,7 @@ static int efx_init_struct(struct efx_nic *efx, struct efx_nic_type *type,
1876 efx->rx_checksum_enabled = true; 1898 efx->rx_checksum_enabled = true;
1877 spin_lock_init(&efx->netif_stop_lock); 1899 spin_lock_init(&efx->netif_stop_lock);
1878 spin_lock_init(&efx->stats_lock); 1900 spin_lock_init(&efx->stats_lock);
1901 efx->stats_disable_count = 1;
1879 mutex_init(&efx->mac_lock); 1902 mutex_init(&efx->mac_lock);
1880 efx->mac_op = &efx_dummy_mac_operations; 1903 efx->mac_op = &efx_dummy_mac_operations;
1881 efx->phy_op = &efx_dummy_phy_operations; 1904 efx->phy_op = &efx_dummy_phy_operations;
diff --git a/drivers/net/sfc/efx.h b/drivers/net/sfc/efx.h
index 0dd7a532c78a..55d0f131b0e9 100644
--- a/drivers/net/sfc/efx.h
+++ b/drivers/net/sfc/efx.h
@@ -36,13 +36,16 @@ extern void efx_process_channel_now(struct efx_channel *channel);
36extern void efx_flush_queues(struct efx_nic *efx); 36extern void efx_flush_queues(struct efx_nic *efx);
37 37
38/* Ports */ 38/* Ports */
39extern void efx_stats_disable(struct efx_nic *efx);
40extern void efx_stats_enable(struct efx_nic *efx);
39extern void efx_reconfigure_port(struct efx_nic *efx); 41extern void efx_reconfigure_port(struct efx_nic *efx);
40extern void __efx_reconfigure_port(struct efx_nic *efx); 42extern void __efx_reconfigure_port(struct efx_nic *efx);
41 43
42/* Reset handling */ 44/* Reset handling */
43extern void efx_reset_down(struct efx_nic *efx, struct ethtool_cmd *ecmd); 45extern void efx_reset_down(struct efx_nic *efx, enum reset_type method,
44extern int efx_reset_up(struct efx_nic *efx, struct ethtool_cmd *ecmd, 46 struct ethtool_cmd *ecmd);
45 bool ok); 47extern int efx_reset_up(struct efx_nic *efx, enum reset_type method,
48 struct ethtool_cmd *ecmd, bool ok);
46 49
47/* Global */ 50/* Global */
48extern void efx_schedule_reset(struct efx_nic *efx, enum reset_type type); 51extern void efx_schedule_reset(struct efx_nic *efx, enum reset_type type);
diff --git a/drivers/net/sfc/ethtool.c b/drivers/net/sfc/ethtool.c
index 53d259e90187..7b5924c039b3 100644
--- a/drivers/net/sfc/ethtool.c
+++ b/drivers/net/sfc/ethtool.c
@@ -219,9 +219,6 @@ int efx_ethtool_set_settings(struct net_device *net_dev,
219 struct efx_nic *efx = netdev_priv(net_dev); 219 struct efx_nic *efx = netdev_priv(net_dev);
220 int rc; 220 int rc;
221 221
222 if (EFX_WORKAROUND_13963(efx) && !ecmd->autoneg)
223 return -EINVAL;
224
225 /* Falcon GMAC does not support 1000Mbps HD */ 222 /* Falcon GMAC does not support 1000Mbps HD */
226 if (ecmd->speed == SPEED_1000 && ecmd->duplex != DUPLEX_FULL) { 223 if (ecmd->speed == SPEED_1000 && ecmd->duplex != DUPLEX_FULL) {
227 EFX_LOG(efx, "rejecting unsupported 1000Mbps HD" 224 EFX_LOG(efx, "rejecting unsupported 1000Mbps HD"
diff --git a/drivers/net/sfc/falcon.c b/drivers/net/sfc/falcon.c
index 5b9f2d9cc4ed..d5378e60fcdd 100644
--- a/drivers/net/sfc/falcon.c
+++ b/drivers/net/sfc/falcon.c
@@ -824,10 +824,6 @@ static void falcon_handle_rx_not_ok(struct efx_rx_queue *rx_queue,
824 rx_ev_pause_frm ? " [PAUSE]" : ""); 824 rx_ev_pause_frm ? " [PAUSE]" : "");
825 } 825 }
826#endif 826#endif
827
828 if (unlikely(rx_ev_eth_crc_err && EFX_WORKAROUND_10750(efx) &&
829 efx->phy_type == PHY_TYPE_SFX7101))
830 tenxpress_crc_err(efx);
831} 827}
832 828
833/* Handle receive events that are not in-order. */ 829/* Handle receive events that are not in-order. */
@@ -1887,7 +1883,7 @@ static int falcon_reset_macs(struct efx_nic *efx)
1887 1883
1888 /* MAC stats will fail whilst the TX fifo is draining. Serialise 1884 /* MAC stats will fail whilst the TX fifo is draining. Serialise
1889 * the drain sequence with the statistics fetch */ 1885 * the drain sequence with the statistics fetch */
1890 spin_lock(&efx->stats_lock); 1886 efx_stats_disable(efx);
1891 1887
1892 falcon_read(efx, &reg, MAC0_CTRL_REG_KER); 1888 falcon_read(efx, &reg, MAC0_CTRL_REG_KER);
1893 EFX_SET_OWORD_FIELD(reg, TXFIFO_DRAIN_EN_B0, 1); 1889 EFX_SET_OWORD_FIELD(reg, TXFIFO_DRAIN_EN_B0, 1);
@@ -1917,7 +1913,7 @@ static int falcon_reset_macs(struct efx_nic *efx)
1917 udelay(10); 1913 udelay(10);
1918 } 1914 }
1919 1915
1920 spin_unlock(&efx->stats_lock); 1916 efx_stats_enable(efx);
1921 1917
1922 /* If we've reset the EM block and the link is up, then 1918 /* If we've reset the EM block and the link is up, then
1923 * we'll have to kick the XAUI link so the PHY can recover */ 1919 * we'll have to kick the XAUI link so the PHY can recover */
@@ -2277,6 +2273,10 @@ int falcon_switch_mac(struct efx_nic *efx)
2277 struct efx_mac_operations *old_mac_op = efx->mac_op; 2273 struct efx_mac_operations *old_mac_op = efx->mac_op;
2278 efx_oword_t nic_stat; 2274 efx_oword_t nic_stat;
2279 unsigned strap_val; 2275 unsigned strap_val;
2276 int rc = 0;
2277
2278 /* Don't try to fetch MAC stats while we're switching MACs */
2279 efx_stats_disable(efx);
2280 2280
2281 /* Internal loopbacks override the phy speed setting */ 2281 /* Internal loopbacks override the phy speed setting */
2282 if (efx->loopback_mode == LOOPBACK_GMAC) { 2282 if (efx->loopback_mode == LOOPBACK_GMAC) {
@@ -2287,16 +2287,12 @@ int falcon_switch_mac(struct efx_nic *efx)
2287 efx->link_fd = true; 2287 efx->link_fd = true;
2288 } 2288 }
2289 2289
2290 WARN_ON(!mutex_is_locked(&efx->mac_lock));
2290 efx->mac_op = (EFX_IS10G(efx) ? 2291 efx->mac_op = (EFX_IS10G(efx) ?
2291 &falcon_xmac_operations : &falcon_gmac_operations); 2292 &falcon_xmac_operations : &falcon_gmac_operations);
2292 if (old_mac_op == efx->mac_op)
2293 return 0;
2294
2295 WARN_ON(!mutex_is_locked(&efx->mac_lock));
2296
2297 /* Not all macs support a mac-level link state */
2298 efx->mac_up = true;
2299 2293
2294 /* Always push the NIC_STAT_REG setting even if the mac hasn't
2295 * changed, because this function is run post online reset */
2300 falcon_read(efx, &nic_stat, NIC_STAT_REG); 2296 falcon_read(efx, &nic_stat, NIC_STAT_REG);
2301 strap_val = EFX_IS10G(efx) ? 5 : 3; 2297 strap_val = EFX_IS10G(efx) ? 5 : 3;
2302 if (falcon_rev(efx) >= FALCON_REV_B0) { 2298 if (falcon_rev(efx) >= FALCON_REV_B0) {
@@ -2309,9 +2305,17 @@ int falcon_switch_mac(struct efx_nic *efx)
2309 BUG_ON(EFX_OWORD_FIELD(nic_stat, STRAP_PINS) != strap_val); 2305 BUG_ON(EFX_OWORD_FIELD(nic_stat, STRAP_PINS) != strap_val);
2310 } 2306 }
2311 2307
2308 if (old_mac_op == efx->mac_op)
2309 goto out;
2312 2310
2313 EFX_LOG(efx, "selected %cMAC\n", EFX_IS10G(efx) ? 'X' : 'G'); 2311 EFX_LOG(efx, "selected %cMAC\n", EFX_IS10G(efx) ? 'X' : 'G');
2314 return falcon_reset_macs(efx); 2312 /* Not all macs support a mac-level link state */
2313 efx->mac_up = true;
2314
2315 rc = falcon_reset_macs(efx);
2316out:
2317 efx_stats_enable(efx);
2318 return rc;
2315} 2319}
2316 2320
2317/* This call is responsible for hooking in the MAC and PHY operations */ 2321/* This call is responsible for hooking in the MAC and PHY operations */
diff --git a/drivers/net/sfc/mdio_10g.c b/drivers/net/sfc/mdio_10g.c
index f6a16428113d..f9e2f95c3b48 100644
--- a/drivers/net/sfc/mdio_10g.c
+++ b/drivers/net/sfc/mdio_10g.c
@@ -15,6 +15,7 @@
15#include "net_driver.h" 15#include "net_driver.h"
16#include "mdio_10g.h" 16#include "mdio_10g.h"
17#include "boards.h" 17#include "boards.h"
18#include "workarounds.h"
18 19
19int mdio_clause45_reset_mmd(struct efx_nic *port, int mmd, 20int mdio_clause45_reset_mmd(struct efx_nic *port, int mmd,
20 int spins, int spintime) 21 int spins, int spintime)
@@ -179,17 +180,12 @@ bool mdio_clause45_links_ok(struct efx_nic *efx, unsigned int mmd_mask)
179 return false; 180 return false;
180 else if (efx_phy_mode_disabled(efx->phy_mode)) 181 else if (efx_phy_mode_disabled(efx->phy_mode))
181 return false; 182 return false;
182 else if (efx->loopback_mode == LOOPBACK_PHYXS) { 183 else if (efx->loopback_mode == LOOPBACK_PHYXS)
183 mmd_mask &= ~(MDIO_MMDREG_DEVS_PHYXS | 184 mmd_mask &= ~(MDIO_MMDREG_DEVS_PHYXS |
184 MDIO_MMDREG_DEVS_PCS | 185 MDIO_MMDREG_DEVS_PCS |
185 MDIO_MMDREG_DEVS_PMAPMD | 186 MDIO_MMDREG_DEVS_PMAPMD |
186 MDIO_MMDREG_DEVS_AN); 187 MDIO_MMDREG_DEVS_AN);
187 if (!mmd_mask) { 188 else if (efx->loopback_mode == LOOPBACK_PCS)
188 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_PHYXS,
189 MDIO_PHYXS_STATUS2);
190 return !(reg & (1 << MDIO_PHYXS_STATUS2_RX_FAULT_LBN));
191 }
192 } else if (efx->loopback_mode == LOOPBACK_PCS)
193 mmd_mask &= ~(MDIO_MMDREG_DEVS_PCS | 189 mmd_mask &= ~(MDIO_MMDREG_DEVS_PCS |
194 MDIO_MMDREG_DEVS_PMAPMD | 190 MDIO_MMDREG_DEVS_PMAPMD |
195 MDIO_MMDREG_DEVS_AN); 191 MDIO_MMDREG_DEVS_AN);
@@ -197,6 +193,13 @@ bool mdio_clause45_links_ok(struct efx_nic *efx, unsigned int mmd_mask)
197 mmd_mask &= ~(MDIO_MMDREG_DEVS_PMAPMD | 193 mmd_mask &= ~(MDIO_MMDREG_DEVS_PMAPMD |
198 MDIO_MMDREG_DEVS_AN); 194 MDIO_MMDREG_DEVS_AN);
199 195
196 if (!mmd_mask) {
197 /* Use presence of XGMII faults in leui of link state */
198 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_PHYXS,
199 MDIO_PHYXS_STATUS2);
200 return !(reg & (1 << MDIO_PHYXS_STATUS2_RX_FAULT_LBN));
201 }
202
200 while (mmd_mask) { 203 while (mmd_mask) {
201 if (mmd_mask & 1) { 204 if (mmd_mask & 1) {
202 /* Double reads because link state is latched, and a 205 /* Double reads because link state is latched, and a
@@ -263,7 +266,7 @@ void mdio_clause45_set_mmds_lpower(struct efx_nic *efx,
263 } 266 }
264} 267}
265 268
266static u32 mdio_clause45_get_an(struct efx_nic *efx, u16 addr, u32 xnp) 269static u32 mdio_clause45_get_an(struct efx_nic *efx, u16 addr)
267{ 270{
268 int phy_id = efx->mii.phy_id; 271 int phy_id = efx->mii.phy_id;
269 u32 result = 0; 272 u32 result = 0;
@@ -278,9 +281,6 @@ static u32 mdio_clause45_get_an(struct efx_nic *efx, u16 addr, u32 xnp)
278 result |= ADVERTISED_100baseT_Half; 281 result |= ADVERTISED_100baseT_Half;
279 if (reg & ADVERTISE_100FULL) 282 if (reg & ADVERTISE_100FULL)
280 result |= ADVERTISED_100baseT_Full; 283 result |= ADVERTISED_100baseT_Full;
281 if (reg & LPA_RESV)
282 result |= xnp;
283
284 return result; 284 return result;
285} 285}
286 286
@@ -310,7 +310,7 @@ void mdio_clause45_get_settings(struct efx_nic *efx,
310 */ 310 */
311void mdio_clause45_get_settings_ext(struct efx_nic *efx, 311void mdio_clause45_get_settings_ext(struct efx_nic *efx,
312 struct ethtool_cmd *ecmd, 312 struct ethtool_cmd *ecmd,
313 u32 xnp, u32 xnp_lpa) 313 u32 npage_adv, u32 npage_lpa)
314{ 314{
315 int phy_id = efx->mii.phy_id; 315 int phy_id = efx->mii.phy_id;
316 int reg; 316 int reg;
@@ -361,8 +361,8 @@ void mdio_clause45_get_settings_ext(struct efx_nic *efx,
361 ecmd->autoneg = AUTONEG_ENABLE; 361 ecmd->autoneg = AUTONEG_ENABLE;
362 ecmd->advertising |= 362 ecmd->advertising |=
363 ADVERTISED_Autoneg | 363 ADVERTISED_Autoneg |
364 mdio_clause45_get_an(efx, 364 mdio_clause45_get_an(efx, MDIO_AN_ADVERTISE) |
365 MDIO_AN_ADVERTISE, xnp); 365 npage_adv;
366 } else 366 } else
367 ecmd->autoneg = AUTONEG_DISABLE; 367 ecmd->autoneg = AUTONEG_DISABLE;
368 } else 368 } else
@@ -371,27 +371,30 @@ void mdio_clause45_get_settings_ext(struct efx_nic *efx,
371 if (ecmd->autoneg) { 371 if (ecmd->autoneg) {
372 /* If AN is complete, report best common mode, 372 /* If AN is complete, report best common mode,
373 * otherwise report best advertised mode. */ 373 * otherwise report best advertised mode. */
374 u32 common = ecmd->advertising; 374 u32 modes = 0;
375 if (mdio_clause45_read(efx, phy_id, MDIO_MMD_AN, 375 if (mdio_clause45_read(efx, phy_id, MDIO_MMD_AN,
376 MDIO_MMDREG_STAT1) & 376 MDIO_MMDREG_STAT1) &
377 (1 << MDIO_AN_STATUS_AN_DONE_LBN)) { 377 (1 << MDIO_AN_STATUS_AN_DONE_LBN))
378 common &= mdio_clause45_get_an(efx, MDIO_AN_LPA, 378 modes = (ecmd->advertising &
379 xnp_lpa); 379 (mdio_clause45_get_an(efx, MDIO_AN_LPA) |
380 } 380 npage_lpa));
381 if (common & ADVERTISED_10000baseT_Full) { 381 if (modes == 0)
382 modes = ecmd->advertising;
383
384 if (modes & ADVERTISED_10000baseT_Full) {
382 ecmd->speed = SPEED_10000; 385 ecmd->speed = SPEED_10000;
383 ecmd->duplex = DUPLEX_FULL; 386 ecmd->duplex = DUPLEX_FULL;
384 } else if (common & (ADVERTISED_1000baseT_Full | 387 } else if (modes & (ADVERTISED_1000baseT_Full |
385 ADVERTISED_1000baseT_Half)) { 388 ADVERTISED_1000baseT_Half)) {
386 ecmd->speed = SPEED_1000; 389 ecmd->speed = SPEED_1000;
387 ecmd->duplex = !!(common & ADVERTISED_1000baseT_Full); 390 ecmd->duplex = !!(modes & ADVERTISED_1000baseT_Full);
388 } else if (common & (ADVERTISED_100baseT_Full | 391 } else if (modes & (ADVERTISED_100baseT_Full |
389 ADVERTISED_100baseT_Half)) { 392 ADVERTISED_100baseT_Half)) {
390 ecmd->speed = SPEED_100; 393 ecmd->speed = SPEED_100;
391 ecmd->duplex = !!(common & ADVERTISED_100baseT_Full); 394 ecmd->duplex = !!(modes & ADVERTISED_100baseT_Full);
392 } else { 395 } else {
393 ecmd->speed = SPEED_10; 396 ecmd->speed = SPEED_10;
394 ecmd->duplex = !!(common & ADVERTISED_10baseT_Full); 397 ecmd->duplex = !!(modes & ADVERTISED_10baseT_Full);
395 } 398 }
396 } else { 399 } else {
397 /* Report forced settings */ 400 /* Report forced settings */
@@ -415,7 +418,7 @@ int mdio_clause45_set_settings(struct efx_nic *efx,
415 int phy_id = efx->mii.phy_id; 418 int phy_id = efx->mii.phy_id;
416 struct ethtool_cmd prev; 419 struct ethtool_cmd prev;
417 u32 required; 420 u32 required;
418 int ctrl1_bits, reg; 421 int reg;
419 422
420 efx->phy_op->get_settings(efx, &prev); 423 efx->phy_op->get_settings(efx, &prev);
421 424
@@ -430,99 +433,83 @@ int mdio_clause45_set_settings(struct efx_nic *efx,
430 if (prev.port != PORT_TP || ecmd->port != PORT_TP) 433 if (prev.port != PORT_TP || ecmd->port != PORT_TP)
431 return -EINVAL; 434 return -EINVAL;
432 435
433 /* Check that PHY supports these settings and work out the 436 /* Check that PHY supports these settings */
434 * basic control bits */ 437 if (ecmd->autoneg) {
435 if (ecmd->duplex) { 438 required = SUPPORTED_Autoneg;
439 } else if (ecmd->duplex) {
436 switch (ecmd->speed) { 440 switch (ecmd->speed) {
437 case SPEED_10: 441 case SPEED_10: required = SUPPORTED_10baseT_Full; break;
438 ctrl1_bits = BMCR_FULLDPLX; 442 case SPEED_100: required = SUPPORTED_100baseT_Full; break;
439 required = SUPPORTED_10baseT_Full; 443 default: return -EINVAL;
440 break;
441 case SPEED_100:
442 ctrl1_bits = BMCR_SPEED100 | BMCR_FULLDPLX;
443 required = SUPPORTED_100baseT_Full;
444 break;
445 case SPEED_1000:
446 ctrl1_bits = BMCR_SPEED1000 | BMCR_FULLDPLX;
447 required = SUPPORTED_1000baseT_Full;
448 break;
449 case SPEED_10000:
450 ctrl1_bits = (BMCR_SPEED1000 | BMCR_SPEED100 |
451 BMCR_FULLDPLX);
452 required = SUPPORTED_10000baseT_Full;
453 break;
454 default:
455 return -EINVAL;
456 } 444 }
457 } else { 445 } else {
458 switch (ecmd->speed) { 446 switch (ecmd->speed) {
459 case SPEED_10: 447 case SPEED_10: required = SUPPORTED_10baseT_Half; break;
460 ctrl1_bits = 0; 448 case SPEED_100: required = SUPPORTED_100baseT_Half; break;
461 required = SUPPORTED_10baseT_Half; 449 default: return -EINVAL;
462 break;
463 case SPEED_100:
464 ctrl1_bits = BMCR_SPEED100;
465 required = SUPPORTED_100baseT_Half;
466 break;
467 case SPEED_1000:
468 ctrl1_bits = BMCR_SPEED1000;
469 required = SUPPORTED_1000baseT_Half;
470 break;
471 default:
472 return -EINVAL;
473 } 450 }
474 } 451 }
475 if (ecmd->autoneg)
476 required |= SUPPORTED_Autoneg;
477 required |= ecmd->advertising; 452 required |= ecmd->advertising;
478 if (required & ~prev.supported) 453 if (required & ~prev.supported)
479 return -EINVAL; 454 return -EINVAL;
480 455
481 /* Set the basic control bits */ 456 if (ecmd->autoneg) {
482 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_PMAPMD, 457 bool xnp = (ecmd->advertising & ADVERTISED_10000baseT_Full
483 MDIO_MMDREG_CTRL1); 458 || EFX_WORKAROUND_13204(efx));
484 reg &= ~(BMCR_SPEED1000 | BMCR_SPEED100 | BMCR_FULLDPLX | 0x003c); 459
485 reg |= ctrl1_bits; 460 /* Set up the base page */
486 mdio_clause45_write(efx, phy_id, MDIO_MMD_PMAPMD, MDIO_MMDREG_CTRL1, 461 reg = ADVERTISE_CSMA;
487 reg); 462 if (ecmd->advertising & ADVERTISED_10baseT_Half)
488 463 reg |= ADVERTISE_10HALF;
489 /* Set the AN registers */ 464 if (ecmd->advertising & ADVERTISED_10baseT_Full)
490 if (ecmd->autoneg != prev.autoneg || 465 reg |= ADVERTISE_10FULL;
491 ecmd->advertising != prev.advertising) { 466 if (ecmd->advertising & ADVERTISED_100baseT_Half)
492 bool xnp = false; 467 reg |= ADVERTISE_100HALF;
493 468 if (ecmd->advertising & ADVERTISED_100baseT_Full)
494 if (efx->phy_op->set_xnp_advertise) 469 reg |= ADVERTISE_100FULL;
495 xnp = efx->phy_op->set_xnp_advertise(efx, 470 if (xnp)
496 ecmd->advertising); 471 reg |= ADVERTISE_RESV;
497 472 else if (ecmd->advertising & (ADVERTISED_1000baseT_Half |
498 if (ecmd->autoneg) { 473 ADVERTISED_1000baseT_Full))
499 reg = 0; 474 reg |= ADVERTISE_NPAGE;
500 if (ecmd->advertising & ADVERTISED_10baseT_Half) 475 reg |= efx_fc_advertise(efx->wanted_fc);
501 reg |= ADVERTISE_10HALF; 476 mdio_clause45_write(efx, phy_id, MDIO_MMD_AN,
502 if (ecmd->advertising & ADVERTISED_10baseT_Full) 477 MDIO_AN_ADVERTISE, reg);
503 reg |= ADVERTISE_10FULL; 478
504 if (ecmd->advertising & ADVERTISED_100baseT_Half) 479 /* Set up the (extended) next page if necessary */
505 reg |= ADVERTISE_100HALF; 480 if (efx->phy_op->set_npage_adv)
506 if (ecmd->advertising & ADVERTISED_100baseT_Full) 481 efx->phy_op->set_npage_adv(efx, ecmd->advertising);
507 reg |= ADVERTISE_100FULL;
508 if (xnp)
509 reg |= ADVERTISE_RESV;
510 mdio_clause45_write(efx, phy_id, MDIO_MMD_AN,
511 MDIO_AN_ADVERTISE, reg);
512 }
513 482
483 /* Enable and restart AN */
514 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_AN, 484 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_AN,
515 MDIO_MMDREG_CTRL1); 485 MDIO_MMDREG_CTRL1);
516 if (ecmd->autoneg) 486 reg |= BMCR_ANENABLE;
517 reg |= BMCR_ANENABLE | BMCR_ANRESTART; 487 if (!(EFX_WORKAROUND_15195(efx) &&
518 else 488 LOOPBACK_MASK(efx) & efx->phy_op->loopbacks))
519 reg &= ~BMCR_ANENABLE; 489 reg |= BMCR_ANRESTART;
520 if (xnp) 490 if (xnp)
521 reg |= 1 << MDIO_AN_CTRL_XNP_LBN; 491 reg |= 1 << MDIO_AN_CTRL_XNP_LBN;
522 else 492 else
523 reg &= ~(1 << MDIO_AN_CTRL_XNP_LBN); 493 reg &= ~(1 << MDIO_AN_CTRL_XNP_LBN);
524 mdio_clause45_write(efx, phy_id, MDIO_MMD_AN, 494 mdio_clause45_write(efx, phy_id, MDIO_MMD_AN,
525 MDIO_MMDREG_CTRL1, reg); 495 MDIO_MMDREG_CTRL1, reg);
496 } else {
497 /* Disable AN */
498 mdio_clause45_set_flag(efx, phy_id, MDIO_MMD_AN,
499 MDIO_MMDREG_CTRL1,
500 __ffs(BMCR_ANENABLE), false);
501
502 /* Set the basic control bits */
503 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_PMAPMD,
504 MDIO_MMDREG_CTRL1);
505 reg &= ~(BMCR_SPEED1000 | BMCR_SPEED100 | BMCR_FULLDPLX |
506 0x003c);
507 if (ecmd->speed == SPEED_100)
508 reg |= BMCR_SPEED100;
509 if (ecmd->duplex)
510 reg |= BMCR_FULLDPLX;
511 mdio_clause45_write(efx, phy_id, MDIO_MMD_PMAPMD,
512 MDIO_MMDREG_CTRL1, reg);
526 } 513 }
527 514
528 return 0; 515 return 0;
diff --git a/drivers/net/sfc/mdio_10g.h b/drivers/net/sfc/mdio_10g.h
index 09bf801d0569..8ba49773ce7e 100644
--- a/drivers/net/sfc/mdio_10g.h
+++ b/drivers/net/sfc/mdio_10g.h
@@ -155,7 +155,8 @@
155#define MDIO_AN_XNP 22 155#define MDIO_AN_XNP 22
156#define MDIO_AN_LPA_XNP 25 156#define MDIO_AN_LPA_XNP 25
157 157
158#define MDIO_AN_10GBT_ADVERTISE 32 158#define MDIO_AN_10GBT_CTRL 32
159#define MDIO_AN_10GBT_CTRL_ADV_10G_LBN 12
159#define MDIO_AN_10GBT_STATUS (33) 160#define MDIO_AN_10GBT_STATUS (33)
160#define MDIO_AN_10GBT_STATUS_MS_FLT_LBN (15) /* MASTER/SLAVE config fault */ 161#define MDIO_AN_10GBT_STATUS_MS_FLT_LBN (15) /* MASTER/SLAVE config fault */
161#define MDIO_AN_10GBT_STATUS_MS_LBN (14) /* MASTER/SLAVE config */ 162#define MDIO_AN_10GBT_STATUS_MS_LBN (14) /* MASTER/SLAVE config */
diff --git a/drivers/net/sfc/net_driver.h b/drivers/net/sfc/net_driver.h
index 5f255f75754e..e019ad1fb9a0 100644
--- a/drivers/net/sfc/net_driver.h
+++ b/drivers/net/sfc/net_driver.h
@@ -566,7 +566,7 @@ struct efx_mac_operations {
566 * @poll: Poll for hardware state. Serialised by the mac_lock. 566 * @poll: Poll for hardware state. Serialised by the mac_lock.
567 * @get_settings: Get ethtool settings. Serialised by the mac_lock. 567 * @get_settings: Get ethtool settings. Serialised by the mac_lock.
568 * @set_settings: Set ethtool settings. Serialised by the mac_lock. 568 * @set_settings: Set ethtool settings. Serialised by the mac_lock.
569 * @set_xnp_advertise: Set abilities advertised in Extended Next Page 569 * @set_npage_adv: Set abilities advertised in (Extended) Next Page
570 * (only needed where AN bit is set in mmds) 570 * (only needed where AN bit is set in mmds)
571 * @num_tests: Number of PHY-specific tests/results 571 * @num_tests: Number of PHY-specific tests/results
572 * @test_names: Names of the tests/results 572 * @test_names: Names of the tests/results
@@ -586,7 +586,7 @@ struct efx_phy_operations {
586 struct ethtool_cmd *ecmd); 586 struct ethtool_cmd *ecmd);
587 int (*set_settings) (struct efx_nic *efx, 587 int (*set_settings) (struct efx_nic *efx,
588 struct ethtool_cmd *ecmd); 588 struct ethtool_cmd *ecmd);
589 bool (*set_xnp_advertise) (struct efx_nic *efx, u32); 589 void (*set_npage_adv) (struct efx_nic *efx, u32);
590 u32 num_tests; 590 u32 num_tests;
591 const char *const *test_names; 591 const char *const *test_names;
592 int (*run_tests) (struct efx_nic *efx, int *results, unsigned flags); 592 int (*run_tests) (struct efx_nic *efx, int *results, unsigned flags);
@@ -754,8 +754,7 @@ union efx_multicast_hash {
754 * &struct net_device_stats. 754 * &struct net_device_stats.
755 * @stats_buffer: DMA buffer for statistics 755 * @stats_buffer: DMA buffer for statistics
756 * @stats_lock: Statistics update lock. Serialises statistics fetches 756 * @stats_lock: Statistics update lock. Serialises statistics fetches
757 * @stats_enabled: Temporarily disable statistics fetches. 757 * @stats_disable_count: Nest count for disabling statistics fetches
758 * Serialised by @stats_lock
759 * @mac_op: MAC interface 758 * @mac_op: MAC interface
760 * @mac_address: Permanent MAC address 759 * @mac_address: Permanent MAC address
761 * @phy_type: PHY type 760 * @phy_type: PHY type
@@ -837,7 +836,7 @@ struct efx_nic {
837 struct efx_mac_stats mac_stats; 836 struct efx_mac_stats mac_stats;
838 struct efx_buffer stats_buffer; 837 struct efx_buffer stats_buffer;
839 spinlock_t stats_lock; 838 spinlock_t stats_lock;
840 bool stats_enabled; 839 unsigned int stats_disable_count;
841 840
842 struct efx_mac_operations *mac_op; 841 struct efx_mac_operations *mac_op;
843 unsigned char mac_address[ETH_ALEN]; 842 unsigned char mac_address[ETH_ALEN];
diff --git a/drivers/net/sfc/phy.h b/drivers/net/sfc/phy.h
index 58c493ef81bb..07e855c148bc 100644
--- a/drivers/net/sfc/phy.h
+++ b/drivers/net/sfc/phy.h
@@ -17,7 +17,6 @@ extern struct efx_phy_operations falcon_sfx7101_phy_ops;
17extern struct efx_phy_operations falcon_sft9001_phy_ops; 17extern struct efx_phy_operations falcon_sft9001_phy_ops;
18 18
19extern void tenxpress_phy_blink(struct efx_nic *efx, bool blink); 19extern void tenxpress_phy_blink(struct efx_nic *efx, bool blink);
20extern void tenxpress_crc_err(struct efx_nic *efx);
21 20
22/**************************************************************************** 21/****************************************************************************
23 * Exported functions from the driver for XFP optical PHYs 22 * Exported functions from the driver for XFP optical PHYs
diff --git a/drivers/net/sfc/selftest.c b/drivers/net/sfc/selftest.c
index dba0d64d50cd..0a598084c513 100644
--- a/drivers/net/sfc/selftest.c
+++ b/drivers/net/sfc/selftest.c
@@ -665,6 +665,7 @@ int efx_selftest(struct efx_nic *efx, struct efx_self_tests *tests,
665{ 665{
666 enum efx_loopback_mode loopback_mode = efx->loopback_mode; 666 enum efx_loopback_mode loopback_mode = efx->loopback_mode;
667 int phy_mode = efx->phy_mode; 667 int phy_mode = efx->phy_mode;
668 enum reset_type reset_method = RESET_TYPE_INVISIBLE;
668 struct ethtool_cmd ecmd; 669 struct ethtool_cmd ecmd;
669 struct efx_channel *channel; 670 struct efx_channel *channel;
670 int rc_test = 0, rc_reset = 0, rc; 671 int rc_test = 0, rc_reset = 0, rc;
@@ -718,21 +719,21 @@ int efx_selftest(struct efx_nic *efx, struct efx_self_tests *tests,
718 mutex_unlock(&efx->mac_lock); 719 mutex_unlock(&efx->mac_lock);
719 720
720 /* free up all consumers of SRAM (including all the queues) */ 721 /* free up all consumers of SRAM (including all the queues) */
721 efx_reset_down(efx, &ecmd); 722 efx_reset_down(efx, reset_method, &ecmd);
722 723
723 rc = efx_test_chip(efx, tests); 724 rc = efx_test_chip(efx, tests);
724 if (rc && !rc_test) 725 if (rc && !rc_test)
725 rc_test = rc; 726 rc_test = rc;
726 727
727 /* reset the chip to recover from the register test */ 728 /* reset the chip to recover from the register test */
728 rc_reset = falcon_reset_hw(efx, RESET_TYPE_ALL); 729 rc_reset = falcon_reset_hw(efx, reset_method);
729 730
730 /* Ensure that the phy is powered and out of loopback 731 /* Ensure that the phy is powered and out of loopback
731 * for the bist and loopback tests */ 732 * for the bist and loopback tests */
732 efx->phy_mode &= ~PHY_MODE_LOW_POWER; 733 efx->phy_mode &= ~PHY_MODE_LOW_POWER;
733 efx->loopback_mode = LOOPBACK_NONE; 734 efx->loopback_mode = LOOPBACK_NONE;
734 735
735 rc = efx_reset_up(efx, &ecmd, rc_reset == 0); 736 rc = efx_reset_up(efx, reset_method, &ecmd, rc_reset == 0);
736 if (rc && !rc_reset) 737 if (rc && !rc_reset)
737 rc_reset = rc; 738 rc_reset = rc;
738 739
diff --git a/drivers/net/sfc/sfe4001.c b/drivers/net/sfc/sfe4001.c
index 16b80acb9992..cb25ae5b257a 100644
--- a/drivers/net/sfc/sfe4001.c
+++ b/drivers/net/sfc/sfe4001.c
@@ -186,19 +186,22 @@ static int sfn4111t_reset(struct efx_nic *efx)
186{ 186{
187 efx_oword_t reg; 187 efx_oword_t reg;
188 188
189 /* GPIO pins are also used for I2C, so block that temporarily */ 189 /* GPIO 3 and the GPIO register are shared with I2C, so block that */
190 mutex_lock(&efx->i2c_adap.bus_lock); 190 mutex_lock(&efx->i2c_adap.bus_lock);
191 191
192 /* Pull RST_N (GPIO 2) low then let it up again, setting the
193 * FLASH_CFG_1 strap (GPIO 3) appropriately. Only change the
194 * output enables; the output levels should always be 0 (low)
195 * and we rely on external pull-ups. */
192 falcon_read(efx, &reg, GPIO_CTL_REG_KER); 196 falcon_read(efx, &reg, GPIO_CTL_REG_KER);
193 EFX_SET_OWORD_FIELD(reg, GPIO2_OEN, true); 197 EFX_SET_OWORD_FIELD(reg, GPIO2_OEN, true);
194 EFX_SET_OWORD_FIELD(reg, GPIO2_OUT, false);
195 falcon_write(efx, &reg, GPIO_CTL_REG_KER); 198 falcon_write(efx, &reg, GPIO_CTL_REG_KER);
196 msleep(1000); 199 msleep(1000);
197 EFX_SET_OWORD_FIELD(reg, GPIO2_OUT, true); 200 EFX_SET_OWORD_FIELD(reg, GPIO2_OEN, false);
198 EFX_SET_OWORD_FIELD(reg, GPIO3_OEN, true); 201 EFX_SET_OWORD_FIELD(reg, GPIO3_OEN,
199 EFX_SET_OWORD_FIELD(reg, GPIO3_OUT, 202 !!(efx->phy_mode & PHY_MODE_SPECIAL));
200 !(efx->phy_mode & PHY_MODE_SPECIAL));
201 falcon_write(efx, &reg, GPIO_CTL_REG_KER); 203 falcon_write(efx, &reg, GPIO_CTL_REG_KER);
204 msleep(1);
202 205
203 mutex_unlock(&efx->i2c_adap.bus_lock); 206 mutex_unlock(&efx->i2c_adap.bus_lock);
204 207
@@ -232,12 +235,18 @@ static ssize_t set_phy_flash_cfg(struct device *dev,
232 } else if (efx->state != STATE_RUNNING || netif_running(efx->net_dev)) { 235 } else if (efx->state != STATE_RUNNING || netif_running(efx->net_dev)) {
233 err = -EBUSY; 236 err = -EBUSY;
234 } else { 237 } else {
238 /* Reset the PHY, reconfigure the MAC and enable/disable
239 * MAC stats accordingly. */
235 efx->phy_mode = new_mode; 240 efx->phy_mode = new_mode;
241 if (new_mode & PHY_MODE_SPECIAL)
242 efx_stats_disable(efx);
236 if (efx->board_info.type == EFX_BOARD_SFE4001) 243 if (efx->board_info.type == EFX_BOARD_SFE4001)
237 err = sfe4001_poweron(efx); 244 err = sfe4001_poweron(efx);
238 else 245 else
239 err = sfn4111t_reset(efx); 246 err = sfn4111t_reset(efx);
240 efx_reconfigure_port(efx); 247 efx_reconfigure_port(efx);
248 if (!(new_mode & PHY_MODE_SPECIAL))
249 efx_stats_enable(efx);
241 } 250 }
242 rtnl_unlock(); 251 rtnl_unlock();
243 252
@@ -326,6 +335,11 @@ int sfe4001_init(struct efx_nic *efx)
326 efx->board_info.monitor = sfe4001_check_hw; 335 efx->board_info.monitor = sfe4001_check_hw;
327 efx->board_info.fini = sfe4001_fini; 336 efx->board_info.fini = sfe4001_fini;
328 337
338 if (efx->phy_mode & PHY_MODE_SPECIAL) {
339 /* PHY won't generate a 156.25 MHz clock and MAC stats fetch
340 * will fail. */
341 efx_stats_disable(efx);
342 }
329 rc = sfe4001_poweron(efx); 343 rc = sfe4001_poweron(efx);
330 if (rc) 344 if (rc)
331 goto fail_ioexp; 345 goto fail_ioexp;
@@ -372,17 +386,25 @@ static void sfn4111t_fini(struct efx_nic *efx)
372 i2c_unregister_device(efx->board_info.hwmon_client); 386 i2c_unregister_device(efx->board_info.hwmon_client);
373} 387}
374 388
375static struct i2c_board_info sfn4111t_hwmon_info = { 389static struct i2c_board_info sfn4111t_a0_hwmon_info = {
376 I2C_BOARD_INFO("max6647", 0x4e), 390 I2C_BOARD_INFO("max6647", 0x4e),
377 .irq = -1, 391 .irq = -1,
378}; 392};
379 393
394static struct i2c_board_info sfn4111t_r5_hwmon_info = {
395 I2C_BOARD_INFO("max6646", 0x4d),
396 .irq = -1,
397};
398
380int sfn4111t_init(struct efx_nic *efx) 399int sfn4111t_init(struct efx_nic *efx)
381{ 400{
382 int rc; 401 int rc;
383 402
384 efx->board_info.hwmon_client = 403 efx->board_info.hwmon_client =
385 i2c_new_device(&efx->i2c_adap, &sfn4111t_hwmon_info); 404 i2c_new_device(&efx->i2c_adap,
405 (efx->board_info.minor < 5) ?
406 &sfn4111t_a0_hwmon_info :
407 &sfn4111t_r5_hwmon_info);
386 if (!efx->board_info.hwmon_client) 408 if (!efx->board_info.hwmon_client)
387 return -EIO; 409 return -EIO;
388 410
@@ -394,8 +416,10 @@ int sfn4111t_init(struct efx_nic *efx)
394 if (rc) 416 if (rc)
395 goto fail_hwmon; 417 goto fail_hwmon;
396 418
397 if (efx->phy_mode & PHY_MODE_SPECIAL) 419 if (efx->phy_mode & PHY_MODE_SPECIAL) {
420 efx_stats_disable(efx);
398 sfn4111t_reset(efx); 421 sfn4111t_reset(efx);
422 }
399 423
400 return 0; 424 return 0;
401 425
diff --git a/drivers/net/sfc/tenxpress.c b/drivers/net/sfc/tenxpress.c
index 9ecb77da9545..f0efd246962c 100644
--- a/drivers/net/sfc/tenxpress.c
+++ b/drivers/net/sfc/tenxpress.c
@@ -67,6 +67,8 @@
67#define PMA_PMD_EXT_CLK312_WIDTH 1 67#define PMA_PMD_EXT_CLK312_WIDTH 1
68#define PMA_PMD_EXT_LPOWER_LBN 12 68#define PMA_PMD_EXT_LPOWER_LBN 12
69#define PMA_PMD_EXT_LPOWER_WIDTH 1 69#define PMA_PMD_EXT_LPOWER_WIDTH 1
70#define PMA_PMD_EXT_ROBUST_LBN 14
71#define PMA_PMD_EXT_ROBUST_WIDTH 1
70#define PMA_PMD_EXT_SSR_LBN 15 72#define PMA_PMD_EXT_SSR_LBN 15
71#define PMA_PMD_EXT_SSR_WIDTH 1 73#define PMA_PMD_EXT_SSR_WIDTH 1
72 74
@@ -177,35 +179,24 @@
177#define C22EXT_STATUS_LINK_LBN 2 179#define C22EXT_STATUS_LINK_LBN 2
178#define C22EXT_STATUS_LINK_WIDTH 1 180#define C22EXT_STATUS_LINK_WIDTH 1
179 181
180#define C22EXT_MSTSLV_REG 49162 182#define C22EXT_MSTSLV_CTRL 49161
181#define C22EXT_MSTSLV_1000_HD_LBN 10 183#define C22EXT_MSTSLV_CTRL_ADV_1000_HD_LBN 8
182#define C22EXT_MSTSLV_1000_HD_WIDTH 1 184#define C22EXT_MSTSLV_CTRL_ADV_1000_FD_LBN 9
183#define C22EXT_MSTSLV_1000_FD_LBN 11 185
184#define C22EXT_MSTSLV_1000_FD_WIDTH 1 186#define C22EXT_MSTSLV_STATUS 49162
187#define C22EXT_MSTSLV_STATUS_LP_1000_HD_LBN 10
188#define C22EXT_MSTSLV_STATUS_LP_1000_FD_LBN 11
185 189
186/* Time to wait between powering down the LNPGA and turning off the power 190/* Time to wait between powering down the LNPGA and turning off the power
187 * rails */ 191 * rails */
188#define LNPGA_PDOWN_WAIT (HZ / 5) 192#define LNPGA_PDOWN_WAIT (HZ / 5)
189 193
190static int crc_error_reset_threshold = 100;
191module_param(crc_error_reset_threshold, int, 0644);
192MODULE_PARM_DESC(crc_error_reset_threshold,
193 "Max number of CRC errors before XAUI reset");
194
195struct tenxpress_phy_data { 194struct tenxpress_phy_data {
196 enum efx_loopback_mode loopback_mode; 195 enum efx_loopback_mode loopback_mode;
197 atomic_t bad_crc_count;
198 enum efx_phy_mode phy_mode; 196 enum efx_phy_mode phy_mode;
199 int bad_lp_tries; 197 int bad_lp_tries;
200}; 198};
201 199
202void tenxpress_crc_err(struct efx_nic *efx)
203{
204 struct tenxpress_phy_data *phy_data = efx->phy_data;
205 if (phy_data != NULL)
206 atomic_inc(&phy_data->bad_crc_count);
207}
208
209static ssize_t show_phy_short_reach(struct device *dev, 200static ssize_t show_phy_short_reach(struct device *dev,
210 struct device_attribute *attr, char *buf) 201 struct device_attribute *attr, char *buf)
211{ 202{
@@ -284,7 +275,9 @@ static int tenxpress_init(struct efx_nic *efx)
284 PMA_PMD_XCONTROL_REG); 275 PMA_PMD_XCONTROL_REG);
285 reg |= ((1 << PMA_PMD_EXT_GMII_EN_LBN) | 276 reg |= ((1 << PMA_PMD_EXT_GMII_EN_LBN) |
286 (1 << PMA_PMD_EXT_CLK_OUT_LBN) | 277 (1 << PMA_PMD_EXT_CLK_OUT_LBN) |
287 (1 << PMA_PMD_EXT_CLK312_LBN)); 278 (1 << PMA_PMD_EXT_CLK312_LBN) |
279 (1 << PMA_PMD_EXT_ROBUST_LBN));
280
288 mdio_clause45_write(efx, phy_id, MDIO_MMD_PMAPMD, 281 mdio_clause45_write(efx, phy_id, MDIO_MMD_PMAPMD,
289 PMA_PMD_XCONTROL_REG, reg); 282 PMA_PMD_XCONTROL_REG, reg);
290 mdio_clause45_set_flag(efx, phy_id, MDIO_MMD_C22EXT, 283 mdio_clause45_set_flag(efx, phy_id, MDIO_MMD_C22EXT,
@@ -346,6 +339,7 @@ static int tenxpress_phy_init(struct efx_nic *efx)
346 rc = tenxpress_init(efx); 339 rc = tenxpress_init(efx);
347 if (rc < 0) 340 if (rc < 0)
348 goto fail; 341 goto fail;
342 mdio_clause45_set_pause(efx);
349 343
350 if (efx->phy_type == PHY_TYPE_SFT9001B) { 344 if (efx->phy_type == PHY_TYPE_SFT9001B) {
351 rc = device_create_file(&efx->pci_dev->dev, 345 rc = device_create_file(&efx->pci_dev->dev,
@@ -376,8 +370,8 @@ static int tenxpress_special_reset(struct efx_nic *efx)
376 370
377 /* The XGMAC clock is driven from the SFC7101/SFT9001 312MHz clock, so 371 /* The XGMAC clock is driven from the SFC7101/SFT9001 312MHz clock, so
378 * a special software reset can glitch the XGMAC sufficiently for stats 372 * a special software reset can glitch the XGMAC sufficiently for stats
379 * requests to fail. Since we don't often special_reset, just lock. */ 373 * requests to fail. */
380 spin_lock(&efx->stats_lock); 374 efx_stats_disable(efx);
381 375
382 /* Initiate reset */ 376 /* Initiate reset */
383 reg = mdio_clause45_read(efx, efx->mii.phy_id, 377 reg = mdio_clause45_read(efx, efx->mii.phy_id,
@@ -392,17 +386,17 @@ static int tenxpress_special_reset(struct efx_nic *efx)
392 rc = mdio_clause45_wait_reset_mmds(efx, 386 rc = mdio_clause45_wait_reset_mmds(efx,
393 TENXPRESS_REQUIRED_DEVS); 387 TENXPRESS_REQUIRED_DEVS);
394 if (rc < 0) 388 if (rc < 0)
395 goto unlock; 389 goto out;
396 390
397 /* Try and reconfigure the device */ 391 /* Try and reconfigure the device */
398 rc = tenxpress_init(efx); 392 rc = tenxpress_init(efx);
399 if (rc < 0) 393 if (rc < 0)
400 goto unlock; 394 goto out;
401 395
402 /* Wait for the XGXS state machine to churn */ 396 /* Wait for the XGXS state machine to churn */
403 mdelay(10); 397 mdelay(10);
404unlock: 398out:
405 spin_unlock(&efx->stats_lock); 399 efx_stats_enable(efx);
406 return rc; 400 return rc;
407} 401}
408 402
@@ -520,7 +514,7 @@ static void tenxpress_phy_reconfigure(struct efx_nic *efx)
520{ 514{
521 struct tenxpress_phy_data *phy_data = efx->phy_data; 515 struct tenxpress_phy_data *phy_data = efx->phy_data;
522 struct ethtool_cmd ecmd; 516 struct ethtool_cmd ecmd;
523 bool phy_mode_change, loop_reset, loop_toggle, loopback; 517 bool phy_mode_change, loop_reset;
524 518
525 if (efx->phy_mode & (PHY_MODE_OFF | PHY_MODE_SPECIAL)) { 519 if (efx->phy_mode & (PHY_MODE_OFF | PHY_MODE_SPECIAL)) {
526 phy_data->phy_mode = efx->phy_mode; 520 phy_data->phy_mode = efx->phy_mode;
@@ -531,12 +525,10 @@ static void tenxpress_phy_reconfigure(struct efx_nic *efx)
531 525
532 phy_mode_change = (efx->phy_mode == PHY_MODE_NORMAL && 526 phy_mode_change = (efx->phy_mode == PHY_MODE_NORMAL &&
533 phy_data->phy_mode != PHY_MODE_NORMAL); 527 phy_data->phy_mode != PHY_MODE_NORMAL);
534 loopback = LOOPBACK_MASK(efx) & efx->phy_op->loopbacks;
535 loop_toggle = LOOPBACK_CHANGED(phy_data, efx, efx->phy_op->loopbacks);
536 loop_reset = (LOOPBACK_OUT_OF(phy_data, efx, efx->phy_op->loopbacks) || 528 loop_reset = (LOOPBACK_OUT_OF(phy_data, efx, efx->phy_op->loopbacks) ||
537 LOOPBACK_CHANGED(phy_data, efx, 1 << LOOPBACK_GPHY)); 529 LOOPBACK_CHANGED(phy_data, efx, 1 << LOOPBACK_GPHY));
538 530
539 if (loop_reset || loop_toggle || loopback || phy_mode_change) { 531 if (loop_reset || phy_mode_change) {
540 int rc; 532 int rc;
541 533
542 efx->phy_op->get_settings(efx, &ecmd); 534 efx->phy_op->get_settings(efx, &ecmd);
@@ -551,20 +543,6 @@ static void tenxpress_phy_reconfigure(struct efx_nic *efx)
551 falcon_reset_xaui(efx); 543 falcon_reset_xaui(efx);
552 } 544 }
553 545
554 if (efx->phy_type != PHY_TYPE_SFX7101) {
555 /* Only change autoneg once, on coming out or
556 * going into loopback */
557 if (loop_toggle)
558 ecmd.autoneg = !loopback;
559 if (loopback) {
560 ecmd.duplex = DUPLEX_FULL;
561 if (efx->loopback_mode == LOOPBACK_GPHY)
562 ecmd.speed = SPEED_1000;
563 else
564 ecmd.speed = SPEED_10000;
565 }
566 }
567
568 rc = efx->phy_op->set_settings(efx, &ecmd); 546 rc = efx->phy_op->set_settings(efx, &ecmd);
569 WARN_ON(rc); 547 WARN_ON(rc);
570 } 548 }
@@ -623,13 +601,6 @@ static void tenxpress_phy_poll(struct efx_nic *efx)
623 601
624 if (phy_data->phy_mode != PHY_MODE_NORMAL) 602 if (phy_data->phy_mode != PHY_MODE_NORMAL)
625 return; 603 return;
626
627 if (EFX_WORKAROUND_10750(efx) &&
628 atomic_read(&phy_data->bad_crc_count) > crc_error_reset_threshold) {
629 EFX_ERR(efx, "Resetting XAUI due to too many CRC errors\n");
630 falcon_reset_xaui(efx);
631 atomic_set(&phy_data->bad_crc_count, 0);
632 }
633} 604}
634 605
635static void tenxpress_phy_fini(struct efx_nic *efx) 606static void tenxpress_phy_fini(struct efx_nic *efx)
@@ -772,107 +743,76 @@ reset:
772 return rc; 743 return rc;
773} 744}
774 745
775static u32 tenxpress_get_xnp_lpa(struct efx_nic *efx) 746static void
747tenxpress_get_settings(struct efx_nic *efx, struct ethtool_cmd *ecmd)
776{ 748{
777 int phy = efx->mii.phy_id; 749 int phy_id = efx->mii.phy_id;
778 u32 lpa = 0; 750 u32 adv = 0, lpa = 0;
779 int reg; 751 int reg;
780 752
781 if (efx->phy_type != PHY_TYPE_SFX7101) { 753 if (efx->phy_type != PHY_TYPE_SFX7101) {
782 reg = mdio_clause45_read(efx, phy, MDIO_MMD_C22EXT, 754 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_C22EXT,
783 C22EXT_MSTSLV_REG); 755 C22EXT_MSTSLV_CTRL);
784 if (reg & (1 << C22EXT_MSTSLV_1000_HD_LBN)) 756 if (reg & (1 << C22EXT_MSTSLV_CTRL_ADV_1000_FD_LBN))
757 adv |= ADVERTISED_1000baseT_Full;
758 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_C22EXT,
759 C22EXT_MSTSLV_STATUS);
760 if (reg & (1 << C22EXT_MSTSLV_STATUS_LP_1000_HD_LBN))
785 lpa |= ADVERTISED_1000baseT_Half; 761 lpa |= ADVERTISED_1000baseT_Half;
786 if (reg & (1 << C22EXT_MSTSLV_1000_FD_LBN)) 762 if (reg & (1 << C22EXT_MSTSLV_STATUS_LP_1000_FD_LBN))
787 lpa |= ADVERTISED_1000baseT_Full; 763 lpa |= ADVERTISED_1000baseT_Full;
788 } 764 }
789 reg = mdio_clause45_read(efx, phy, MDIO_MMD_AN, MDIO_AN_10GBT_STATUS); 765 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_AN,
766 MDIO_AN_10GBT_CTRL);
767 if (reg & (1 << MDIO_AN_10GBT_CTRL_ADV_10G_LBN))
768 adv |= ADVERTISED_10000baseT_Full;
769 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_AN,
770 MDIO_AN_10GBT_STATUS);
790 if (reg & (1 << MDIO_AN_10GBT_STATUS_LP_10G_LBN)) 771 if (reg & (1 << MDIO_AN_10GBT_STATUS_LP_10G_LBN))
791 lpa |= ADVERTISED_10000baseT_Full; 772 lpa |= ADVERTISED_10000baseT_Full;
792 return lpa;
793}
794 773
795static void sfx7101_get_settings(struct efx_nic *efx, struct ethtool_cmd *ecmd) 774 mdio_clause45_get_settings_ext(efx, ecmd, adv, lpa);
796{ 775
797 mdio_clause45_get_settings_ext(efx, ecmd, ADVERTISED_10000baseT_Full, 776 if (efx->phy_type != PHY_TYPE_SFX7101)
798 tenxpress_get_xnp_lpa(efx)); 777 ecmd->supported |= (SUPPORTED_100baseT_Full |
799 ecmd->supported |= SUPPORTED_10000baseT_Full; 778 SUPPORTED_1000baseT_Full);
800 ecmd->advertising |= ADVERTISED_10000baseT_Full; 779
780 /* In loopback, the PHY automatically brings up the correct interface,
781 * but doesn't advertise the correct speed. So override it */
782 if (efx->loopback_mode == LOOPBACK_GPHY)
783 ecmd->speed = SPEED_1000;
784 else if (LOOPBACK_MASK(efx) & efx->phy_op->loopbacks)
785 ecmd->speed = SPEED_10000;
801} 786}
802 787
803static void sft9001_get_settings(struct efx_nic *efx, struct ethtool_cmd *ecmd) 788static int tenxpress_set_settings(struct efx_nic *efx, struct ethtool_cmd *ecmd)
804{ 789{
805 int phy_id = efx->mii.phy_id; 790 if (!ecmd->autoneg)
806 u32 xnp_adv = 0; 791 return -EINVAL;
807 int reg;
808
809 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_PMAPMD,
810 PMA_PMD_SPEED_ENABLE_REG);
811 if (EFX_WORKAROUND_13204(efx) && (reg & (1 << PMA_PMD_100TX_ADV_LBN)))
812 xnp_adv |= ADVERTISED_100baseT_Full;
813 if (reg & (1 << PMA_PMD_1000T_ADV_LBN))
814 xnp_adv |= ADVERTISED_1000baseT_Full;
815 if (reg & (1 << PMA_PMD_10000T_ADV_LBN))
816 xnp_adv |= ADVERTISED_10000baseT_Full;
817
818 mdio_clause45_get_settings_ext(efx, ecmd, xnp_adv,
819 tenxpress_get_xnp_lpa(efx));
820
821 ecmd->supported |= (SUPPORTED_100baseT_Half |
822 SUPPORTED_100baseT_Full |
823 SUPPORTED_1000baseT_Full);
824 792
825 /* Use the vendor defined C22ext register for duplex settings */ 793 return mdio_clause45_set_settings(efx, ecmd);
826 if (ecmd->speed != SPEED_10000 && !ecmd->autoneg) {
827 reg = mdio_clause45_read(efx, phy_id, MDIO_MMD_C22EXT,
828 GPHY_XCONTROL_REG);
829 ecmd->duplex = (reg & (1 << GPHY_DUPLEX_LBN) ?
830 DUPLEX_FULL : DUPLEX_HALF);
831 }
832} 794}
833 795
834static int sft9001_set_settings(struct efx_nic *efx, struct ethtool_cmd *ecmd) 796static void sfx7101_set_npage_adv(struct efx_nic *efx, u32 advertising)
835{ 797{
836 int phy_id = efx->mii.phy_id; 798 mdio_clause45_set_flag(efx, efx->mii.phy_id, MDIO_MMD_AN,
837 int rc; 799 MDIO_AN_10GBT_CTRL,
838 800 MDIO_AN_10GBT_CTRL_ADV_10G_LBN,
839 rc = mdio_clause45_set_settings(efx, ecmd); 801 advertising & ADVERTISED_10000baseT_Full);
840 if (rc)
841 return rc;
842
843 if (ecmd->speed != SPEED_10000 && !ecmd->autoneg)
844 mdio_clause45_set_flag(efx, phy_id, MDIO_MMD_C22EXT,
845 GPHY_XCONTROL_REG, GPHY_DUPLEX_LBN,
846 ecmd->duplex == DUPLEX_FULL);
847
848 return rc;
849} 802}
850 803
851static bool sft9001_set_xnp_advertise(struct efx_nic *efx, u32 advertising) 804static void sft9001_set_npage_adv(struct efx_nic *efx, u32 advertising)
852{ 805{
853 int phy = efx->mii.phy_id; 806 int phy_id = efx->mii.phy_id;
854 int reg = mdio_clause45_read(efx, phy, MDIO_MMD_PMAPMD, 807
855 PMA_PMD_SPEED_ENABLE_REG); 808 mdio_clause45_set_flag(efx, phy_id, MDIO_MMD_C22EXT,
856 bool enabled; 809 C22EXT_MSTSLV_CTRL,
857 810 C22EXT_MSTSLV_CTRL_ADV_1000_FD_LBN,
858 reg &= ~((1 << 2) | (1 << 3)); 811 advertising & ADVERTISED_1000baseT_Full);
859 if (EFX_WORKAROUND_13204(efx) && 812 mdio_clause45_set_flag(efx, phy_id, MDIO_MMD_AN,
860 (advertising & ADVERTISED_100baseT_Full)) 813 MDIO_AN_10GBT_CTRL,
861 reg |= 1 << PMA_PMD_100TX_ADV_LBN; 814 MDIO_AN_10GBT_CTRL_ADV_10G_LBN,
862 if (advertising & ADVERTISED_1000baseT_Full) 815 advertising & ADVERTISED_10000baseT_Full);
863 reg |= 1 << PMA_PMD_1000T_ADV_LBN;
864 if (advertising & ADVERTISED_10000baseT_Full)
865 reg |= 1 << PMA_PMD_10000T_ADV_LBN;
866 mdio_clause45_write(efx, phy, MDIO_MMD_PMAPMD,
867 PMA_PMD_SPEED_ENABLE_REG, reg);
868
869 enabled = (advertising &
870 (ADVERTISED_1000baseT_Half |
871 ADVERTISED_1000baseT_Full |
872 ADVERTISED_10000baseT_Full));
873 if (EFX_WORKAROUND_13204(efx))
874 enabled |= (advertising & ADVERTISED_100baseT_Full);
875 return enabled;
876} 816}
877 817
878struct efx_phy_operations falcon_sfx7101_phy_ops = { 818struct efx_phy_operations falcon_sfx7101_phy_ops = {
@@ -882,8 +822,9 @@ struct efx_phy_operations falcon_sfx7101_phy_ops = {
882 .poll = tenxpress_phy_poll, 822 .poll = tenxpress_phy_poll,
883 .fini = tenxpress_phy_fini, 823 .fini = tenxpress_phy_fini,
884 .clear_interrupt = efx_port_dummy_op_void, 824 .clear_interrupt = efx_port_dummy_op_void,
885 .get_settings = sfx7101_get_settings, 825 .get_settings = tenxpress_get_settings,
886 .set_settings = mdio_clause45_set_settings, 826 .set_settings = tenxpress_set_settings,
827 .set_npage_adv = sfx7101_set_npage_adv,
887 .num_tests = ARRAY_SIZE(sfx7101_test_names), 828 .num_tests = ARRAY_SIZE(sfx7101_test_names),
888 .test_names = sfx7101_test_names, 829 .test_names = sfx7101_test_names,
889 .run_tests = sfx7101_run_tests, 830 .run_tests = sfx7101_run_tests,
@@ -898,9 +839,9 @@ struct efx_phy_operations falcon_sft9001_phy_ops = {
898 .poll = tenxpress_phy_poll, 839 .poll = tenxpress_phy_poll,
899 .fini = tenxpress_phy_fini, 840 .fini = tenxpress_phy_fini,
900 .clear_interrupt = efx_port_dummy_op_void, 841 .clear_interrupt = efx_port_dummy_op_void,
901 .get_settings = sft9001_get_settings, 842 .get_settings = tenxpress_get_settings,
902 .set_settings = sft9001_set_settings, 843 .set_settings = tenxpress_set_settings,
903 .set_xnp_advertise = sft9001_set_xnp_advertise, 844 .set_npage_adv = sft9001_set_npage_adv,
904 .num_tests = ARRAY_SIZE(sft9001_test_names), 845 .num_tests = ARRAY_SIZE(sft9001_test_names),
905 .test_names = sft9001_test_names, 846 .test_names = sft9001_test_names,
906 .run_tests = sft9001_run_tests, 847 .run_tests = sft9001_run_tests,
diff --git a/drivers/net/sfc/workarounds.h b/drivers/net/sfc/workarounds.h
index 82e03e1d7371..78de68f4a95b 100644
--- a/drivers/net/sfc/workarounds.h
+++ b/drivers/net/sfc/workarounds.h
@@ -18,8 +18,8 @@
18#define EFX_WORKAROUND_ALWAYS(efx) 1 18#define EFX_WORKAROUND_ALWAYS(efx) 1
19#define EFX_WORKAROUND_FALCON_A(efx) (falcon_rev(efx) <= FALCON_REV_A1) 19#define EFX_WORKAROUND_FALCON_A(efx) (falcon_rev(efx) <= FALCON_REV_A1)
20#define EFX_WORKAROUND_10G(efx) EFX_IS10G(efx) 20#define EFX_WORKAROUND_10G(efx) EFX_IS10G(efx)
21#define EFX_WORKAROUND_SFX7101(efx) ((efx)->phy_type == PHY_TYPE_SFX7101) 21#define EFX_WORKAROUND_SFT9001(efx) ((efx)->phy_type == PHY_TYPE_SFT9001A || \
22#define EFX_WORKAROUND_SFT9001A(efx) ((efx)->phy_type == PHY_TYPE_SFT9001A) 22 (efx)->phy_type == PHY_TYPE_SFT9001B)
23 23
24/* XAUI resets if link not detected */ 24/* XAUI resets if link not detected */
25#define EFX_WORKAROUND_5147 EFX_WORKAROUND_ALWAYS 25#define EFX_WORKAROUND_5147 EFX_WORKAROUND_ALWAYS
@@ -29,8 +29,6 @@
29#define EFX_WORKAROUND_7884 EFX_WORKAROUND_10G 29#define EFX_WORKAROUND_7884 EFX_WORKAROUND_10G
30/* TX pkt parser problem with <= 16 byte TXes */ 30/* TX pkt parser problem with <= 16 byte TXes */
31#define EFX_WORKAROUND_9141 EFX_WORKAROUND_ALWAYS 31#define EFX_WORKAROUND_9141 EFX_WORKAROUND_ALWAYS
32/* Low rate CRC errors require XAUI reset */
33#define EFX_WORKAROUND_10750 EFX_WORKAROUND_SFX7101
34/* TX_EV_PKT_ERR can be caused by a dangling TX descriptor 32/* TX_EV_PKT_ERR can be caused by a dangling TX descriptor
35 * or a PCIe error (bug 11028) */ 33 * or a PCIe error (bug 11028) */
36#define EFX_WORKAROUND_10727 EFX_WORKAROUND_ALWAYS 34#define EFX_WORKAROUND_10727 EFX_WORKAROUND_ALWAYS
@@ -55,8 +53,8 @@
55#define EFX_WORKAROUND_8071 EFX_WORKAROUND_FALCON_A 53#define EFX_WORKAROUND_8071 EFX_WORKAROUND_FALCON_A
56 54
57/* Need to send XNP pages for 100BaseT */ 55/* Need to send XNP pages for 100BaseT */
58#define EFX_WORKAROUND_13204 EFX_WORKAROUND_SFT9001A 56#define EFX_WORKAROUND_13204 EFX_WORKAROUND_SFT9001
59/* Need to keep AN enabled */ 57/* Don't restart AN in near-side loopback */
60#define EFX_WORKAROUND_13963 EFX_WORKAROUND_SFT9001A 58#define EFX_WORKAROUND_15195 EFX_WORKAROUND_SFT9001
61 59
62#endif /* EFX_WORKAROUNDS_H */ 60#endif /* EFX_WORKAROUNDS_H */
diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c
index 3668e81e474d..994703cc0db3 100644
--- a/drivers/net/sky2.c
+++ b/drivers/net/sky2.c
@@ -1403,9 +1403,6 @@ static int sky2_up(struct net_device *dev)
1403 1403
1404 } 1404 }
1405 1405
1406 if (netif_msg_ifup(sky2))
1407 printk(KERN_INFO PFX "%s: enabling interface\n", dev->name);
1408
1409 netif_carrier_off(dev); 1406 netif_carrier_off(dev);
1410 1407
1411 /* must be power of 2 */ 1408 /* must be power of 2 */
@@ -1484,6 +1481,9 @@ static int sky2_up(struct net_device *dev)
1484 sky2_write32(hw, B0_IMSK, imask); 1481 sky2_write32(hw, B0_IMSK, imask);
1485 1482
1486 sky2_set_multicast(dev); 1483 sky2_set_multicast(dev);
1484
1485 if (netif_msg_ifup(sky2))
1486 printk(KERN_INFO PFX "%s: enabling interface\n", dev->name);
1487 return 0; 1487 return 0;
1488 1488
1489err_out: 1489err_out:
diff --git a/ipc/shm.c b/ipc/shm.c
index a9e09ad2263e..c0a021f7f41a 100644
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -368,14 +368,14 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
368 file = hugetlb_file_setup(name, size); 368 file = hugetlb_file_setup(name, size);
369 shp->mlock_user = current_user(); 369 shp->mlock_user = current_user();
370 } else { 370 } else {
371 int acctflag = VM_ACCOUNT; 371 int acctflag = 0;
372 /* 372 /*
373 * Do not allow no accounting for OVERCOMMIT_NEVER, even 373 * Do not allow no accounting for OVERCOMMIT_NEVER, even
374 * if it's asked for. 374 * if it's asked for.
375 */ 375 */
376 if ((shmflg & SHM_NORESERVE) && 376 if ((shmflg & SHM_NORESERVE) &&
377 sysctl_overcommit_memory != OVERCOMMIT_NEVER) 377 sysctl_overcommit_memory != OVERCOMMIT_NEVER)
378 acctflag = 0; 378 acctflag = VM_NORESERVE;
379 file = shmem_file_setup(name, size, acctflag); 379 file = shmem_file_setup(name, size, acctflag);
380 } 380 }
381 error = PTR_ERR(file); 381 error = PTR_ERR(file);
diff --git a/mm/mmap.c b/mm/mmap.c
index c581df14d0de..214b6a258eeb 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -1090,6 +1090,15 @@ int vma_wants_writenotify(struct vm_area_struct *vma)
1090 mapping_cap_account_dirty(vma->vm_file->f_mapping); 1090 mapping_cap_account_dirty(vma->vm_file->f_mapping);
1091} 1091}
1092 1092
1093/*
1094 * We account for memory if it's a private writeable mapping,
1095 * and VM_NORESERVE wasn't set.
1096 */
1097static inline int accountable_mapping(unsigned int vm_flags)
1098{
1099 return (vm_flags & (VM_NORESERVE | VM_SHARED | VM_WRITE)) == VM_WRITE;
1100}
1101
1093unsigned long mmap_region(struct file *file, unsigned long addr, 1102unsigned long mmap_region(struct file *file, unsigned long addr,
1094 unsigned long len, unsigned long flags, 1103 unsigned long len, unsigned long flags,
1095 unsigned int vm_flags, unsigned long pgoff, 1104 unsigned int vm_flags, unsigned long pgoff,
@@ -1117,23 +1126,24 @@ munmap_back:
1117 if (!may_expand_vm(mm, len >> PAGE_SHIFT)) 1126 if (!may_expand_vm(mm, len >> PAGE_SHIFT))
1118 return -ENOMEM; 1127 return -ENOMEM;
1119 1128
1120 if (flags & MAP_NORESERVE) 1129 /*
1130 * Set 'VM_NORESERVE' if we should not account for the
1131 * memory use of this mapping. We only honor MAP_NORESERVE
1132 * if we're allowed to overcommit memory.
1133 */
1134 if ((flags & MAP_NORESERVE) && sysctl_overcommit_memory != OVERCOMMIT_NEVER)
1135 vm_flags |= VM_NORESERVE;
1136 if (!accountable)
1121 vm_flags |= VM_NORESERVE; 1137 vm_flags |= VM_NORESERVE;
1122 1138
1123 if (accountable && (!(flags & MAP_NORESERVE) || 1139 /*
1124 sysctl_overcommit_memory == OVERCOMMIT_NEVER)) { 1140 * Private writable mapping: check memory availability
1125 if (vm_flags & VM_SHARED) { 1141 */
1126 /* Check memory availability in shmem_file_setup? */ 1142 if (accountable_mapping(vm_flags)) {
1127 vm_flags |= VM_ACCOUNT; 1143 charged = len >> PAGE_SHIFT;
1128 } else if (vm_flags & VM_WRITE) { 1144 if (security_vm_enough_memory(charged))
1129 /* 1145 return -ENOMEM;
1130 * Private writable mapping: check memory availability 1146 vm_flags |= VM_ACCOUNT;
1131 */
1132 charged = len >> PAGE_SHIFT;
1133 if (security_vm_enough_memory(charged))
1134 return -ENOMEM;
1135 vm_flags |= VM_ACCOUNT;
1136 }
1137 } 1147 }
1138 1148
1139 /* 1149 /*
@@ -1184,14 +1194,6 @@ munmap_back:
1184 goto free_vma; 1194 goto free_vma;
1185 } 1195 }
1186 1196
1187 /* We set VM_ACCOUNT in a shared mapping's vm_flags, to inform
1188 * shmem_zero_setup (perhaps called through /dev/zero's ->mmap)
1189 * that memory reservation must be checked; but that reservation
1190 * belongs to shared memory object, not to vma: so now clear it.
1191 */
1192 if ((vm_flags & (VM_SHARED|VM_ACCOUNT)) == (VM_SHARED|VM_ACCOUNT))
1193 vma->vm_flags &= ~VM_ACCOUNT;
1194
1195 /* Can addr have changed?? 1197 /* Can addr have changed??
1196 * 1198 *
1197 * Answer: Yes, several device drivers can do it in their 1199 * Answer: Yes, several device drivers can do it in their
diff --git a/mm/shmem.c b/mm/shmem.c
index 5d0de96c9789..19d566ccdeea 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -2628,7 +2628,7 @@ struct file *shmem_file_setup(char *name, loff_t size, unsigned long flags)
2628 goto close_file; 2628 goto close_file;
2629 2629
2630#ifdef CONFIG_SHMEM 2630#ifdef CONFIG_SHMEM
2631 SHMEM_I(inode)->flags = flags & VM_ACCOUNT; 2631 SHMEM_I(inode)->flags = (flags & VM_NORESERVE) ? 0 : VM_ACCOUNT;
2632#endif 2632#endif
2633 d_instantiate(dentry, inode); 2633 d_instantiate(dentry, inode);
2634 inode->i_size = size; 2634 inode->i_size = size;
diff --git a/net/ipv6/ip6mr.c b/net/ipv6/ip6mr.c
index d19a84b79503..228be551e9c1 100644
--- a/net/ipv6/ip6mr.c
+++ b/net/ipv6/ip6mr.c
@@ -48,6 +48,7 @@
48#include <linux/pim.h> 48#include <linux/pim.h>
49#include <net/addrconf.h> 49#include <net/addrconf.h>
50#include <linux/netfilter_ipv6.h> 50#include <linux/netfilter_ipv6.h>
51#include <net/ip6_checksum.h>
51 52
52/* Big lock, protecting vif table, mrt cache and mroute socket state. 53/* Big lock, protecting vif table, mrt cache and mroute socket state.
53 Note that the changes are semaphored via rtnl_lock. 54 Note that the changes are semaphored via rtnl_lock.
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 5f94db2f3e9e..9454d4ae46df 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -77,6 +77,7 @@
77#include <linux/poll.h> 77#include <linux/poll.h>
78#include <linux/module.h> 78#include <linux/module.h>
79#include <linux/init.h> 79#include <linux/init.h>
80#include <linux/mutex.h>
80 81
81#ifdef CONFIG_INET 82#ifdef CONFIG_INET
82#include <net/inet_common.h> 83#include <net/inet_common.h>
@@ -175,6 +176,7 @@ struct packet_sock {
175#endif 176#endif
176 struct packet_type prot_hook; 177 struct packet_type prot_hook;
177 spinlock_t bind_lock; 178 spinlock_t bind_lock;
179 struct mutex pg_vec_lock;
178 unsigned int running:1, /* prot_hook is attached*/ 180 unsigned int running:1, /* prot_hook is attached*/
179 auxdata:1, 181 auxdata:1,
180 origdev:1; 182 origdev:1;
@@ -1069,6 +1071,7 @@ static int packet_create(struct net *net, struct socket *sock, int protocol)
1069 */ 1071 */
1070 1072
1071 spin_lock_init(&po->bind_lock); 1073 spin_lock_init(&po->bind_lock);
1074 mutex_init(&po->pg_vec_lock);
1072 po->prot_hook.func = packet_rcv; 1075 po->prot_hook.func = packet_rcv;
1073 1076
1074 if (sock->type == SOCK_PACKET) 1077 if (sock->type == SOCK_PACKET)
@@ -1865,6 +1868,7 @@ static int packet_set_ring(struct sock *sk, struct tpacket_req *req, int closing
1865 synchronize_net(); 1868 synchronize_net();
1866 1869
1867 err = -EBUSY; 1870 err = -EBUSY;
1871 mutex_lock(&po->pg_vec_lock);
1868 if (closing || atomic_read(&po->mapped) == 0) { 1872 if (closing || atomic_read(&po->mapped) == 0) {
1869 err = 0; 1873 err = 0;
1870#define XC(a, b) ({ __typeof__ ((a)) __t; __t = (a); (a) = (b); __t; }) 1874#define XC(a, b) ({ __typeof__ ((a)) __t; __t = (a); (a) = (b); __t; })
@@ -1886,6 +1890,7 @@ static int packet_set_ring(struct sock *sk, struct tpacket_req *req, int closing
1886 if (atomic_read(&po->mapped)) 1890 if (atomic_read(&po->mapped))
1887 printk(KERN_DEBUG "packet_mmap: vma is busy: %d\n", atomic_read(&po->mapped)); 1891 printk(KERN_DEBUG "packet_mmap: vma is busy: %d\n", atomic_read(&po->mapped));
1888 } 1892 }
1893 mutex_unlock(&po->pg_vec_lock);
1889 1894
1890 spin_lock(&po->bind_lock); 1895 spin_lock(&po->bind_lock);
1891 if (was_running && !po->running) { 1896 if (was_running && !po->running) {
@@ -1918,7 +1923,7 @@ static int packet_mmap(struct file *file, struct socket *sock, struct vm_area_st
1918 1923
1919 size = vma->vm_end - vma->vm_start; 1924 size = vma->vm_end - vma->vm_start;
1920 1925
1921 lock_sock(sk); 1926 mutex_lock(&po->pg_vec_lock);
1922 if (po->pg_vec == NULL) 1927 if (po->pg_vec == NULL)
1923 goto out; 1928 goto out;
1924 if (size != po->pg_vec_len*po->pg_vec_pages*PAGE_SIZE) 1929 if (size != po->pg_vec_len*po->pg_vec_pages*PAGE_SIZE)
@@ -1941,7 +1946,7 @@ static int packet_mmap(struct file *file, struct socket *sock, struct vm_area_st
1941 err = 0; 1946 err = 0;
1942 1947
1943out: 1948out:
1944 release_sock(sk); 1949 mutex_unlock(&po->pg_vec_lock);
1945 return err; 1950 return err;
1946} 1951}
1947#endif 1952#endif