From 33f0f88f1c51ae5c2d593d26960c760ea154c2e2 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Mon, 9 Jan 2006 20:54:13 -0800 Subject: [PATCH] TTY layer buffering revamp The API and code have been through various bits of initial review by serial driver people but they definitely need to live somewhere for a while so the unconverted drivers can get knocked into shape, existing drivers that have been updated can be better tuned and bugs whacked out. This replaces the tty flip buffers with kmalloc objects in rings. In the normal situation for an IRQ driven serial port at typical speeds the behaviour is pretty much the same, two buffers end up allocated and the kernel cycles between them as before. When there are delays or at high speed we now behave far better as the buffer pool can grow a bit rather than lose characters. This also means that we can operate at higher speeds reliably. For drivers that receive characters in blocks (DMA based, USB and especially virtualisation) the layer allows a lot of driver specific code that works around the tty layer with private secondary queues to be removed. The IBM folks need this sort of layer, the smart serial port people do, the virtualisers do (because a virtualised tty typically operates at infinite speed rather than emulating 9600 baud). Finally many drivers had invalid and unsafe attempts to avoid buffer overflows by directly invoking tty methods extracted out of the innards of work queue structs. These are no longer needed and all go away. That fixes various random hangs with serial ports on overflow. The other change in here is to optimise the receive_room path that is used by some callers. It turns out that only one ldisc uses receive room except asa constant and it updates it far far less than the value is read. We thus make it a variable not a function call. I expect the code to contain bugs due to the size alone but I'll be watching and squashing them and feeding out new patches as it goes. Because the buffers now dynamically expand you should only run out of buffering when the kernel runs out of memory for real. That means a lot of the horrible hacks high performance drivers used to do just aren't needed any more. Description: tty_insert_flip_char is an old API and continues to work as before, as does tty_flip_buffer_push() [this is why many drivers dont need modification]. It does now also return the number of chars inserted There are also tty_buffer_request_room(tty, len) which asks for a buffer block of the length requested and returns the space found. This improves efficiency with hardware that knows how much to transfer. and tty_insert_flip_string_flags(tty, str, flags, len) to insert a string of characters and flags For a smart interface the usual code is len = tty_request_buffer_room(tty, amount_hardware_says); tty_insert_flip_string(tty, buffer_from_card, len); More description! At the moment tty buffers are attached directly to the tty. This is causing a lot of the problems related to tty layer locking, also problems at high speed and also with bursty data (such as occurs in virtualised environments) I'm working on ripping out the flip buffers and replacing them with a pool of dynamically allocated buffers. This allows both for old style "byte I/O" devices and also helps virtualisation and smart devices where large blocks of data suddenely materialise and need storing. So far so good. Lots of drivers reference tty->flip.*. Several of them also call directly and unsafely into function pointers it provides. This will all break. Most drivers can use tty_insert_flip_char which can be kept as an API but others need more. At the moment I've added the following interfaces, if people think more will be needed now is a good time to say int tty_buffer_request_room(tty, size) Try and ensure at least size bytes are available, returns actual room (may be zero). At the moment it just uses the flipbuf space but that will change. Repeated calls without characters being added are not cumulative. (ie if you call it with 1, 1, 1, and then 4 you'll have four characters of space. The other functions will also try and grow buffers in future but this will be a more efficient way when you know block sizes. int tty_insert_flip_char(tty, ch, flag) As before insert a character if there is room. Now returns 1 for success, 0 for failure. int tty_insert_flip_string(tty, str, len) Insert a block of non error characters. Returns the number inserted. int tty_prepare_flip_string(tty, strptr, len) Adjust the buffer to allow len characters to be added. Returns a buffer pointer in strptr and the length available. This allows for hardware that needs to use functions like insl or mencpy_fromio. Signed-off-by: Alan Cox Cc: Paul Fulghum Signed-off-by: Hirokazu Takata Signed-off-by: Serge Hallyn Signed-off-by: Jeff Dike Signed-off-by: John Hawkes Signed-off-by: Martin Schwidefsky Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/char/viocons.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'drivers/char/viocons.c') diff --git a/drivers/char/viocons.c b/drivers/char/viocons.c index 4d75c261f98a..cb82ebf4cb07 100644 --- a/drivers/char/viocons.c +++ b/drivers/char/viocons.c @@ -993,11 +993,10 @@ static void vioHandleData(struct HvLpEvent *event) * Don't attempt to copy more data into the buffer than we * have room for because it would fail without indication. */ - if ((tty->flip.count + 1) > TTY_FLIPBUF_SIZE) { + if(tty_insert_flip_char(tty, cevent->data[index], TTY_NORMAL) == 0) { printk(VIOCONS_KERN_WARN "input buffer overflow!\n"); break; } - tty_insert_flip_char(tty, cevent->data[index], TTY_NORMAL); } /* if cevent->len == 0 then no data was added to the buffer and flip.count == 0 */ -- cgit v1.2.3 From bf6a7112bda99aadd6675526423a96be6b356a3d Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Wed, 11 Jan 2006 11:54:08 +1100 Subject: [PATCH] powerpc: Early debugging support for iSeries Connect iSeries up to the standard early debugging infrastructure. To actually use this you need to enable the iSeries early debugging in setup_64.c. Then after the messages are logged hit Ctrl-x Ctrl-x on your console to dump the Hypervisor console buffer. Signed-off-by: Michael Ellerman Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/setup_64.c | 14 +++++++++----- arch/powerpc/platforms/iseries/setup.c | 18 +++++++++++++++--- drivers/char/viocons.c | 4 ++-- 3 files changed, 26 insertions(+), 10 deletions(-) (limited to 'drivers/char/viocons.c') diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index c4b76961d6de..177d8249078d 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -75,22 +75,26 @@ * but your kernel will not boot on anything else if you do so */ -/* This one is for use on LPAR machines that support an HVC console - * on vterm 0 - */ +/* For use on LPAR machines that support an HVC console on vterm 0 */ extern void udbg_init_debug_lpar(void); -/* This one is for use on Apple G5 machines - */ + +/* This one is for use on Apple G5 machines */ extern void udbg_init_pmac_realmode(void); + /* That's RTAS panel debug */ extern void call_rtas_display_status_delay(unsigned char c); + /* Here's maple real mode debug */ extern void udbg_init_maple_realmode(void); +/* For iSeries - hit Ctrl-x Ctrl-x to see the output */ +extern void udbg_init_iseries(void); + #define EARLY_DEBUG_INIT() do {} while(0) #if 0 #define EARLY_DEBUG_INIT() udbg_init_debug_lpar() +#define EARLY_DEBUG_INIT() udbg_init_iseries() #define EARLY_DEBUG_INIT() udbg_init_maple_realmode() #define EARLY_DEBUG_INIT() udbg_init_pmac_realmode() #define EARLY_DEBUG_INIT() \ diff --git a/arch/powerpc/platforms/iseries/setup.c b/arch/powerpc/platforms/iseries/setup.c index ad5ef80500ce..25cdebdf6f2e 100644 --- a/arch/powerpc/platforms/iseries/setup.c +++ b/arch/powerpc/platforms/iseries/setup.c @@ -52,6 +52,7 @@ #include #include #include +#include #include "naca.h" #include "setup.h" @@ -62,10 +63,8 @@ #include "call_sm.h" #include "call_hpt.h" -extern void hvlog(char *fmt, ...); - #ifdef DEBUG -#define DBG(fmt...) hvlog(fmt) +#define DBG(fmt...) udbg_printf(fmt) #else #define DBG(fmt...) #endif @@ -984,3 +983,16 @@ static int __init early_parsemem(char *p) return 0; } early_param("mem", early_parsemem); + +static void hvputc(char c) +{ + if (c == '\n') + hvputc('\r'); + + HvCall_writeLogBuffer(&c, 1); +} + +void __init udbg_init_iseries(void) +{ + udbg_putc = hvputc; +} diff --git a/drivers/char/viocons.c b/drivers/char/viocons.c index cb82ebf4cb07..bead38a4a68a 100644 --- a/drivers/char/viocons.c +++ b/drivers/char/viocons.c @@ -131,7 +131,7 @@ static void initDataEvent(struct viocharlpevent *viochar, HvLpIndex lp); static struct tty_driver *viotty_driver; -void hvlog(char *fmt, ...) +static void hvlog(char *fmt, ...) { int i; unsigned long flags; @@ -147,7 +147,7 @@ void hvlog(char *fmt, ...) spin_unlock_irqrestore(&consoleloglock, flags); } -void hvlogOutput(const char *buf, int count) +static void hvlogOutput(const char *buf, int count) { unsigned long flags; int begin; -- cgit v1.2.3 From 415c2e083a85f0daf6cbb4988691ac3ea9279509 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 11 Jan 2006 13:51:22 +1100 Subject: [PATCH] powerpc: fix up iSeries console after TTY layer buffering revamp Signed-off-by: Stephen Rothwell Signed-off-by: Paul Mackerras --- drivers/char/viocons.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'drivers/char/viocons.c') diff --git a/drivers/char/viocons.c b/drivers/char/viocons.c index bead38a4a68a..faee5e7acaf7 100644 --- a/drivers/char/viocons.c +++ b/drivers/char/viocons.c @@ -904,6 +904,7 @@ static void vioHandleData(struct HvLpEvent *event) struct viocharlpevent *cevent = (struct viocharlpevent *)event; struct port_info *pi; int index; + int num_pushed; u8 port = cevent->virtual_device; if (port >= VTTY_PORTS) { @@ -964,6 +965,7 @@ static void vioHandleData(struct HvLpEvent *event) * functionality will only work if built into the kernel and * then only if sysrq is enabled through the proc filesystem. */ + num_pushed = 0; for (index = 0; index < cevent->len; index++) { #ifdef CONFIG_MAGIC_SYSRQ if (sysrq_enabled) { @@ -997,11 +999,10 @@ static void vioHandleData(struct HvLpEvent *event) printk(VIOCONS_KERN_WARN "input buffer overflow!\n"); break; } + num_pushed++; } - /* if cevent->len == 0 then no data was added to the buffer and flip.count == 0 */ - if (tty->flip.count) - /* The next call resets flip.count when the data is flushed. */ + if (num_pushed) tty_flip_buffer_push(tty); } -- cgit v1.2.3 From 677f8c0d0455a35aaff08a0292ba9c6232ef4eb1 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Thu, 12 Jan 2006 13:47:43 +1100 Subject: [PATCH] powerpc: remove bitfields from HvLpEvent Signed-off-by: Stephen Rothwell Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/iseries/irq.c | 13 ++-------- arch/powerpc/platforms/iseries/lpevents.c | 12 ++++----- arch/powerpc/platforms/iseries/mf.c | 16 +++--------- arch/powerpc/platforms/iseries/viopath.c | 12 ++++----- drivers/block/viodasd.c | 32 +++++++++++------------- drivers/cdrom/viocd.c | 6 ++--- drivers/char/viocons.c | 31 +++++++++++------------ drivers/net/iseries_veth.c | 4 +-- include/asm-powerpc/iseries/hv_lp_event.h | 41 +++++++++++++++++++++++-------- 9 files changed, 82 insertions(+), 85 deletions(-) (limited to 'drivers/char/viocons.c') diff --git a/arch/powerpc/platforms/iseries/irq.c b/arch/powerpc/platforms/iseries/irq.c index 42e978e4897a..3d023856405a 100644 --- a/arch/powerpc/platforms/iseries/irq.c +++ b/arch/powerpc/platforms/iseries/irq.c @@ -147,20 +147,11 @@ static void int_received(struct pci_event *event, struct pt_regs *regs) static void pci_event_handler(struct HvLpEvent *event, struct pt_regs *regs) { if (event && (event->xType == HvLpEvent_Type_PciIo)) { - switch (event->xFlags.xFunction) { - case HvLpEvent_Function_Int: + if (hvlpevent_is_int(event)) int_received((struct pci_event *)event, regs); - break; - case HvLpEvent_Function_Ack: + else printk(KERN_ERR "pci_event_handler: unexpected ack received\n"); - break; - default: - printk(KERN_ERR - "pci_event_handler: unexpected event function %d\n", - (int)event->xFlags.xFunction); - break; - } } else if (event) printk(KERN_ERR "pci_event_handler: Unrecognized PCI event type 0x%x\n", diff --git a/arch/powerpc/platforms/iseries/lpevents.c b/arch/powerpc/platforms/iseries/lpevents.c index e9fb98bf895f..0b885300d1d1 100644 --- a/arch/powerpc/platforms/iseries/lpevents.c +++ b/arch/powerpc/platforms/iseries/lpevents.c @@ -53,7 +53,7 @@ static struct HvLpEvent * get_next_hvlpevent(void) struct HvLpEvent * event; event = (struct HvLpEvent *)hvlpevent_queue.xSlicCurEventPtr; - if (event->xFlags.xValid) { + if (hvlpevent_is_valid(event)) { /* rmb() needed only for weakly consistent machines (regatta) */ rmb(); /* Set pointer to next potential event */ @@ -84,7 +84,7 @@ int hvlpevent_is_pending(void) next_event = (struct HvLpEvent *)hvlpevent_queue.xSlicCurEventPtr; - return next_event->xFlags.xValid | + return hvlpevent_is_valid(next_event) || hvlpevent_queue.xPlicOverflowIntPending; } @@ -101,18 +101,18 @@ static void hvlpevent_clear_valid(struct HvLpEvent * event) switch (extra) { case 3: tmp = (struct HvLpEvent*)((char*)event + 3 * LpEventAlign); - tmp->xFlags.xValid = 0; + hvlpevent_invalidate(tmp); case 2: tmp = (struct HvLpEvent*)((char*)event + 2 * LpEventAlign); - tmp->xFlags.xValid = 0; + hvlpevent_invalidate(tmp); case 1: tmp = (struct HvLpEvent*)((char*)event + 1 * LpEventAlign); - tmp->xFlags.xValid = 0; + hvlpevent_invalidate(tmp); } mb(); - event->xFlags.xValid = 0; + hvlpevent_invalidate(event); } void process_hvlpevents(struct pt_regs *regs) diff --git a/arch/powerpc/platforms/iseries/mf.c b/arch/powerpc/platforms/iseries/mf.c index 49e7e4b85847..a41d8b78c0cd 100644 --- a/arch/powerpc/platforms/iseries/mf.c +++ b/arch/powerpc/platforms/iseries/mf.c @@ -251,10 +251,7 @@ static struct pending_event *new_pending_event(void) } memset(ev, 0, sizeof(struct pending_event)); hev = &ev->event.hp_lp_event; - hev->xFlags.xValid = 1; - hev->xFlags.xAckType = HvLpEvent_AckType_ImmediateAck; - hev->xFlags.xAckInd = HvLpEvent_AckInd_DoAck; - hev->xFlags.xFunction = HvLpEvent_Function_Int; + hev->flags = HV_LP_EVENT_VALID | HV_LP_EVENT_DO_ACK | HV_LP_EVENT_INT; hev->xType = HvLpEvent_Type_MachineFac; hev->xSourceLp = HvLpConfig_getLpIndex(); hev->xTargetLp = primary_lp; @@ -518,17 +515,10 @@ static void handle_ack(struct io_mf_lp_event *event) static void hv_handler(struct HvLpEvent *event, struct pt_regs *regs) { if ((event != NULL) && (event->xType == HvLpEvent_Type_MachineFac)) { - switch(event->xFlags.xFunction) { - case HvLpEvent_Function_Ack: + if (hvlpevent_is_ack(event)) handle_ack((struct io_mf_lp_event *)event); - break; - case HvLpEvent_Function_Int: + else handle_int((struct io_mf_lp_event *)event); - break; - default: - printk(KERN_ERR "mf.c: non ack/int event received\n"); - break; - } } else printk(KERN_ERR "mf.c: alien event received\n"); } diff --git a/arch/powerpc/platforms/iseries/viopath.c b/arch/powerpc/platforms/iseries/viopath.c index 842672695598..622a30149b48 100644 --- a/arch/powerpc/platforms/iseries/viopath.c +++ b/arch/powerpc/platforms/iseries/viopath.c @@ -270,7 +270,7 @@ static void handleMonitorEvent(struct HvLpEvent *event) * First see if this is just a normal monitor message from the * other partition */ - if (event->xFlags.xFunction == HvLpEvent_Function_Int) { + if (hvlpevent_is_int(event)) { remoteLp = event->xSourceLp; if (!viopathStatus[remoteLp].isActive) sendMonMsg(remoteLp); @@ -331,13 +331,12 @@ static void handleConfig(struct HvLpEvent *event) { if (!event) return; - if (event->xFlags.xFunction == HvLpEvent_Function_Int) { + if (hvlpevent_is_int(event)) { printk(VIOPATH_KERN_WARN "unexpected config request from partition %d", event->xSourceLp); - if ((event->xFlags.xFunction == HvLpEvent_Function_Int) && - (event->xFlags.xAckInd == HvLpEvent_AckInd_DoAck)) { + if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); } @@ -377,7 +376,7 @@ static void vio_handleEvent(struct HvLpEvent *event, struct pt_regs *regs) int subtype = (event->xSubtype & VIOMAJOR_SUBTYPE_MASK) >> VIOMAJOR_SUBTYPE_SHIFT; - if (event->xFlags.xFunction == HvLpEvent_Function_Int) { + if (hvlpevent_is_int(event)) { remoteLp = event->xSourceLp; /* * The isActive is checked because if the hosting partition @@ -436,8 +435,7 @@ static void vio_handleEvent(struct HvLpEvent *event, struct pt_regs *regs) "unexpected virtual io event subtype %d from partition %d\n", event->xSubtype, remoteLp); /* No handler. Ack if necessary */ - if ((event->xFlags.xFunction == HvLpEvent_Function_Int) && - (event->xFlags.xAckInd == HvLpEvent_AckInd_DoAck)) { + if (hvlpevent_is_int(event) && hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); } diff --git a/drivers/block/viodasd.c b/drivers/block/viodasd.c index d1aaf31bd97e..f63e07bd9f9c 100644 --- a/drivers/block/viodasd.c +++ b/drivers/block/viodasd.c @@ -293,6 +293,7 @@ static int send_request(struct request *req) u16 viocmd; HvLpEvent_Rc hvrc; struct vioblocklpevent *bevent; + struct HvLpEvent *hev; struct scatterlist sg[VIOMAXBLOCKDMA]; int sgindex; int statindex; @@ -347,22 +348,19 @@ static int send_request(struct request *req) * token so we can match the response up later */ memset(bevent, 0, sizeof(struct vioblocklpevent)); - bevent->event.xFlags.xValid = 1; - bevent->event.xFlags.xFunction = HvLpEvent_Function_Int; - bevent->event.xFlags.xAckInd = HvLpEvent_AckInd_DoAck; - bevent->event.xFlags.xAckType = HvLpEvent_AckType_ImmediateAck; - bevent->event.xType = HvLpEvent_Type_VirtualIo; - bevent->event.xSubtype = viocmd; - bevent->event.xSourceLp = HvLpConfig_getLpIndex(); - bevent->event.xTargetLp = viopath_hostLp; - bevent->event.xSizeMinus1 = + hev = &bevent->event; + hev->flags = HV_LP_EVENT_VALID | HV_LP_EVENT_DO_ACK | + HV_LP_EVENT_INT; + hev->xType = HvLpEvent_Type_VirtualIo; + hev->xSubtype = viocmd; + hev->xSourceLp = HvLpConfig_getLpIndex(); + hev->xTargetLp = viopath_hostLp; + hev->xSizeMinus1 = offsetof(struct vioblocklpevent, u.rw_data.dma_info) + (sizeof(bevent->u.rw_data.dma_info[0]) * nsg) - 1; - bevent->event.xSourceInstanceId = - viopath_sourceinst(viopath_hostLp); - bevent->event.xTargetInstanceId = - viopath_targetinst(viopath_hostLp); - bevent->event.xCorrelationToken = (u64)req; + hev->xSourceInstanceId = viopath_sourceinst(viopath_hostLp); + hev->xTargetInstanceId = viopath_targetinst(viopath_hostLp); + hev->xCorrelationToken = (u64)req; bevent->version = VIOVERSION; bevent->disk = DEVICE_NO(d); bevent->u.rw_data.offset = start; @@ -649,10 +647,10 @@ static void handle_block_event(struct HvLpEvent *event) /* Notification that a partition went away! */ return; /* First, we should NEVER get an int here...only acks */ - if (event->xFlags.xFunction == HvLpEvent_Function_Int) { + if (hvlpevent_is_int(event)) { printk(VIOD_KERN_WARNING "Yikes! got an int in viodasd event handler!\n"); - if (event->xFlags.xAckInd == HvLpEvent_AckInd_DoAck) { + if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); } @@ -695,7 +693,7 @@ static void handle_block_event(struct HvLpEvent *event) default: printk(VIOD_KERN_WARNING "invalid subtype!"); - if (event->xFlags.xAckInd == HvLpEvent_AckInd_DoAck) { + if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); } diff --git a/drivers/cdrom/viocd.c b/drivers/cdrom/viocd.c index b5191780ecca..193446e6a08a 100644 --- a/drivers/cdrom/viocd.c +++ b/drivers/cdrom/viocd.c @@ -542,10 +542,10 @@ static void vio_handle_cd_event(struct HvLpEvent *event) /* Notification that a partition went away! */ return; /* First, we should NEVER get an int here...only acks */ - if (event->xFlags.xFunction == HvLpEvent_Function_Int) { + if (hvlpevent_is_int(event)) { printk(VIOCD_KERN_WARNING "Yikes! got an int in viocd event handler!\n"); - if (event->xFlags.xAckInd == HvLpEvent_AckInd_DoAck) { + if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); } @@ -616,7 +616,7 @@ return_complete: printk(VIOCD_KERN_WARNING "message with invalid subtype %0x04X!\n", event->xSubtype & VIOMINOR_SUBTYPE_MASK); - if (event->xFlags.xAckInd == HvLpEvent_AckInd_DoAck) { + if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); } diff --git a/drivers/char/viocons.c b/drivers/char/viocons.c index faee5e7acaf7..4e5360388748 100644 --- a/drivers/char/viocons.c +++ b/drivers/char/viocons.c @@ -476,19 +476,19 @@ static struct port_info *get_port_data(struct tty_struct *tty) */ static void initDataEvent(struct viocharlpevent *viochar, HvLpIndex lp) { + struct HvLpEvent *hev = &viochar->event; + memset(viochar, 0, sizeof(struct viocharlpevent)); - viochar->event.xFlags.xValid = 1; - viochar->event.xFlags.xFunction = HvLpEvent_Function_Int; - viochar->event.xFlags.xAckInd = HvLpEvent_AckInd_NoAck; - viochar->event.xFlags.xAckType = HvLpEvent_AckType_DeferredAck; - viochar->event.xType = HvLpEvent_Type_VirtualIo; - viochar->event.xSubtype = viomajorsubtype_chario | viochardata; - viochar->event.xSourceLp = HvLpConfig_getLpIndex(); - viochar->event.xTargetLp = lp; - viochar->event.xSizeMinus1 = sizeof(struct viocharlpevent); - viochar->event.xSourceInstanceId = viopath_sourceinst(lp); - viochar->event.xTargetInstanceId = viopath_targetinst(lp); + hev->flags = HV_LP_EVENT_VALID | HV_LP_EVENT_DEFERRED_ACK | + HV_LP_EVENT_INT; + hev->xType = HvLpEvent_Type_VirtualIo; + hev->xSubtype = viomajorsubtype_chario | viochardata; + hev->xSourceLp = HvLpConfig_getLpIndex(); + hev->xTargetLp = lp; + hev->xSizeMinus1 = sizeof(struct viocharlpevent); + hev->xSourceInstanceId = viopath_sourceinst(lp); + hev->xTargetInstanceId = viopath_targetinst(lp); } /* @@ -752,7 +752,7 @@ static void vioHandleOpenEvent(struct HvLpEvent *event) struct port_info *pi; int reject = 0; - if (event->xFlags.xFunction == HvLpEvent_Function_Ack) { + if (hvlpevent_is_ack(event)) { if (port >= VTTY_PORTS) return; @@ -788,7 +788,7 @@ static void vioHandleOpenEvent(struct HvLpEvent *event) } /* This had better require an ack, otherwise complain */ - if (event->xFlags.xAckInd != HvLpEvent_AckInd_DoAck) { + if (!hvlpevent_need_ack(event)) { printk(VIOCONS_KERN_WARN "viocharopen without ack bit!\n"); return; } @@ -856,7 +856,7 @@ static void vioHandleCloseEvent(struct HvLpEvent *event) struct viocharlpevent *cevent = (struct viocharlpevent *)event; u8 port = cevent->virtual_device; - if (event->xFlags.xFunction == HvLpEvent_Function_Int) { + if (hvlpevent_is_int(event)) { if (port >= VTTY_PORTS) { printk(VIOCONS_KERN_WARN "close message from invalid virtual device.\n"); @@ -1056,8 +1056,7 @@ static void vioHandleCharEvent(struct HvLpEvent *event) vioHandleConfig(event); break; default: - if ((event->xFlags.xFunction == HvLpEvent_Function_Int) && - (event->xFlags.xAckInd == HvLpEvent_AckInd_DoAck)) { + if (hvlpevent_is_int(event) && hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); } diff --git a/drivers/net/iseries_veth.c b/drivers/net/iseries_veth.c index 77eadf84cb2c..f0f04be989d6 100644 --- a/drivers/net/iseries_veth.c +++ b/drivers/net/iseries_veth.c @@ -590,9 +590,9 @@ static void veth_handle_event(struct HvLpEvent *event, struct pt_regs *regs) { struct veth_lpevent *veth_event = (struct veth_lpevent *)event; - if (event->xFlags.xFunction == HvLpEvent_Function_Ack) + if (hvlpevent_is_ack(event)) veth_handle_ack(veth_event); - else if (event->xFlags.xFunction == HvLpEvent_Function_Int) + else veth_handle_int(veth_event); } diff --git a/include/asm-powerpc/iseries/hv_lp_event.h b/include/asm-powerpc/iseries/hv_lp_event.h index 499ab1ad0185..8ee89b67b8fd 100644 --- a/include/asm-powerpc/iseries/hv_lp_event.h +++ b/include/asm-powerpc/iseries/hv_lp_event.h @@ -32,17 +32,8 @@ * partitions through PLIC. */ -struct HvEventFlags { - u8 xValid:1; /* Indicates a valid request x00-x00 */ - u8 xRsvd1:4; /* Reserved ... */ - u8 xAckType:1; /* Immediate or deferred ... */ - u8 xAckInd:1; /* Indicates if ACK required ... */ - u8 xFunction:1; /* Interrupt or Acknowledge ... */ -}; - - struct HvLpEvent { - struct HvEventFlags xFlags; /* Event flags x00-x00 */ + u8 flags; /* Event flags x00-x00 */ u8 xType; /* Type of message x01-x01 */ u16 xSubtype; /* Subtype for event x02-x03 */ u8 xSourceLp; /* Source LP x04-x04 */ @@ -126,6 +117,11 @@ extern int HvLpEvent_closePath(HvLpEvent_Type eventType, HvLpIndex lpIndex); #define HvLpEvent_AckType_ImmediateAck 0 #define HvLpEvent_AckType_DeferredAck 1 +#define HV_LP_EVENT_INT 0x01 +#define HV_LP_EVENT_DO_ACK 0x02 +#define HV_LP_EVENT_DEFERRED_ACK 0x04 +#define HV_LP_EVENT_VALID 0x80 + #define HvLpDma_Direction_LocalToRemote 0 #define HvLpDma_Direction_RemoteToLocal 1 @@ -139,4 +135,29 @@ extern int HvLpEvent_closePath(HvLpEvent_Type eventType, HvLpIndex lpIndex); #define HvLpDma_Rc_InvalidAddress 4 #define HvLpDma_Rc_InvalidLength 5 +static inline int hvlpevent_is_valid(struct HvLpEvent *h) +{ + return h->flags & HV_LP_EVENT_VALID; +} + +static inline void hvlpevent_invalidate(struct HvLpEvent *h) +{ + h->flags &= ~ HV_LP_EVENT_VALID; +} + +static inline int hvlpevent_is_int(struct HvLpEvent *h) +{ + return h->flags & HV_LP_EVENT_INT; +} + +static inline int hvlpevent_is_ack(struct HvLpEvent *h) +{ + return !hvlpevent_is_int(h); +} + +static inline int hvlpevent_need_ack(struct HvLpEvent *h) +{ + return h->flags & HV_LP_EVENT_DO_ACK; +} + #endif /* _ASM_POWERPC_ISERIES_HV_LP_EVENT_H */ -- cgit v1.2.3