* USB on MPC885 vs. MPC850/823
From: Wolfgang Grandegger @ 2007-09-21 7:00 UTC (permalink / raw)
To: linuxppc-embedded
Hello,
I heard that the MPC885 has an improved USB hardware/firmware compared
to the one on the rather old MPC823 or MPC850. Is it similar to the USB
on the MPC8272? It would be nice, if somebody could shed some light on
the usability and the support of USB in Linux for the MPC885 (or similar
MPC8xx chips).
Thanks.
Wolfgang.
^ permalink raw reply
* Re: 44x bug: funny TLB writes?
From: David Gibson @ 2007-09-21 6:47 UTC (permalink / raw)
To: Hollis Blanchard; +Cc: linuxppc-dev
In-Reply-To: <1190356714.25483.19.camel@basalt>
On Fri, Sep 21, 2007 at 01:38:34AM -0500, Hollis Blanchard wrote:
> On Fri, 2007-09-21 at 15:42 +1000, David Gibson wrote:
> > On Thu, Sep 20, 2007 at 10:34:12PM -0500, Hollis Blanchard wrote:
> > > I seem to have come across a strange bug while doing KVM development. It
> > > seems that the final tlbwe in finish_tlb (head_44x.S) is actually
> > > leaking RPN bits into the "attribute" word.
> > >
> > > When I set a breakpoint there and press enter on the serial console, I
> > > see r12=ef600703, which is the physical address of the UART on this chip
> > > (440EP), plus the correct permission bits at the bottom.
> > >
> > > Am I crazy? I'm not really looking to step through that assembly right
> > > now... Clearly (current) hardware is just ignoring these errant writes,
> > > but it should be fixed.
> >
> > A quick glance at the code suggests this is indeed wrong. Hurrah.
> > Another reason to rewrite the 44x tlb miss handling.
>
> Just a quick fix would be fine too... ;)
I suppose.
> I'm just glad it's not a KVM bug, because when I dumped the TLB state
> and saw bizarre values I was getting really worried.
>
> > PS. "errant" and "error" are not cognate, even if the chip doc
> > writers think so...
>
> According to Merriam Webster, errant 2c is "c : behaving wrongly <an
> errant child>", so I'm OK with it.
Good heavens. No such usage mentioned in the Shorter OED, only as in
"knight errant". You Americans and your strange language :-p.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
^ permalink raw reply
* Re: [PATCH v2 2/4] Implement generic time of day clocksource for powerpc machines.
From: David Gibson @ 2007-09-21 6:43 UTC (permalink / raw)
To: Paul Mackerras
Cc: Daniel Walker, Thomas Gleixner, Realtime Kernel, linuxppc-dev
In-Reply-To: <18163.20403.967082.774118@cargo.ozlabs.ibm.com>
On Fri, Sep 21, 2007 at 02:59:31PM +1000, Paul Mackerras wrote:
> Daniel Walker writes:
>
> > I don't think the RTC frequency isn't the same as the timebase? Seems
> > like the RTC only case about seconds at the lowest level. If that's the
> > case then the jiffies clock might be better to use .. The other thing I
> > wonder is if the __USE_RTC boards might have lower level clocks that
> > could be used instead ..
>
> It's OK, the RTC isn't what you think it is, it's a pair of
> CPU-internal registers which count seconds and nanoseconds. On
> processors with the RTC, tb_ticks_per_sec is initialized to
> 1000000000. Trust me, this code is OK. :)
Indeed.. I'm wondering if we should do a s/rtc/ppc601rtc/ or
something. Otherwise people will keep making this mistake and be
confused.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
^ permalink raw reply
* Re: 44x bug: funny TLB writes?
From: Hollis Blanchard @ 2007-09-21 6:38 UTC (permalink / raw)
To: David Gibson; +Cc: linuxppc-dev
In-Reply-To: <20070921054218.GA13470@localhost.localdomain>
On Fri, 2007-09-21 at 15:42 +1000, David Gibson wrote:
> On Thu, Sep 20, 2007 at 10:34:12PM -0500, Hollis Blanchard wrote:
> > I seem to have come across a strange bug while doing KVM development. It
> > seems that the final tlbwe in finish_tlb (head_44x.S) is actually
> > leaking RPN bits into the "attribute" word.
> >
> > When I set a breakpoint there and press enter on the serial console, I
> > see r12=ef600703, which is the physical address of the UART on this chip
> > (440EP), plus the correct permission bits at the bottom.
> >
> > Am I crazy? I'm not really looking to step through that assembly right
> > now... Clearly (current) hardware is just ignoring these errant writes,
> > but it should be fixed.
>
> A quick glance at the code suggests this is indeed wrong. Hurrah.
> Another reason to rewrite the 44x tlb miss handling.
Just a quick fix would be fine too... ;)
I'm just glad it's not a KVM bug, because when I dumped the TLB state
and saw bizarre values I was getting really worried.
> PS. "errant" and "error" are not cognate, even if the chip doc
> writers think so...
According to Merriam Webster, errant 2c is "c : behaving wrongly <an
errant child>", so I'm OK with it.
--
Hollis Blanchard
IBM Linux Technology Center
^ permalink raw reply
* Re: 44x bug: funny TLB writes?
From: David Gibson @ 2007-09-21 5:42 UTC (permalink / raw)
To: Hollis Blanchard; +Cc: linuxppc-dev
In-Reply-To: <1190345652.25483.6.camel@basalt>
On Thu, Sep 20, 2007 at 10:34:12PM -0500, Hollis Blanchard wrote:
> I seem to have come across a strange bug while doing KVM development. It
> seems that the final tlbwe in finish_tlb (head_44x.S) is actually
> leaking RPN bits into the "attribute" word.
>
> When I set a breakpoint there and press enter on the serial console, I
> see r12=ef600703, which is the physical address of the UART on this chip
> (440EP), plus the correct permission bits at the bottom.
>
> Am I crazy? I'm not really looking to step through that assembly right
> now... Clearly (current) hardware is just ignoring these errant writes,
> but it should be fixed.
A quick glance at the code suggests this is indeed wrong. Hurrah.
Another reason to rewrite the 44x tlb miss handling.
PS. "errant" and "error" are not cognate, even if the chip doc
writers think so...
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
^ permalink raw reply
* Re: [PATCH v2 2/4] Implement generic time of day clocksource for powerpc machines.
From: Paul Mackerras @ 2007-09-21 4:59 UTC (permalink / raw)
To: Daniel Walker; +Cc: linuxppc-dev, Thomas Gleixner, Realtime Kernel
In-Reply-To: <1190347546.15347.27.camel@imap.mvista.com>
Daniel Walker writes:
> I don't think the RTC frequency isn't the same as the timebase? Seems
> like the RTC only case about seconds at the lowest level. If that's the
> case then the jiffies clock might be better to use .. The other thing I
> wonder is if the __USE_RTC boards might have lower level clocks that
> could be used instead ..
It's OK, the RTC isn't what you think it is, it's a pair of
CPU-internal registers which count seconds and nanoseconds. On
processors with the RTC, tb_ticks_per_sec is initialized to
1000000000. Trust me, this code is OK. :)
Paul.
^ permalink raw reply
* Re: [PATCH v2 2/4] Implement generic time of day clocksource for powerpc machines.
From: Stephen Rothwell @ 2007-09-21 4:52 UTC (permalink / raw)
To: Tony Breeds
Cc: linuxppc-dev, Thomas, Paul Mackerras, Gleixner, Realtime Kernel
In-Reply-To: <20070921032603.07AF232C889@thor>
[-- Attachment #1: Type: text/plain, Size: 712 bytes --]
Small comments.
On Fri, 21 Sep 2007 13:26:02 +1000 Tony Breeds <tony@bakeyournoodle.com> wrote:
>
> +void update_vsyscall(struct timespec *wall_time, struct clocksource *clock)
> +{
> + u64 t2x, stamp_xsec;
> +
> + if (__USE_RTC() || clock != &clocksource_timebase)
^^^^^^^^^^^
I think this is redundant as if __USE_RTC() is true, you register
clocksource_rtc below.
> +#ifdef CONFIG_ISERIES
> + if (!firmware_has_feature(FW_FEATURE_ISERIES))
> +#endif
The #ifdef is redundant since if CONFIG_ISERIES is not set,
firmware_has_feature(FW_FEATURE_ISERIES) is constant 0.
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply
* [PATCH 11/11] [POWERPC] iSeries: move viodasd probing
From: Stephen Rothwell @ 2007-09-21 4:40 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143954.3b707ab5.sfr@canb.auug.org.au>
This way we only have entries in the device tree for disks that actually
exist. A slight complication is that disks may be attached to LPARs
at runtime.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/platforms/iseries/dt.c | 6 -
arch/powerpc/platforms/iseries/vio.c | 305 ++++++++++++++++++++++++----------
drivers/block/viodasd.c | 77 +++------
include/asm-powerpc/iseries/vio.h | 47 +++++
4 files changed, 286 insertions(+), 149 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/platforms/iseries/dt.c b/arch/powerpc/platforms/iseries/dt.c
index 2e4ad6b..4543c4b 100644
--- a/arch/powerpc/platforms/iseries/dt.c
+++ b/arch/powerpc/platforms/iseries/dt.c
@@ -72,7 +72,6 @@ static char __initdata device_type_cpu[] = "cpu";
static char __initdata device_type_memory[] = "memory";
static char __initdata device_type_serial[] = "serial";
static char __initdata device_type_network[] = "network";
-static char __initdata device_type_block[] = "block";
static char __initdata device_type_pci[] = "pci";
static char __initdata device_type_vdevice[] = "vdevice";
static char __initdata device_type_vscsi[] = "vscsi";
@@ -374,11 +373,6 @@ static void __init dt_vdevices(struct iseries_flat_dt *dt)
dt_end_node(dt);
}
- reg += HVMAXARCHITECTEDVIRTUALLANS;
-
- for (i = 0; i < HVMAXARCHITECTEDVIRTUALDISKS; i++)
- dt_do_vdevice(dt, "viodasd", reg, i, device_type_block,
- "IBM,iSeries-viodasd", 1);
dt_end_node(dt);
}
diff --git a/arch/powerpc/platforms/iseries/vio.c b/arch/powerpc/platforms/iseries/vio.c
index c7d2a83..f52ecb7 100644
--- a/arch/powerpc/platforms/iseries/vio.c
+++ b/arch/powerpc/platforms/iseries/vio.c
@@ -25,8 +25,10 @@
#include <linux/gfp.h>
#include <linux/completion.h>
#include <linux/proc_fs.h>
+#include <linux/module.h>
#include <asm/firmware.h>
+#include <asm/vio.h>
#include <asm/iseries/vio.h>
#include <asm/iseries/iommu.h>
#include <asm/iseries/hv_types.h>
@@ -57,7 +59,7 @@ struct vio_resource {
char model[3];
};
-static struct property * __init new_property(const char *name, int length,
+static struct property *new_property(const char *name, int length,
const void *value)
{
struct property *np = kzalloc(sizeof(*np) + strlen(name) + 1 + length,
@@ -78,7 +80,7 @@ static void __init free_property(struct property *np)
kfree(np);
}
-static struct device_node * __init new_node(const char *path,
+static struct device_node *new_node(const char *path,
struct device_node *parent)
{
struct device_node *np = kzalloc(sizeof(*np), GFP_KERNEL);
@@ -97,7 +99,7 @@ static struct device_node * __init new_node(const char *path,
return np;
}
-static void __init free_node(struct device_node *np)
+static void free_node(struct device_node *np)
{
struct property *next;
struct property *prop;
@@ -113,7 +115,7 @@ static void __init free_node(struct device_node *np)
kfree(np);
}
-static int __init add_string_property(struct device_node *np, const char *name,
+static int add_string_property(struct device_node *np, const char *name,
const char *value)
{
struct property *nprop = new_property(name, strlen(value) + 1, value);
@@ -124,7 +126,7 @@ static int __init add_string_property(struct device_node *np, const char *name,
return 1;
}
-static int __init add_raw_property(struct device_node *np, const char *name,
+static int add_raw_property(struct device_node *np, const char *name,
int length, const void *value)
{
struct property *nprop = new_property(name, length, value);
@@ -135,6 +137,203 @@ static int __init add_raw_property(struct device_node *np, const char *name,
return 1;
}
+static struct device_node *do_device_node(struct device_node *parent,
+ const char *name, u32 reg, u32 unit, const char *type,
+ const char *compat, struct vio_resource *res)
+{
+ struct device_node *np;
+ char path[32];
+
+ snprintf(path, sizeof(path), "/vdevice/%s@%08x", name, reg);
+ np = new_node(path, parent);
+ if (!np)
+ return NULL;
+ if (!add_string_property(np, "name", name) ||
+ !add_string_property(np, "device_type", type) ||
+ !add_string_property(np, "compatible", compat) ||
+ !add_raw_property(np, "reg", sizeof(reg), ®) ||
+ !add_raw_property(np, "linux,unit_address",
+ sizeof(unit), &unit)) {
+ goto node_free;
+ }
+ if (res) {
+ if (!add_raw_property(np, "linux,vio_rsrcname",
+ sizeof(res->rsrcname), res->rsrcname) ||
+ !add_raw_property(np, "linux,vio_type",
+ sizeof(res->type), res->type) ||
+ !add_raw_property(np, "linux,vio_model",
+ sizeof(res->model), res->model))
+ goto node_free;
+ }
+ np->name = of_get_property(np, "name", NULL);
+ np->type = of_get_property(np, "device_type", NULL);
+ of_attach_node(np);
+#ifdef CONFIG_PROC_DEVICETREE
+ if (parent->pde) {
+ struct proc_dir_entry *ent;
+
+ ent = proc_mkdir(strrchr(np->full_name, '/') + 1, parent->pde);
+ if (ent)
+ proc_device_tree_add_node(np, ent);
+ }
+#endif
+ return np;
+
+ node_free:
+ free_node(np);
+ return NULL;
+}
+
+#if defined(CONFIG_VIODASD) || defined(CONFIG_VIODASD_MODULE)
+/*
+ * This is here so that we can dynamically add viodasd
+ * devices without exposing all the above infrastructure.
+ */
+struct vio_dev *vio_create_viodasd(u32 unit)
+{
+ struct device_node *vio_root;
+ struct device_node *np;
+ struct vio_dev *vdev = NULL;
+
+ vio_root = of_find_node_by_path("/vdevice");
+ if (!vio_root)
+ return NULL;
+ np = do_device_node(vio_root, "viodasd", FIRST_VIODASD + unit, unit,
+ "block", "IBM,iSeries-viodasd", NULL);
+ of_node_put(vio_root);
+ if (np) {
+ vdev = vio_register_device_node(np);
+ if (!vdev)
+ free_node(np);
+ }
+ return vdev;
+}
+EXPORT_SYMBOL_GPL(vio_create_viodasd);
+
+static void __init handle_block_event(struct HvLpEvent *event)
+{
+ struct vioblocklpevent *bevent = (struct vioblocklpevent *)event;
+ struct vio_waitevent *pwe;
+
+ if (event == NULL)
+ /* Notification that a partition went away! */
+ return;
+ /* First, we should NEVER get an int here...only acks */
+ if (hvlpevent_is_int(event)) {
+ printk(KERN_WARNING "handle_viod_request: "
+ "Yikes! got an int in viodasd event handler!\n");
+ if (hvlpevent_need_ack(event)) {
+ event->xRc = HvLpEvent_Rc_InvalidSubtype;
+ HvCallEvent_ackLpEvent(event);
+ }
+ return;
+ }
+
+ switch (event->xSubtype & VIOMINOR_SUBTYPE_MASK) {
+ case vioblockopen:
+ /*
+ * Handle a response to an open request. We get all the
+ * disk information in the response, so update it. The
+ * correlation token contains a pointer to a waitevent
+ * structure that has a completion in it. update the
+ * return code in the waitevent structure and post the
+ * completion to wake up the guy who sent the request
+ */
+ pwe = (struct vio_waitevent *)event->xCorrelationToken;
+ pwe->rc = event->xRc;
+ pwe->sub_result = bevent->sub_result;
+ complete(&pwe->com);
+ break;
+ case vioblockclose:
+ break;
+ default:
+ printk(KERN_WARNING "handle_viod_request: unexpected subtype!");
+ if (hvlpevent_need_ack(event)) {
+ event->xRc = HvLpEvent_Rc_InvalidSubtype;
+ HvCallEvent_ackLpEvent(event);
+ }
+ }
+}
+
+static void __init probe_disk(struct device_node *vio_root, u32 unit)
+{
+ HvLpEvent_Rc hvrc;
+ struct vio_waitevent we;
+ u16 flags = 0;
+
+retry:
+ init_completion(&we.com);
+
+ /* Send the open event to OS/400 */
+ hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp,
+ HvLpEvent_Type_VirtualIo,
+ viomajorsubtype_blockio | vioblockopen,
+ HvLpEvent_AckInd_DoAck, HvLpEvent_AckType_ImmediateAck,
+ viopath_sourceinst(viopath_hostLp),
+ viopath_targetinst(viopath_hostLp),
+ (u64)(unsigned long)&we, VIOVERSION << 16,
+ ((u64)unit << 48) | ((u64)flags<< 32),
+ 0, 0, 0);
+ if (hvrc != 0) {
+ printk(KERN_WARNING "probe_disk: bad rc on HV open %d\n",
+ (int)hvrc);
+ return;
+ }
+
+ wait_for_completion(&we.com);
+
+ if (we.rc != 0) {
+ if (flags != 0)
+ return;
+ /* try again with read only flag set */
+ flags = vioblockflags_ro;
+ goto retry;
+ }
+
+ /* Send the close event to OS/400. We DON'T expect a response */
+ hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp,
+ HvLpEvent_Type_VirtualIo,
+ viomajorsubtype_blockio | vioblockclose,
+ HvLpEvent_AckInd_NoAck, HvLpEvent_AckType_ImmediateAck,
+ viopath_sourceinst(viopath_hostLp),
+ viopath_targetinst(viopath_hostLp),
+ 0, VIOVERSION << 16,
+ ((u64)unit << 48) | ((u64)flags << 32),
+ 0, 0, 0);
+ if (hvrc != 0) {
+ printk(KERN_WARNING "probe_disk: "
+ "bad rc sending event to OS/400 %d\n", (int)hvrc);
+ return;
+ }
+
+ do_device_node(vio_root, "viodasd", FIRST_VIODASD + unit, unit,
+ "block", "IBM,iSeries-viodasd", NULL);
+}
+
+static void __init get_viodasd_info(struct device_node *vio_root)
+{
+ int rc;
+ u32 unit;
+
+ rc = viopath_open(viopath_hostLp, viomajorsubtype_blockio, 2);
+ if (rc) {
+ printk(KERN_WARNING "get_viodasd_info: "
+ "error opening path to host partition %d\n",
+ viopath_hostLp);
+ return;
+ }
+
+ /* Initialize our request handler */
+ vio_setHandler(viomajorsubtype_blockio, handle_block_event);
+
+ for (unit = 0; unit < HVMAXARCHITECTEDVIRTUALDISKS; unit++)
+ probe_disk(vio_root, unit);
+
+ vio_clearHandler(viomajorsubtype_blockio);
+ viopath_close(viopath_hostLp, viomajorsubtype_blockio, 2);
+}
+#endif /* defined(CONFIG_VIODASD) || defined(CONFIG_VIODASD_MODULE) */
+
#if defined(CONFIG_VIOCD) || defined(CONFIG_VIOCD_MODULE)
static void __init handle_cd_event(struct HvLpEvent *event)
{
@@ -234,49 +433,9 @@ static void __init get_viocd_info(struct device_node *vio_root)
for (unit = 0; (unit < HVMAXARCHITECTEDVIRTUALCDROMS) &&
unitinfo[unit].rsrcname[0]; unit++) {
- struct device_node *np;
- char name[64];
- u32 reg = FIRST_VIOCD + unit;
-
- snprintf(name, sizeof(name), "/vdevice/viocd@%08x", reg);
- np = new_node(name, vio_root);
- if (!np)
- goto hv_free;
- if (!add_string_property(np, "name", "viocd") ||
- !add_string_property(np, "device_type", "block") ||
- !add_string_property(np, "compatible",
- "IBM,iSeries-viocd") ||
- !add_raw_property(np, "reg", sizeof(reg), ®) ||
- !add_raw_property(np, "linux,unit_address",
- sizeof(unit), &unit) ||
- !add_raw_property(np, "linux,vio_rsrcname",
- sizeof(unitinfo[unit].rsrcname),
- unitinfo[unit].rsrcname) ||
- !add_raw_property(np, "linux,vio_type",
- sizeof(unitinfo[unit].type),
- unitinfo[unit].type) ||
- !add_raw_property(np, "linux,vio_model",
- sizeof(unitinfo[unit].model),
- unitinfo[unit].model))
- goto node_free;
- np->name = of_get_property(np, "name", NULL);
- np->type = of_get_property(np, "device_type", NULL);
- of_attach_node(np);
-#ifdef CONFIG_PROC_DEVICETREE
- if (vio_root->pde) {
- struct proc_dir_entry *ent;
-
- ent = proc_mkdir(strrchr(np->full_name, '/') + 1,
- vio_root->pde);
- if (ent)
- proc_device_tree_add_node(np, ent);
- }
-#endif
- continue;
-
- node_free:
- free_node(np);
- break;
+ if (!do_device_node(vio_root, "viocd", FIRST_VIOCD + unit, unit,
+ "block", "IBM,iSeries-viocd", &unitinfo[unit]))
+ break;
}
hv_free:
@@ -353,49 +512,10 @@ static void __init get_viotape_info(struct device_node *vio_root)
for (unit = 0; (unit < HVMAXARCHITECTEDVIRTUALTAPES) &&
unitinfo[unit].rsrcname[0]; unit++) {
- struct device_node *np;
- char name[64];
- u32 reg = FIRST_VIOTAPE + unit;
-
- snprintf(name, sizeof(name), "/vdevice/viotape@%08x", reg);
- np = new_node(name, vio_root);
- if (!np)
- goto hv_free;
- if (!add_string_property(np, "name", "viotape") ||
- !add_string_property(np, "device_type", "byte") ||
- !add_string_property(np, "compatible",
- "IBM,iSeries-viotape") ||
- !add_raw_property(np, "reg", sizeof(reg), ®) ||
- !add_raw_property(np, "linux,unit_address",
- sizeof(unit), &unit) ||
- !add_raw_property(np, "linux,vio_rsrcname",
- sizeof(unitinfo[unit].rsrcname),
- unitinfo[unit].rsrcname) ||
- !add_raw_property(np, "linux,vio_type",
- sizeof(unitinfo[unit].type),
- unitinfo[unit].type) ||
- !add_raw_property(np, "linux,vio_model",
- sizeof(unitinfo[unit].model),
- unitinfo[unit].model))
- goto node_free;
- np->name = of_get_property(np, "name", NULL);
- np->type = of_get_property(np, "device_type", NULL);
- of_attach_node(np);
-#ifdef CONFIG_PROC_DEVICETREE
- if (vio_root->pde) {
- struct proc_dir_entry *ent;
-
- ent = proc_mkdir(strrchr(np->full_name, '/') + 1,
- vio_root->pde);
- if (ent)
- proc_device_tree_add_node(np, ent);
- }
-#endif
- continue;
-
- node_free:
- free_node(np);
- break;
+ if (!do_device_node(vio_root, "viotape", FIRST_VIOTAPE + unit,
+ unit, "byte", "IBM,iSeries-viotape",
+ &unitinfo[unit]))
+ break;
}
hv_free:
@@ -426,6 +546,9 @@ static int __init iseries_vio_init(void)
goto put_node;
}
+#if defined(CONFIG_VIODASD) || defined(CONFIG_VIODASD_MODULE)
+ get_viodasd_info(vio_root);
+#endif
#if defined(CONFIG_VIOCD) || defined(CONFIG_VIOCD_MODULE)
get_viocd_info(vio_root);
#endif
diff --git a/drivers/block/viodasd.c b/drivers/block/viodasd.c
index af3969a..e824b67 100644
--- a/drivers/block/viodasd.c
+++ b/drivers/block/viodasd.c
@@ -74,53 +74,9 @@ enum {
static DEFINE_SPINLOCK(viodasd_spinlock);
#define VIOMAXREQ 16
-#define VIOMAXBLOCKDMA 12
#define DEVICE_NO(cell) ((struct viodasd_device *)(cell) - &viodasd_devices[0])
-struct open_data {
- u64 disk_size;
- u16 max_disk;
- u16 cylinders;
- u16 tracks;
- u16 sectors;
- u16 bytes_per_sector;
-};
-
-struct rw_data {
- u64 offset;
- struct {
- u32 token;
- u32 reserved;
- u64 len;
- } dma_info[VIOMAXBLOCKDMA];
-};
-
-struct vioblocklpevent {
- struct HvLpEvent event;
- u32 reserved;
- u16 version;
- u16 sub_result;
- u16 disk;
- u16 flags;
- union {
- struct open_data open_data;
- struct rw_data rw_data;
- u64 changed;
- } u;
-};
-
-#define vioblockflags_ro 0x0001
-
-enum vioblocksubtype {
- vioblockopen = 0x0001,
- vioblockclose = 0x0002,
- vioblockread = 0x0003,
- vioblockwrite = 0x0004,
- vioblockflush = 0x0005,
- vioblockcheck = 0x0007
-};
-
struct viodasd_waitevent {
struct completion com;
int rc;
@@ -429,7 +385,7 @@ static void do_viodasd_request(struct request_queue *q)
* Probe a single disk and fill in the viodasd_device structure
* for it.
*/
-static void probe_disk(struct viodasd_device *d)
+static int probe_disk(struct viodasd_device *d)
{
HvLpEvent_Rc hvrc;
struct viodasd_waitevent we;
@@ -453,14 +409,14 @@ retry:
0, 0, 0);
if (hvrc != 0) {
printk(VIOD_KERN_WARNING "bad rc on HV open %d\n", (int)hvrc);
- return;
+ return 0;
}
wait_for_completion(&we.com);
if (we.rc != 0) {
if (flags != 0)
- return;
+ return 0;
/* try again with read only flag set */
flags = vioblockflags_ro;
goto retry;
@@ -490,15 +446,32 @@ retry:
if (hvrc != 0) {
printk(VIOD_KERN_WARNING
"bad rc sending event to OS/400 %d\n", (int)hvrc);
- return;
+ return 0;
}
+
+ if (d->dev == NULL) {
+ /* this is when we reprobe for new disks */
+ if (vio_create_viodasd(dev_no) == NULL) {
+ printk(VIOD_KERN_WARNING
+ "cannot allocate virtual device for disk %d\n",
+ dev_no);
+ return 0;
+ }
+ /*
+ * The vio_create_viodasd will have recursed into this
+ * routine with d->dev set to the new vio device and
+ * will finish the setup of the disk below.
+ */
+ return 1;
+ }
+
/* create the request queue for the disk */
spin_lock_init(&d->q_lock);
q = blk_init_queue(do_viodasd_request, &d->q_lock);
if (q == NULL) {
printk(VIOD_KERN_WARNING "cannot allocate queue for disk %d\n",
dev_no);
- return;
+ return 0;
}
g = alloc_disk(1 << PARTITION_SHIFT);
if (g == NULL) {
@@ -506,7 +479,7 @@ retry:
"cannot allocate disk structure for disk %d\n",
dev_no);
blk_cleanup_queue(q);
- return;
+ return 0;
}
d->disk = g;
@@ -538,6 +511,7 @@ retry:
/* register us in the global list */
add_disk(g);
+ return 1;
}
/* returns the total number of scatterlist elements converted */
@@ -718,8 +692,7 @@ static int viodasd_probe(struct vio_dev *vdev, const struct vio_device_id *id)
struct viodasd_device *d = &viodasd_devices[vdev->unit_address];
d->dev = &vdev->dev;
- probe_disk(d);
- if (d->disk == NULL)
+ if (!probe_disk(d))
return -ENODEV;
return 0;
}
diff --git a/include/asm-powerpc/iseries/vio.h b/include/asm-powerpc/iseries/vio.h
index 2555dfd..f9ac0d0 100644
--- a/include/asm-powerpc/iseries/vio.h
+++ b/include/asm-powerpc/iseries/vio.h
@@ -51,6 +51,51 @@
*/
#define VIO_MAX_SUBTYPES 8
+#define VIOMAXBLOCKDMA 12
+
+struct open_data {
+ u64 disk_size;
+ u16 max_disk;
+ u16 cylinders;
+ u16 tracks;
+ u16 sectors;
+ u16 bytes_per_sector;
+};
+
+struct rw_data {
+ u64 offset;
+ struct {
+ u32 token;
+ u32 reserved;
+ u64 len;
+ } dma_info[VIOMAXBLOCKDMA];
+};
+
+struct vioblocklpevent {
+ struct HvLpEvent event;
+ u32 reserved;
+ u16 version;
+ u16 sub_result;
+ u16 disk;
+ u16 flags;
+ union {
+ struct open_data open_data;
+ struct rw_data rw_data;
+ u64 changed;
+ } u;
+};
+
+#define vioblockflags_ro 0x0001
+
+enum vioblocksubtype {
+ vioblockopen = 0x0001,
+ vioblockclose = 0x0002,
+ vioblockread = 0x0003,
+ vioblockwrite = 0x0004,
+ vioblockflush = 0x0005,
+ vioblockcheck = 0x0007
+};
+
struct viocdlpevent {
struct HvLpEvent event;
u32 reserved;
@@ -133,6 +178,8 @@ extern void vio_set_hostlp(void);
extern void *vio_get_event_buffer(int subtype);
extern void vio_free_event_buffer(int subtype, void *buffer);
+extern struct vio_dev *vio_create_viodasd(u32 unit);
+
extern HvLpIndex viopath_hostLp;
extern HvLpIndex viopath_ourLp;
--
1.5.3.2
^ permalink raw reply related
* [PATCH 10/11] [POWERPC] Clean up vio.h
From: Stephen Rothwell @ 2007-09-21 4:39 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143841.37ad2a2f.sfr@canb.auug.org.au>
Remove vio_dma_ops declaration (since it no longer exists) and some
unused fields from struct vio_driver.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/kernel/vio.c | 11 -----------
include/asm-powerpc/vio.h | 5 -----
2 files changed, 0 insertions(+), 16 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index aec5fff..cb22a35 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -135,16 +135,6 @@ static int vio_bus_remove(struct device *dev)
return 1;
}
-/* convert from struct device to struct vio_dev and pass to driver. */
-static void vio_bus_shutdown(struct device *dev)
-{
- struct vio_dev *viodev = to_vio_dev(dev);
- struct vio_driver *viodrv = to_vio_driver(dev->driver);
-
- if (dev->driver && viodrv->shutdown)
- viodrv->shutdown(viodev);
-}
-
/**
* vio_register_driver: - Register a new vio driver
* @drv: The vio_driver structure to be registered.
@@ -361,7 +351,6 @@ static struct bus_type vio_bus_type = {
.match = vio_bus_match,
.probe = vio_bus_probe,
.remove = vio_bus_remove,
- .shutdown = vio_bus_shutdown,
};
/**
diff --git a/include/asm-powerpc/vio.h b/include/asm-powerpc/vio.h
index 598d111..9204c15 100644
--- a/include/asm-powerpc/vio.h
+++ b/include/asm-powerpc/vio.h
@@ -53,17 +53,12 @@ struct vio_dev {
};
struct vio_driver {
- struct list_head node;
const struct vio_device_id *id_table;
int (*probe)(struct vio_dev *dev, const struct vio_device_id *id);
int (*remove)(struct vio_dev *dev);
- void (*shutdown)(struct vio_dev *dev);
- unsigned long driver_data;
struct device_driver driver;
};
-extern struct dma_mapping_ops vio_dma_ops;
-
extern int vio_register_driver(struct vio_driver *drv);
extern void vio_unregister_driver(struct vio_driver *drv);
--
1.5.3.2
^ permalink raw reply related
* [PATCH 09/11] [POWERPC] iSeries: move detection of virtual tapes
From: Stephen Rothwell @ 2007-09-21 4:38 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143750.4114b0b5.sfr@canb.auug.org.au>
Now we will only have entries in the device tree for the actual existing
devices (including their OS/400 properties). This way viotape.c gets
all the information about the devices from the device tree.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/platforms/iseries/dt.c | 7 --
arch/powerpc/platforms/iseries/vio.c | 153 ++++++++++++++++++++++++++++++----
drivers/char/viotape.c | 124 ++++-----------------------
include/asm-powerpc/iseries/vio.h | 41 +++++++++
4 files changed, 196 insertions(+), 129 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/platforms/iseries/dt.c b/arch/powerpc/platforms/iseries/dt.c
index 84fcee1..2e4ad6b 100644
--- a/arch/powerpc/platforms/iseries/dt.c
+++ b/arch/powerpc/platforms/iseries/dt.c
@@ -73,7 +73,6 @@ static char __initdata device_type_memory[] = "memory";
static char __initdata device_type_serial[] = "serial";
static char __initdata device_type_network[] = "network";
static char __initdata device_type_block[] = "block";
-static char __initdata device_type_byte[] = "byte";
static char __initdata device_type_pci[] = "pci";
static char __initdata device_type_vdevice[] = "vdevice";
static char __initdata device_type_vscsi[] = "vscsi";
@@ -380,12 +379,6 @@ static void __init dt_vdevices(struct iseries_flat_dt *dt)
for (i = 0; i < HVMAXARCHITECTEDVIRTUALDISKS; i++)
dt_do_vdevice(dt, "viodasd", reg, i, device_type_block,
"IBM,iSeries-viodasd", 1);
- reg += HVMAXARCHITECTEDVIRTUALDISKS;
- reg += HVMAXARCHITECTEDVIRTUALCDROMS;
-
- for (i = 0; i < HVMAXARCHITECTEDVIRTUALTAPES; i++)
- dt_do_vdevice(dt, "viotape", reg, i, device_type_byte,
- "IBM,iSeries-viotape", 1);
dt_end_node(dt);
}
diff --git a/arch/powerpc/platforms/iseries/vio.c b/arch/powerpc/platforms/iseries/vio.c
index ad36d89..c7d2a83 100644
--- a/arch/powerpc/platforms/iseries/vio.c
+++ b/arch/powerpc/platforms/iseries/vio.c
@@ -45,6 +45,18 @@
#define FIRST_VIOTAPE (FIRST_VIOCD + NUM_VIOCDS)
#define NUM_VIOTAPES HVMAXARCHITECTEDVIRTUALTAPES
+struct vio_waitevent {
+ struct completion com;
+ int rc;
+ u16 sub_result;
+};
+
+struct vio_resource {
+ char rsrcname[10];
+ char type[4];
+ char model[3];
+};
+
static struct property * __init new_property(const char *name, int length,
const void *value)
{
@@ -124,22 +136,10 @@ static int __init add_raw_property(struct device_node *np, const char *name,
}
#if defined(CONFIG_VIOCD) || defined(CONFIG_VIOCD_MODULE)
-struct viocd_waitevent {
- struct completion com;
- int rc;
- u16 sub_result;
-};
-
-struct cdrom_info {
- char rsrcname[10];
- char type[4];
- char model[3];
-};
-
static void __init handle_cd_event(struct HvLpEvent *event)
{
struct viocdlpevent *bevent;
- struct viocd_waitevent *pwe;
+ struct vio_waitevent *pwe;
if (!event)
/* Notification that a partition went away! */
@@ -159,7 +159,7 @@ static void __init handle_cd_event(struct HvLpEvent *event)
switch (event->xSubtype & VIOMINOR_SUBTYPE_MASK) {
case viocdgetinfo:
- pwe = (struct viocd_waitevent *)event->xCorrelationToken;
+ pwe = (struct vio_waitevent *)event->xCorrelationToken;
pwe->rc = event->xRc;
pwe->sub_result = bevent->sub_result;
complete(&pwe->com);
@@ -180,8 +180,8 @@ static void __init get_viocd_info(struct device_node *vio_root)
{
HvLpEvent_Rc hvrc;
u32 unit;
- struct viocd_waitevent we;
- struct cdrom_info *unitinfo;
+ struct vio_waitevent we;
+ struct vio_resource *unitinfo;
dma_addr_t unitinfo_dmaaddr;
int ret;
@@ -288,6 +288,124 @@ static void __init get_viocd_info(struct device_node *vio_root)
}
#endif /* defined(CONFIG_VIOCD) || defined(CONFIG_VIOCD_MODULE) */
+#if defined(CONFIG_VIOTAPE) || defined(CONFIG_VIOTAPE_MODULE)
+/* Handle interrupt events for tape */
+static void __init handle_tape_event(struct HvLpEvent *event)
+{
+ struct vio_waitevent *we;
+ struct viotapelpevent *tevent = (struct viotapelpevent *)event;
+
+ if (event == NULL)
+ /* Notification that a partition went away! */
+ return;
+
+ we = (struct vio_waitevent *)event->xCorrelationToken;
+ switch (event->xSubtype & VIOMINOR_SUBTYPE_MASK) {
+ case viotapegetinfo:
+ we->rc = tevent->sub_type_result;
+ complete(&we->com);
+ break;
+ default:
+ printk(KERN_WARNING "handle_tape_event: weird ack\n");
+ }
+}
+
+static void __init get_viotape_info(struct device_node *vio_root)
+{
+ HvLpEvent_Rc hvrc;
+ u32 unit;
+ struct vio_resource *unitinfo;
+ dma_addr_t unitinfo_dmaaddr;
+ size_t len = sizeof(*unitinfo) * HVMAXARCHITECTEDVIRTUALTAPES;
+ struct vio_waitevent we;
+ int ret;
+
+ ret = viopath_open(viopath_hostLp, viomajorsubtype_tape, 2);
+ if (ret) {
+ printk(KERN_WARNING "get_viotape_info: "
+ "error on viopath_open to hostlp %d\n", ret);
+ return;
+ }
+
+ vio_setHandler(viomajorsubtype_tape, handle_tape_event);
+
+ unitinfo = iseries_hv_alloc(len, &unitinfo_dmaaddr, GFP_ATOMIC);
+ if (!unitinfo)
+ goto clear_handler;
+
+ memset(unitinfo, 0, len);
+
+ hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp,
+ HvLpEvent_Type_VirtualIo,
+ viomajorsubtype_tape | viotapegetinfo,
+ HvLpEvent_AckInd_DoAck, HvLpEvent_AckType_ImmediateAck,
+ viopath_sourceinst(viopath_hostLp),
+ viopath_targetinst(viopath_hostLp),
+ (u64)(unsigned long)&we, VIOVERSION << 16,
+ unitinfo_dmaaddr, len, 0, 0);
+ if (hvrc != HvLpEvent_Rc_Good) {
+ printk(KERN_WARNING "get_viotape_info: hv error on op %d\n",
+ (int)hvrc);
+ goto hv_free;
+ }
+
+ wait_for_completion(&we.com);
+
+ for (unit = 0; (unit < HVMAXARCHITECTEDVIRTUALTAPES) &&
+ unitinfo[unit].rsrcname[0]; unit++) {
+ struct device_node *np;
+ char name[64];
+ u32 reg = FIRST_VIOTAPE + unit;
+
+ snprintf(name, sizeof(name), "/vdevice/viotape@%08x", reg);
+ np = new_node(name, vio_root);
+ if (!np)
+ goto hv_free;
+ if (!add_string_property(np, "name", "viotape") ||
+ !add_string_property(np, "device_type", "byte") ||
+ !add_string_property(np, "compatible",
+ "IBM,iSeries-viotape") ||
+ !add_raw_property(np, "reg", sizeof(reg), ®) ||
+ !add_raw_property(np, "linux,unit_address",
+ sizeof(unit), &unit) ||
+ !add_raw_property(np, "linux,vio_rsrcname",
+ sizeof(unitinfo[unit].rsrcname),
+ unitinfo[unit].rsrcname) ||
+ !add_raw_property(np, "linux,vio_type",
+ sizeof(unitinfo[unit].type),
+ unitinfo[unit].type) ||
+ !add_raw_property(np, "linux,vio_model",
+ sizeof(unitinfo[unit].model),
+ unitinfo[unit].model))
+ goto node_free;
+ np->name = of_get_property(np, "name", NULL);
+ np->type = of_get_property(np, "device_type", NULL);
+ of_attach_node(np);
+#ifdef CONFIG_PROC_DEVICETREE
+ if (vio_root->pde) {
+ struct proc_dir_entry *ent;
+
+ ent = proc_mkdir(strrchr(np->full_name, '/') + 1,
+ vio_root->pde);
+ if (ent)
+ proc_device_tree_add_node(np, ent);
+ }
+#endif
+ continue;
+
+ node_free:
+ free_node(np);
+ break;
+ }
+
+ hv_free:
+ iseries_hv_free(len, unitinfo, unitinfo_dmaaddr);
+ clear_handler:
+ vio_clearHandler(viomajorsubtype_tape);
+ viopath_close(viopath_hostLp, viomajorsubtype_tape, 2);
+}
+#endif /* defined(CONFIG_VIOTAPE) || defined(CONFIG_VIOTAPE_MODULE) */
+
static int __init iseries_vio_init(void)
{
struct device_node *vio_root;
@@ -311,6 +429,9 @@ static int __init iseries_vio_init(void)
#if defined(CONFIG_VIOCD) || defined(CONFIG_VIOCD_MODULE)
get_viocd_info(vio_root);
#endif
+#if defined(CONFIG_VIOTAPE) || defined(CONFIG_VIOTAPE_MODULE)
+ get_viotape_info(vio_root);
+#endif
return 0;
diff --git a/drivers/char/viotape.c b/drivers/char/viotape.c
index 064c091..f1d60f0 100644
--- a/drivers/char/viotape.c
+++ b/drivers/char/viotape.c
@@ -92,47 +92,6 @@ struct viot_devinfo_struct {
#define VIOTAPOP_SETPART 14
#define VIOTAPOP_UNLOAD 15
-struct viotapelpevent {
- struct HvLpEvent event;
- u32 reserved;
- u16 version;
- u16 sub_type_result;
- u16 tape;
- u16 flags;
- u32 token;
- u64 len;
- union {
- struct {
- u32 tape_op;
- u32 count;
- } op;
- struct {
- u32 type;
- u32 resid;
- u32 dsreg;
- u32 gstat;
- u32 erreg;
- u32 file_no;
- u32 block_no;
- } get_status;
- struct {
- u32 block_no;
- } get_pos;
- } u;
-};
-
-enum viotapesubtype {
- viotapeopen = 0x0001,
- viotapeclose = 0x0002,
- viotaperead = 0x0003,
- viotapewrite = 0x0004,
- viotapegetinfo = 0x0005,
- viotapeop = 0x0006,
- viotapegetpos = 0x0007,
- viotapesetpos = 0x0008,
- viotapegetstatus = 0x0009
-};
-
enum viotaperc {
viotape_InvalidRange = 0x0601,
viotape_InvalidToken = 0x0602,
@@ -223,14 +182,11 @@ static const struct vio_error_entry viotape_err_table[] = {
#define VIOT_WRITING 2
/* Our info on the tapes */
-struct tape_descr {
- char rsrcname[10];
- char type[4];
- char model[3];
-};
-
-static struct tape_descr *viotape_unitinfo;
-static dma_addr_t viotape_unitinfo_token;
+static struct {
+ const char *rsrcname;
+ const char *type;
+ const char *model;
+} viotape_unitinfo[VIOTAPE_MAX_TAPE];
static struct mtget viomtget[VIOTAPE_MAX_TAPE];
@@ -381,53 +337,6 @@ int tape_rc_to_errno(int tape_rc, char *operation, int tapeno)
return -err->errno;
}
-/* Get info on all tapes from OS/400 */
-static int get_viotape_info(void)
-{
- HvLpEvent_Rc hvrc;
- int i;
- size_t len = sizeof(*viotape_unitinfo) * VIOTAPE_MAX_TAPE;
- struct op_struct *op = get_op_struct();
-
- if (op == NULL)
- return -ENOMEM;
-
- viotape_unitinfo = iseries_hv_alloc(len, &viotape_unitinfo_token,
- GFP_ATOMIC);
- if (viotape_unitinfo == NULL) {
- free_op_struct(op);
- return -ENOMEM;
- }
-
- memset(viotape_unitinfo, 0, len);
-
- hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp,
- HvLpEvent_Type_VirtualIo,
- viomajorsubtype_tape | viotapegetinfo,
- HvLpEvent_AckInd_DoAck, HvLpEvent_AckType_ImmediateAck,
- viopath_sourceinst(viopath_hostLp),
- viopath_targetinst(viopath_hostLp),
- (u64) (unsigned long) op, VIOVERSION << 16,
- viotape_unitinfo_token, len, 0, 0);
- if (hvrc != HvLpEvent_Rc_Good) {
- printk(VIOTAPE_KERN_WARN "hv error on op %d\n",
- (int)hvrc);
- free_op_struct(op);
- return -EIO;
- }
-
- wait_for_completion(&op->com);
-
- free_op_struct(op);
-
- for (i = 0;
- ((i < VIOTAPE_MAX_TAPE) && (viotape_unitinfo[i].rsrcname[0]));
- i++)
- viotape_numdev++;
- return 0;
-}
-
-
/* Write */
static ssize_t viotap_write(struct file *file, const char *buf,
size_t count, loff_t * ppos)
@@ -899,7 +808,6 @@ static void vioHandleTapeEvent(struct HvLpEvent *event)
tapeminor = event->xSubtype & VIOMINOR_SUBTYPE_MASK;
op = (struct op_struct *)event->xCorrelationToken;
switch (tapeminor) {
- case viotapegetinfo:
case viotapeopen:
case viotapeclose:
op->rc = tevent->sub_type_result;
@@ -942,11 +850,23 @@ static int viotape_probe(struct vio_dev *vdev, const struct vio_device_id *id)
{
int i = vdev->unit_address;
int j;
+ struct device_node *node = vdev->dev.archdata.of_node;
- if (i >= viotape_numdev)
+ if (i > VIOTAPE_MAX_TAPE)
+ return -ENODEV;
+ if (!node)
return -ENODEV;
+ if (i >= viotape_numdev)
+ viotape_numdev = i + 1;
+
tape_device[i] = &vdev->dev;
+ viotape_unitinfo[i].rsrcname = of_get_property(node,
+ "linux,vio_rsrcname", NULL);
+ viotape_unitinfo[i].type = of_get_property(node, "linux,vio_type",
+ NULL);
+ viotape_unitinfo[i].model = of_get_property(node, "linux,vio_model",
+ NULL);
state[i].cur_part = 0;
for (j = 0; j < MAX_PARTITIONS; ++j)
@@ -1044,11 +964,6 @@ int __init viotap_init(void)
goto unreg_chrdev;
}
- if ((ret = get_viotape_info()) < 0) {
- printk(VIOTAPE_KERN_WARN "Unable to obtain virtual device information");
- goto unreg_class;
- }
-
ret = vio_register_driver(&viotape_driver);
if (ret)
goto unreg_class;
@@ -1102,9 +1017,6 @@ static void __exit viotap_exit(void)
vio_unregister_driver(&viotape_driver);
class_destroy(tape_class);
unregister_chrdev(VIOTAPE_MAJOR, "viotape");
- if (viotape_unitinfo)
- iseries_hv_free(sizeof(viotape_unitinfo[0]) * VIOTAPE_MAX_TAPE,
- viotape_unitinfo, viotape_unitinfo_token);
viopath_close(viopath_hostLp, viomajorsubtype_tape, VIOTAPE_MAXREQ + 2);
vio_clearHandler(viomajorsubtype_tape);
clear_op_struct_pool();
diff --git a/include/asm-powerpc/iseries/vio.h b/include/asm-powerpc/iseries/vio.h
index e5a405b..2555dfd 100644
--- a/include/asm-powerpc/iseries/vio.h
+++ b/include/asm-powerpc/iseries/vio.h
@@ -75,6 +75,47 @@ enum viocdsubtype {
viocdcheck = 0x0007
};
+struct viotapelpevent {
+ struct HvLpEvent event;
+ u32 reserved;
+ u16 version;
+ u16 sub_type_result;
+ u16 tape;
+ u16 flags;
+ u32 token;
+ u64 len;
+ union {
+ struct {
+ u32 tape_op;
+ u32 count;
+ } op;
+ struct {
+ u32 type;
+ u32 resid;
+ u32 dsreg;
+ u32 gstat;
+ u32 erreg;
+ u32 file_no;
+ u32 block_no;
+ } get_status;
+ struct {
+ u32 block_no;
+ } get_pos;
+ } u;
+};
+
+enum viotapesubtype {
+ viotapeopen = 0x0001,
+ viotapeclose = 0x0002,
+ viotaperead = 0x0003,
+ viotapewrite = 0x0004,
+ viotapegetinfo = 0x0005,
+ viotapeop = 0x0006,
+ viotapegetpos = 0x0007,
+ viotapesetpos = 0x0008,
+ viotapegetstatus = 0x0009
+};
+
/*
* Each subtype can register a handler to process their events.
* The handler must have this interface.
--
1.5.3.2
^ permalink raw reply related
* [PATCH 08/11] [POWERPC] iSeries: move detection of virtual cdroms
From: Stephen Rothwell @ 2007-09-21 4:37 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143647.7ceebabc.sfr@canb.auug.org.au>
Now we will only have entries in the device tree for the actual existing
devices (including their OS/400 properties). This way viocd.c gets all
the information about the devices from the device tree.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/kernel/vio.c | 3 -
arch/powerpc/platforms/iseries/Makefile | 2 +-
arch/powerpc/platforms/iseries/dt.c | 4 -
arch/powerpc/platforms/iseries/vio.c | 321 +++++++++++++++++++++++++++++++
drivers/cdrom/viocd.c | 116 ++----------
include/asm-powerpc/iseries/vio.h | 24 +++
6 files changed, 365 insertions(+), 105 deletions(-)
create mode 100644 arch/powerpc/platforms/iseries/vio.c
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index 9e1ff21..aec5fff 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -257,9 +257,6 @@ static int __init vio_bus_init(void)
int err;
struct device_node *node_vroot;
- if (firmware_has_feature(FW_FEATURE_ISERIES))
- iommu_vio_init();
-
err = bus_register(&vio_bus_type);
if (err) {
printk(KERN_ERR "failed to register VIO bus\n");
diff --git a/arch/powerpc/platforms/iseries/Makefile b/arch/powerpc/platforms/iseries/Makefile
index 60db509..a65f1b4 100644
--- a/arch/powerpc/platforms/iseries/Makefile
+++ b/arch/powerpc/platforms/iseries/Makefile
@@ -7,7 +7,7 @@ obj-y += hvlog.o hvlpconfig.o lpardata.o setup.o dt_mod.o mf.o lpevents.o \
hvcall.o proc.o htab.o iommu.o misc.o irq.o
obj-$(CONFIG_PCI) += pci.o vpdinfo.o
obj-$(CONFIG_SMP) += smp.o
-obj-$(CONFIG_VIOPATH) += viopath.o
+obj-$(CONFIG_VIOPATH) += viopath.o vio.o
obj-$(CONFIG_MODULES) += ksyms.o
quiet_cmd_dt_strings = DT_STR $@
diff --git a/arch/powerpc/platforms/iseries/dt.c b/arch/powerpc/platforms/iseries/dt.c
index 9e8a334..84fcee1 100644
--- a/arch/powerpc/platforms/iseries/dt.c
+++ b/arch/powerpc/platforms/iseries/dt.c
@@ -381,10 +381,6 @@ static void __init dt_vdevices(struct iseries_flat_dt *dt)
dt_do_vdevice(dt, "viodasd", reg, i, device_type_block,
"IBM,iSeries-viodasd", 1);
reg += HVMAXARCHITECTEDVIRTUALDISKS;
-
- for (i = 0; i < HVMAXARCHITECTEDVIRTUALCDROMS; i++)
- dt_do_vdevice(dt, "viocd", reg, i, device_type_block,
- "IBM,iSeries-viocd", 1);
reg += HVMAXARCHITECTEDVIRTUALCDROMS;
for (i = 0; i < HVMAXARCHITECTEDVIRTUALTAPES; i++)
diff --git a/arch/powerpc/platforms/iseries/vio.c b/arch/powerpc/platforms/iseries/vio.c
new file mode 100644
index 0000000..ad36d89
--- /dev/null
+++ b/arch/powerpc/platforms/iseries/vio.c
@@ -0,0 +1,321 @@
+/*
+ * Legacy iSeries specific vio initialisation
+ * that needs to be built in (not a module).
+ *
+ * © Copyright 2007 IBM Corporation
+ * Author: Stephen Rothwell
+ * Some parts collected from various other files
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
+#include <linux/of.h>
+#include <linux/init.h>
+#include <linux/gfp.h>
+#include <linux/completion.h>
+#include <linux/proc_fs.h>
+
+#include <asm/firmware.h>
+#include <asm/iseries/vio.h>
+#include <asm/iseries/iommu.h>
+#include <asm/iseries/hv_types.h>
+#include <asm/iseries/hv_lp_event.h>
+
+#define FIRST_VTY 0
+#define NUM_VTYS 1
+#define FIRST_VSCSI (FIRST_VTY + NUM_VTYS)
+#define NUM_VSCSIS 1
+#define FIRST_VLAN (FIRST_VSCSI + NUM_VSCSIS)
+#define NUM_VLANS HVMAXARCHITECTEDVIRTUALLANS
+#define FIRST_VIODASD (FIRST_VLAN + NUM_VLANS)
+#define NUM_VIODASDS HVMAXARCHITECTEDVIRTUALDISKS
+#define FIRST_VIOCD (FIRST_VIODASD + NUM_VIODASDS)
+#define NUM_VIOCDS HVMAXARCHITECTEDVIRTUALCDROMS
+#define FIRST_VIOTAPE (FIRST_VIOCD + NUM_VIOCDS)
+#define NUM_VIOTAPES HVMAXARCHITECTEDVIRTUALTAPES
+
+static struct property * __init new_property(const char *name, int length,
+ const void *value)
+{
+ struct property *np = kzalloc(sizeof(*np) + strlen(name) + 1 + length,
+ GFP_KERNEL);
+
+ if (!np)
+ return NULL;
+ np->name = (char *)(np + 1);
+ np->value = np->name + strlen(name) + 1;
+ strcpy(np->name, name);
+ memcpy(np->value, value, length);
+ np->length = length;
+ return np;
+}
+
+static void __init free_property(struct property *np)
+{
+ kfree(np);
+}
+
+static struct device_node * __init new_node(const char *path,
+ struct device_node *parent)
+{
+ struct device_node *np = kzalloc(sizeof(*np), GFP_KERNEL);
+
+ if (!np)
+ return NULL;
+ np->full_name = kmalloc(strlen(path) + 1, GFP_KERNEL);
+ if (!np->full_name) {
+ kfree(np);
+ return NULL;
+ }
+ strcpy(np->full_name, path);
+ of_node_set_flag(np, OF_DYNAMIC);
+ kref_init(&np->kref);
+ np->parent = of_node_get(parent);
+ return np;
+}
+
+static void __init free_node(struct device_node *np)
+{
+ struct property *next;
+ struct property *prop;
+
+ next = np->properties;
+ while (next) {
+ prop = next;
+ next = prop->next;
+ free_property(prop);
+ }
+ of_node_put(np->parent);
+ kfree(np->full_name);
+ kfree(np);
+}
+
+static int __init add_string_property(struct device_node *np, const char *name,
+ const char *value)
+{
+ struct property *nprop = new_property(name, strlen(value) + 1, value);
+
+ if (!nprop)
+ return 0;
+ prom_add_property(np, nprop);
+ return 1;
+}
+
+static int __init add_raw_property(struct device_node *np, const char *name,
+ int length, const void *value)
+{
+ struct property *nprop = new_property(name, length, value);
+
+ if (!nprop)
+ return 0;
+ prom_add_property(np, nprop);
+ return 1;
+}
+
+#if defined(CONFIG_VIOCD) || defined(CONFIG_VIOCD_MODULE)
+struct viocd_waitevent {
+ struct completion com;
+ int rc;
+ u16 sub_result;
+};
+
+struct cdrom_info {
+ char rsrcname[10];
+ char type[4];
+ char model[3];
+};
+
+static void __init handle_cd_event(struct HvLpEvent *event)
+{
+ struct viocdlpevent *bevent;
+ struct viocd_waitevent *pwe;
+
+ if (!event)
+ /* Notification that a partition went away! */
+ return;
+
+ /* First, we should NEVER get an int here...only acks */
+ if (hvlpevent_is_int(event)) {
+ printk(KERN_WARNING "handle_cd_event: got an unexpected int\n");
+ if (hvlpevent_need_ack(event)) {
+ event->xRc = HvLpEvent_Rc_InvalidSubtype;
+ HvCallEvent_ackLpEvent(event);
+ }
+ return;
+ }
+
+ bevent = (struct viocdlpevent *)event;
+
+ switch (event->xSubtype & VIOMINOR_SUBTYPE_MASK) {
+ case viocdgetinfo:
+ pwe = (struct viocd_waitevent *)event->xCorrelationToken;
+ pwe->rc = event->xRc;
+ pwe->sub_result = bevent->sub_result;
+ complete(&pwe->com);
+ break;
+
+ default:
+ printk(KERN_WARNING "handle_cd_event: "
+ "message with unexpected subtype %0x04X!\n",
+ event->xSubtype & VIOMINOR_SUBTYPE_MASK);
+ if (hvlpevent_need_ack(event)) {
+ event->xRc = HvLpEvent_Rc_InvalidSubtype;
+ HvCallEvent_ackLpEvent(event);
+ }
+ }
+}
+
+static void __init get_viocd_info(struct device_node *vio_root)
+{
+ HvLpEvent_Rc hvrc;
+ u32 unit;
+ struct viocd_waitevent we;
+ struct cdrom_info *unitinfo;
+ dma_addr_t unitinfo_dmaaddr;
+ int ret;
+
+ ret = viopath_open(viopath_hostLp, viomajorsubtype_cdio, 2);
+ if (ret) {
+ printk(KERN_WARNING
+ "get_viocd_info: error opening path to host partition %d\n",
+ viopath_hostLp);
+ return;
+ }
+
+ /* Initialize our request handler */
+ vio_setHandler(viomajorsubtype_cdio, handle_cd_event);
+
+ unitinfo = iseries_hv_alloc(
+ sizeof(*unitinfo) * HVMAXARCHITECTEDVIRTUALCDROMS,
+ &unitinfo_dmaaddr, GFP_ATOMIC);
+ if (!unitinfo) {
+ printk(KERN_WARNING
+ "get_viocd_info: error allocating unitinfo\n");
+ goto clear_handler;
+ }
+
+ memset(unitinfo, 0, sizeof(*unitinfo) * HVMAXARCHITECTEDVIRTUALCDROMS);
+
+ init_completion(&we.com);
+
+ hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp,
+ HvLpEvent_Type_VirtualIo,
+ viomajorsubtype_cdio | viocdgetinfo,
+ HvLpEvent_AckInd_DoAck, HvLpEvent_AckType_ImmediateAck,
+ viopath_sourceinst(viopath_hostLp),
+ viopath_targetinst(viopath_hostLp),
+ (u64)&we, VIOVERSION << 16, unitinfo_dmaaddr, 0,
+ sizeof(*unitinfo) * HVMAXARCHITECTEDVIRTUALCDROMS, 0);
+ if (hvrc != HvLpEvent_Rc_Good) {
+ printk(KERN_WARNING
+ "get_viocd_info: cdrom error sending event. rc %d\n",
+ (int)hvrc);
+ goto hv_free;
+ }
+
+ wait_for_completion(&we.com);
+
+ if (we.rc) {
+ printk(KERN_WARNING "get_viocd_info: bad rc %d:0x%04X\n",
+ we.rc, we.sub_result);
+ goto hv_free;
+ }
+
+ for (unit = 0; (unit < HVMAXARCHITECTEDVIRTUALCDROMS) &&
+ unitinfo[unit].rsrcname[0]; unit++) {
+ struct device_node *np;
+ char name[64];
+ u32 reg = FIRST_VIOCD + unit;
+
+ snprintf(name, sizeof(name), "/vdevice/viocd@%08x", reg);
+ np = new_node(name, vio_root);
+ if (!np)
+ goto hv_free;
+ if (!add_string_property(np, "name", "viocd") ||
+ !add_string_property(np, "device_type", "block") ||
+ !add_string_property(np, "compatible",
+ "IBM,iSeries-viocd") ||
+ !add_raw_property(np, "reg", sizeof(reg), ®) ||
+ !add_raw_property(np, "linux,unit_address",
+ sizeof(unit), &unit) ||
+ !add_raw_property(np, "linux,vio_rsrcname",
+ sizeof(unitinfo[unit].rsrcname),
+ unitinfo[unit].rsrcname) ||
+ !add_raw_property(np, "linux,vio_type",
+ sizeof(unitinfo[unit].type),
+ unitinfo[unit].type) ||
+ !add_raw_property(np, "linux,vio_model",
+ sizeof(unitinfo[unit].model),
+ unitinfo[unit].model))
+ goto node_free;
+ np->name = of_get_property(np, "name", NULL);
+ np->type = of_get_property(np, "device_type", NULL);
+ of_attach_node(np);
+#ifdef CONFIG_PROC_DEVICETREE
+ if (vio_root->pde) {
+ struct proc_dir_entry *ent;
+
+ ent = proc_mkdir(strrchr(np->full_name, '/') + 1,
+ vio_root->pde);
+ if (ent)
+ proc_device_tree_add_node(np, ent);
+ }
+#endif
+ continue;
+
+ node_free:
+ free_node(np);
+ break;
+ }
+
+ hv_free:
+ iseries_hv_free(sizeof(*unitinfo) * HVMAXARCHITECTEDVIRTUALCDROMS,
+ unitinfo, unitinfo_dmaaddr);
+ clear_handler:
+ vio_clearHandler(viomajorsubtype_cdio);
+ viopath_close(viopath_hostLp, viomajorsubtype_cdio, 2);
+}
+#endif /* defined(CONFIG_VIOCD) || defined(CONFIG_VIOCD_MODULE) */
+
+static int __init iseries_vio_init(void)
+{
+ struct device_node *vio_root;
+
+ if (!firmware_has_feature(FW_FEATURE_ISERIES))
+ return -ENODEV;
+
+ iommu_vio_init();
+
+ vio_root = of_find_node_by_path("/vdevice");
+ if (!vio_root)
+ return -ENODEV;
+
+ if (viopath_hostLp == HvLpIndexInvalid) {
+ vio_set_hostlp();
+ /* If we don't have a host, bail out */
+ if (viopath_hostLp == HvLpIndexInvalid)
+ goto put_node;
+ }
+
+#if defined(CONFIG_VIOCD) || defined(CONFIG_VIOCD_MODULE)
+ get_viocd_info(vio_root);
+#endif
+
+ return 0;
+
+ put_node:
+ of_node_put(vio_root);
+ return -ENODEV;
+}
+arch_initcall(iseries_vio_init);
diff --git a/drivers/cdrom/viocd.c b/drivers/cdrom/viocd.c
index c081e54..880b5dc 100644
--- a/drivers/cdrom/viocd.c
+++ b/drivers/cdrom/viocd.c
@@ -56,30 +56,6 @@
#define VIOCD_KERN_WARNING KERN_WARNING "viocd: "
#define VIOCD_KERN_INFO KERN_INFO "viocd: "
-struct viocdlpevent {
- struct HvLpEvent event;
- u32 reserved;
- u16 version;
- u16 sub_result;
- u16 disk;
- u16 flags;
- u32 token;
- u64 offset; /* On open, max number of disks */
- u64 len; /* On open, size of the disk */
- u32 block_size; /* Only set on open */
- u32 media_size; /* Only set on open */
-};
-
-enum viocdsubtype {
- viocdopen = 0x0001,
- viocdclose = 0x0002,
- viocdread = 0x0003,
- viocdwrite = 0x0004,
- viocdlockdoor = 0x0005,
- viocdgetinfo = 0x0006,
- viocdcheck = 0x0007
-};
-
/*
* Should probably make this a module parameter....sigh
*/
@@ -131,17 +107,13 @@ static struct capability_entry capability_table[] __initdata = {
/* These are our internal structures for keeping track of devices */
static int viocd_numdev;
-struct cdrom_info {
- char rsrcname[10];
- char type[4];
- char model[3];
-};
-
struct disk_info {
struct gendisk *viocd_disk;
struct cdrom_device_info viocd_info;
struct device *dev;
- struct cdrom_info unitinfo;
+ const char *rsrcname;
+ const char *type;
+ const char *model;
};
static struct disk_info viocd_diskinfo[VIOCD_MAX_CD];
@@ -159,9 +131,9 @@ static int proc_viocd_show(struct seq_file *m, void *v)
for (i = 0; i < viocd_numdev; i++) {
seq_printf(m, "viocd device %d is iSeries resource %10.10s"
"type %4.4s, model %3.3s\n",
- i, viocd_diskinfo[i].unitinfo.rsrcname,
- viocd_diskinfo[i].unitinfo.type,
- viocd_diskinfo[i].unitinfo.model);
+ i, viocd_diskinfo[i].rsrcname,
+ viocd_diskinfo[i].type,
+ viocd_diskinfo[i].model);
}
return 0;
}
@@ -211,61 +183,6 @@ struct block_device_operations viocd_fops = {
.media_changed = viocd_blk_media_changed,
};
-/* Get info on CD devices from OS/400 */
-static void __init get_viocd_info(void)
-{
- HvLpEvent_Rc hvrc;
- int i;
- struct viocd_waitevent we;
- struct cdrom_info *viocd_unitinfo;
- dma_addr_t unitinfo_dmaaddr;
-
- viocd_unitinfo = iseries_hv_alloc(
- sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
- &unitinfo_dmaaddr, GFP_ATOMIC);
- if (viocd_unitinfo == NULL) {
- printk(VIOCD_KERN_WARNING "error allocating unitinfo\n");
- return;
- }
-
- memset(viocd_unitinfo, 0, sizeof(*viocd_unitinfo) * VIOCD_MAX_CD);
-
- init_completion(&we.com);
-
- hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp,
- HvLpEvent_Type_VirtualIo,
- viomajorsubtype_cdio | viocdgetinfo,
- HvLpEvent_AckInd_DoAck, HvLpEvent_AckType_ImmediateAck,
- viopath_sourceinst(viopath_hostLp),
- viopath_targetinst(viopath_hostLp),
- (u64)&we, VIOVERSION << 16, unitinfo_dmaaddr, 0,
- sizeof(*viocd_unitinfo) * VIOCD_MAX_CD, 0);
- if (hvrc != HvLpEvent_Rc_Good) {
- printk(VIOCD_KERN_WARNING "cdrom error sending event. rc %d\n",
- (int)hvrc);
- goto error_ret;
- }
-
- wait_for_completion(&we.com);
-
- if (we.rc) {
- const struct vio_error_entry *err =
- vio_lookup_rc(viocd_err_table, we.sub_result);
- printk(VIOCD_KERN_WARNING "bad rc %d:0x%04X on getinfo: %s\n",
- we.rc, we.sub_result, err->msg);
- goto error_ret;
- }
-
- for (i = 0; (i < VIOCD_MAX_CD) && viocd_unitinfo[i].rsrcname[0]; i++) {
- viocd_diskinfo[viocd_numdev].unitinfo = viocd_unitinfo[i];
- viocd_numdev++;
- }
-
-error_ret:
- iseries_hv_free(sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
- viocd_unitinfo, unitinfo_dmaaddr);
-}
-
static int viocd_open(struct cdrom_device_info *cdi, int purpose)
{
struct disk_info *diskinfo = cdi->handle;
@@ -576,7 +493,6 @@ static void vio_handle_cd_event(struct HvLpEvent *event)
bevent->block_size / 512);
}
/* FALLTHROUGH !! */
- case viocdgetinfo:
case viocdlockdoor:
pwe = (struct viocd_waitevent *)event->xCorrelationToken;
return_complete:
@@ -660,22 +576,30 @@ static int viocd_probe(struct vio_dev *vdev, const struct vio_device_id *id)
int deviceno;
struct disk_info *d;
struct cdrom_device_info *c;
- struct cdrom_info *ci;
struct request_queue *q;
+ struct device_node *node = vdev->dev.archdata.of_node;
deviceno = vdev->unit_address;
- if (deviceno >= viocd_numdev)
+ if (deviceno > VIOCD_MAX_CD)
return -ENODEV;
+ if (!node)
+ return -ENODEV;
+
+ if (deviceno >= viocd_numdev)
+ viocd_numdev = deviceno + 1;
d = &viocd_diskinfo[deviceno];
+ d->rsrcname = of_get_property(node, "linux,vio_rsrcname", NULL);
+ d->type = of_get_property(node, "linux,vio_type", NULL);
+ d->model = of_get_property(node, "linux,vio_model", NULL);
+
c = &d->viocd_info;
- ci = &d->unitinfo;
c->ops = &viocd_dops;
c->speed = 4;
c->capacity = 1;
c->handle = d;
- c->mask = ~find_capability(ci->type);
+ c->mask = ~find_capability(d->type);
sprintf(c->name, VIOCD_DEVICE "%c", 'a' + deviceno);
if (register_cdrom(c) != 0) {
@@ -685,7 +609,7 @@ static int viocd_probe(struct vio_dev *vdev, const struct vio_device_id *id)
}
printk(VIOCD_KERN_INFO "cd %s is iSeries resource %10.10s "
"type %4.4s, model %3.3s\n",
- c->name, ci->rsrcname, ci->type, ci->model);
+ c->name, d->rsrcname, d->type, d->model);
q = blk_init_queue(do_viocd_request, &viocd_reqlock);
if (q == NULL) {
printk(VIOCD_KERN_WARNING "Cannot allocate queue for %s!\n",
@@ -794,8 +718,6 @@ static int __init viocd_init(void)
/* Initialize our request handler */
vio_setHandler(viomajorsubtype_cdio, vio_handle_cd_event);
- get_viocd_info();
-
spin_lock_init(&viocd_reqlock);
ret = vio_register_driver(&viocd_driver);
diff --git a/include/asm-powerpc/iseries/vio.h b/include/asm-powerpc/iseries/vio.h
index 5a5cd0f..e5a405b 100644
--- a/include/asm-powerpc/iseries/vio.h
+++ b/include/asm-powerpc/iseries/vio.h
@@ -51,6 +51,30 @@
*/
#define VIO_MAX_SUBTYPES 8
+struct viocdlpevent {
+ struct HvLpEvent event;
+ u32 reserved;
+ u16 version;
+ u16 sub_result;
+ u16 disk;
+ u16 flags;
+ u32 token;
+ u64 offset; /* On open, max number of disks */
+ u64 len; /* On open, size of the disk */
+ u32 block_size; /* Only set on open */
+ u32 media_size; /* Only set on open */
+};
+
+enum viocdsubtype {
+ viocdopen = 0x0001,
+ viocdclose = 0x0002,
+ viocdread = 0x0003,
+ viocdwrite = 0x0004,
+ viocdlockdoor = 0x0005,
+ viocdgetinfo = 0x0006,
+ viocdcheck = 0x0007
+};
+
/*
* Each subtype can register a handler to process their events.
* The handler must have this interface.
--
1.5.3.2
^ permalink raw reply related
* [PATCH 07/11] [POWERPC] iSeries: missing newline in printk
From: Stephen Rothwell @ 2007-09-21 4:36 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143540.75f3becf.sfr@canb.auug.org.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/platforms/iseries/viopath.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/platforms/iseries/viopath.c b/arch/powerpc/platforms/iseries/viopath.c
index 7106a05..df23331 100644
--- a/arch/powerpc/platforms/iseries/viopath.c
+++ b/arch/powerpc/platforms/iseries/viopath.c
@@ -594,7 +594,7 @@ int viopath_close(HvLpIndex remoteLp, int subtype, int numReq)
numOpen += viopathStatus[remoteLp].users[i];
if ((viopathStatus[remoteLp].isOpen) && (numOpen == 0)) {
- printk(VIOPATH_KERN_INFO "closing connection to partition %d",
+ printk(VIOPATH_KERN_INFO "closing connection to partition %d\n",
remoteLp);
HvCallEvent_closeLpEventPath(remoteLp,
--
1.5.3.2
^ permalink raw reply related
* [PATCH 06/11] [POWERPC] Remove more iSeries specific stuff from vio.c
From: Stephen Rothwell @ 2007-09-21 4:35 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143446.ad55d494.sfr@canb.auug.org.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/kernel/vio.c | 78 ++++++++++---------------------
arch/powerpc/platforms/iseries/iommu.c | 24 +++++++++-
include/asm-powerpc/iseries/iommu.h | 4 ++
3 files changed, 52 insertions(+), 54 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index cf71c88..9e1ff21 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -48,61 +48,33 @@ static struct vio_dev vio_bus_device = { /* fake "parent" device */
.dev.bus = &vio_bus_type,
};
-#ifdef CONFIG_PPC_ISERIES
-static struct iommu_table veth_iommu_table;
-struct iommu_table vio_iommu_table;
-
-static void __init iommu_vio_init(void)
-{
- iommu_table_getparms_iSeries(255, 0, 0xff, &veth_iommu_table);
- veth_iommu_table.it_size /= 2;
- vio_iommu_table = veth_iommu_table;
- vio_iommu_table.it_offset += veth_iommu_table.it_size;
-
- if (!iommu_init_table(&veth_iommu_table, -1))
- printk("Virtual Bus VETH TCE table failed.\n");
- if (!iommu_init_table(&vio_iommu_table, -1))
- printk("Virtual Bus VIO TCE table failed.\n");
-}
-#else
-static void __init iommu_vio_init(void)
-{
-}
-#endif
-
static struct iommu_table *vio_build_iommu_table(struct vio_dev *dev)
{
-#ifdef CONFIG_PPC_ISERIES
- if (firmware_has_feature(FW_FEATURE_ISERIES)) {
- if (strcmp(dev->type, "network") == 0)
- return &veth_iommu_table;
- return &vio_iommu_table;
- } else
-#endif
- {
- const unsigned char *dma_window;
- struct iommu_table *tbl;
- unsigned long offset, size;
-
- dma_window = of_get_property(dev->dev.archdata.of_node,
- "ibm,my-dma-window", NULL);
- if (!dma_window)
- return NULL;
-
- tbl = kmalloc(sizeof(*tbl), GFP_KERNEL);
-
- of_parse_dma_window(dev->dev.archdata.of_node, dma_window,
- &tbl->it_index, &offset, &size);
-
- /* TCE table size - measured in tce entries */
- tbl->it_size = size >> IOMMU_PAGE_SHIFT;
- /* offset for VIO should always be 0 */
- tbl->it_offset = offset >> IOMMU_PAGE_SHIFT;
- tbl->it_busno = 0;
- tbl->it_type = TCE_VB;
-
- return iommu_init_table(tbl, -1);
- }
+ const unsigned char *dma_window;
+ struct iommu_table *tbl;
+ unsigned long offset, size;
+
+ if (firmware_has_feature(FW_FEATURE_ISERIES))
+ return vio_build_iommu_table_iseries(dev);
+
+ dma_window = of_get_property(dev->dev.archdata.of_node,
+ "ibm,my-dma-window", NULL);
+ if (!dma_window)
+ return NULL;
+
+ tbl = kmalloc(sizeof(*tbl), GFP_KERNEL);
+
+ of_parse_dma_window(dev->dev.archdata.of_node, dma_window,
+ &tbl->it_index, &offset, &size);
+
+ /* TCE table size - measured in tce entries */
+ tbl->it_size = size >> IOMMU_PAGE_SHIFT;
+ /* offset for VIO should always be 0 */
+ tbl->it_offset = offset >> IOMMU_PAGE_SHIFT;
+ tbl->it_busno = 0;
+ tbl->it_type = TCE_VB;
+
+ return iommu_init_table(tbl, -1);
}
/**
diff --git a/arch/powerpc/platforms/iseries/iommu.c b/arch/powerpc/platforms/iseries/iommu.c
index 3281f10..49e9c66 100644
--- a/arch/powerpc/platforms/iseries/iommu.c
+++ b/arch/powerpc/platforms/iseries/iommu.c
@@ -31,6 +31,7 @@
#include <linux/module.h>
#include <asm/iommu.h>
+#include <asm/vio.h>
#include <asm/tce.h>
#include <asm/machdep.h>
#include <asm/abs_addr.h>
@@ -191,7 +192,8 @@ void iommu_devnode_init_iSeries(struct pci_dev *pdev, struct device_node *dn)
}
#endif
-extern struct iommu_table vio_iommu_table;
+static struct iommu_table veth_iommu_table;
+static struct iommu_table vio_iommu_table;
void *iseries_hv_alloc(size_t size, dma_addr_t *dma_handle, gfp_t flag)
{
@@ -219,6 +221,26 @@ void iseries_hv_unmap(dma_addr_t dma_handle, size_t size,
iommu_unmap_single(&vio_iommu_table, dma_handle, size, direction);
}
+void __init iommu_vio_init(void)
+{
+ iommu_table_getparms_iSeries(255, 0, 0xff, &veth_iommu_table);
+ veth_iommu_table.it_size /= 2;
+ vio_iommu_table = veth_iommu_table;
+ vio_iommu_table.it_offset += veth_iommu_table.it_size;
+
+ if (!iommu_init_table(&veth_iommu_table, -1))
+ printk("Virtual Bus VETH TCE table failed.\n");
+ if (!iommu_init_table(&vio_iommu_table, -1))
+ printk("Virtual Bus VIO TCE table failed.\n");
+}
+
+struct iommu_table *vio_build_iommu_table_iseries(struct vio_dev *dev)
+{
+ if (strcmp(dev->type, "network") == 0)
+ return &veth_iommu_table;
+ return &vio_iommu_table;
+}
+
void iommu_init_early_iSeries(void)
{
ppc_md.tce_build = tce_build_iSeries;
diff --git a/include/asm-powerpc/iseries/iommu.h b/include/asm-powerpc/iseries/iommu.h
index 6e323a1..c59ee7e 100644
--- a/include/asm-powerpc/iseries/iommu.h
+++ b/include/asm-powerpc/iseries/iommu.h
@@ -22,6 +22,7 @@
*/
struct pci_dev;
+struct vio_dev;
struct device_node;
struct iommu_table;
@@ -34,4 +35,7 @@ extern void iommu_table_getparms_iSeries(unsigned long busno,
unsigned char slotno, unsigned char virtbus,
struct iommu_table *tbl);
+extern struct iommu_table *vio_build_iommu_table_iseries(struct vio_dev *dev);
+extern void iommu_vio_init(void);
+
#endif /* _ASM_POWERPC_ISERIES_IOMMU_H */
--
1.5.3.2
^ permalink raw reply related
* [PATCH 05/11] [POWERPC] remove iSeries_vio_dev
From: Stephen Rothwell @ 2007-09-21 4:34 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143337.0e4a2f0a.sfr@canb.auug.org.au>
It was only being used to carry around dma_iommu_ops and vio_iommu_table
which we can use directly instead. This also means that vio_bus_device
doesn't need to refer to them either.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/kernel/vio.c | 7 +-----
arch/powerpc/platforms/iseries/iommu.c | 30 +++++++++++++++++++++++++++
arch/powerpc/platforms/iseries/mf.c | 23 ++++++++------------
arch/powerpc/platforms/iseries/viopath.c | 6 +---
drivers/cdrom/viocd.c | 5 +--
drivers/char/viotape.c | 7 ++---
include/asm-powerpc/iseries/hv_call_event.h | 10 +++++++++
include/asm-powerpc/iseries/vio.h | 4 ---
8 files changed, 57 insertions(+), 35 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index 1d7b272..cf71c88 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -49,11 +49,8 @@ static struct vio_dev vio_bus_device = { /* fake "parent" device */
};
#ifdef CONFIG_PPC_ISERIES
-struct device *iSeries_vio_dev = &vio_bus_device.dev;
-EXPORT_SYMBOL(iSeries_vio_dev);
-
static struct iommu_table veth_iommu_table;
-static struct iommu_table vio_iommu_table;
+struct iommu_table vio_iommu_table;
static void __init iommu_vio_init(void)
{
@@ -66,8 +63,6 @@ static void __init iommu_vio_init(void)
printk("Virtual Bus VETH TCE table failed.\n");
if (!iommu_init_table(&vio_iommu_table, -1))
printk("Virtual Bus VIO TCE table failed.\n");
- vio_bus_device.dev.archdata.dma_ops = &dma_iommu_ops;
- vio_bus_device.dev.archdata.dma_data = &vio_iommu_table;
}
#else
static void __init iommu_vio_init(void)
diff --git a/arch/powerpc/platforms/iseries/iommu.c b/arch/powerpc/platforms/iseries/iommu.c
index 3b6a966..3281f10 100644
--- a/arch/powerpc/platforms/iseries/iommu.c
+++ b/arch/powerpc/platforms/iseries/iommu.c
@@ -28,6 +28,7 @@
#include <linux/dma-mapping.h>
#include <linux/list.h>
#include <linux/pci.h>
+#include <linux/module.h>
#include <asm/iommu.h>
#include <asm/tce.h>
@@ -36,6 +37,7 @@
#include <asm/prom.h>
#include <asm/pci-bridge.h>
#include <asm/iseries/hv_call_xm.h>
+#include <asm/iseries/hv_call_event.h>
#include <asm/iseries/iommu.h>
static void tce_build_iSeries(struct iommu_table *tbl, long index, long npages,
@@ -189,6 +191,34 @@ void iommu_devnode_init_iSeries(struct pci_dev *pdev, struct device_node *dn)
}
#endif
+extern struct iommu_table vio_iommu_table;
+
+void *iseries_hv_alloc(size_t size, dma_addr_t *dma_handle, gfp_t flag)
+{
+ return iommu_alloc_coherent(&vio_iommu_table, size, dma_handle,
+ DMA_32BIT_MASK, flag, -1);
+}
+EXPORT_SYMBOL_GPL(iseries_hv_alloc);
+
+void iseries_hv_free(size_t size, void *vaddr, dma_addr_t dma_handle)
+{
+ iommu_free_coherent(&vio_iommu_table, size, vaddr, dma_handle);
+}
+EXPORT_SYMBOL_GPL(iseries_hv_free);
+
+dma_addr_t iseries_hv_map(void *vaddr, size_t size,
+ enum dma_data_direction direction)
+{
+ return iommu_map_single(&vio_iommu_table, vaddr, size,
+ DMA_32BIT_MASK, direction);
+}
+
+void iseries_hv_unmap(dma_addr_t dma_handle, size_t size,
+ enum dma_data_direction direction)
+{
+ iommu_unmap_single(&vio_iommu_table, dma_handle, size, direction);
+}
+
void iommu_init_early_iSeries(void)
{
ppc_md.tce_build = tce_build_iSeries;
diff --git a/arch/powerpc/platforms/iseries/mf.c b/arch/powerpc/platforms/iseries/mf.c
index b1187d9..c0f2433 100644
--- a/arch/powerpc/platforms/iseries/mf.c
+++ b/arch/powerpc/platforms/iseries/mf.c
@@ -39,9 +39,9 @@
#include <asm/paca.h>
#include <asm/abs_addr.h>
#include <asm/firmware.h>
-#include <asm/iseries/vio.h>
#include <asm/iseries/mf.h>
#include <asm/iseries/hv_lp_config.h>
+#include <asm/iseries/hv_lp_event.h>
#include <asm/iseries/it_lp_queue.h>
#include "setup.h"
@@ -870,8 +870,7 @@ static int proc_mf_dump_cmdline(char *page, char **start, off_t off,
if ((off + count) > 256)
count = 256 - off;
- dma_addr = dma_map_single(iSeries_vio_dev, page, off + count,
- DMA_FROM_DEVICE);
+ dma_addr = iseries_hv_map(page, off + count, DMA_FROM_DEVICE);
if (dma_mapping_error(dma_addr))
return -ENOMEM;
memset(page, 0, off + count);
@@ -883,8 +882,7 @@ static int proc_mf_dump_cmdline(char *page, char **start, off_t off,
vsp_cmd.sub_data.kern.length = off + count;
mb();
rc = signal_vsp_instruction(&vsp_cmd);
- dma_unmap_single(iSeries_vio_dev, dma_addr, off + count,
- DMA_FROM_DEVICE);
+ iseries_hv_unmap(dma_addr, off + count, DMA_FROM_DEVICE);
if (rc)
return rc;
if (vsp_cmd.result_code != 0)
@@ -919,8 +917,7 @@ static int mf_getVmlinuxChunk(char *buffer, int *size, int offset, u64 side)
int len = *size;
dma_addr_t dma_addr;
- dma_addr = dma_map_single(iSeries_vio_dev, buffer, len,
- DMA_FROM_DEVICE);
+ dma_addr = iseries_hv_map(buffer, len, DMA_FROM_DEVICE);
memset(buffer, 0, len);
memset(&vsp_cmd, 0, sizeof(vsp_cmd));
vsp_cmd.cmd = 32;
@@ -938,7 +935,7 @@ static int mf_getVmlinuxChunk(char *buffer, int *size, int offset, u64 side)
rc = -ENOMEM;
}
- dma_unmap_single(iSeries_vio_dev, dma_addr, len, DMA_FROM_DEVICE);
+ iseries_hv_unmap(dma_addr, len, DMA_FROM_DEVICE);
return rc;
}
@@ -1149,8 +1146,7 @@ static int proc_mf_change_cmdline(struct file *file, const char __user *buffer,
goto out;
dma_addr = 0;
- page = dma_alloc_coherent(iSeries_vio_dev, count, &dma_addr,
- GFP_ATOMIC);
+ page = iseries_hv_alloc(count, &dma_addr, GFP_ATOMIC);
ret = -ENOMEM;
if (page == NULL)
goto out;
@@ -1170,7 +1166,7 @@ static int proc_mf_change_cmdline(struct file *file, const char __user *buffer,
ret = count;
out_free:
- dma_free_coherent(iSeries_vio_dev, count, page, dma_addr);
+ iseries_hv_free(count, page, dma_addr);
out:
return ret;
}
@@ -1190,8 +1186,7 @@ static ssize_t proc_mf_change_vmlinux(struct file *file,
goto out;
dma_addr = 0;
- page = dma_alloc_coherent(iSeries_vio_dev, count, &dma_addr,
- GFP_ATOMIC);
+ page = iseries_hv_alloc(count, &dma_addr, GFP_ATOMIC);
rc = -ENOMEM;
if (page == NULL) {
printk(KERN_ERR "mf.c: couldn't allocate memory to set vmlinux chunk\n");
@@ -1219,7 +1214,7 @@ static ssize_t proc_mf_change_vmlinux(struct file *file,
*ppos += count;
rc = count;
out_free:
- dma_free_coherent(iSeries_vio_dev, count, page, dma_addr);
+ iseries_hv_free(count, page, dma_addr);
out:
return rc;
}
diff --git a/arch/powerpc/platforms/iseries/viopath.c b/arch/powerpc/platforms/iseries/viopath.c
index 6a0060a..7106a05 100644
--- a/arch/powerpc/platforms/iseries/viopath.c
+++ b/arch/powerpc/platforms/iseries/viopath.c
@@ -124,8 +124,7 @@ static int proc_viopath_show(struct seq_file *m, void *v)
if (!buf)
return 0;
- handle = dma_map_single(iSeries_vio_dev, buf, HW_PAGE_SIZE,
- DMA_FROM_DEVICE);
+ handle = iseries_hv_map(buf, HW_PAGE_SIZE, DMA_FROM_DEVICE);
hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp,
HvLpEvent_Type_VirtualIo,
@@ -146,8 +145,7 @@ static int proc_viopath_show(struct seq_file *m, void *v)
buf[HW_PAGE_SIZE-1] = '\0';
seq_printf(m, "%s", buf);
- dma_unmap_single(iSeries_vio_dev, handle, HW_PAGE_SIZE,
- DMA_FROM_DEVICE);
+ iseries_hv_unmap(handle, HW_PAGE_SIZE, DMA_FROM_DEVICE);
kfree(buf);
seq_printf(m, "AVAILABLE_VETH=%x\n", vlanMap);
diff --git a/drivers/cdrom/viocd.c b/drivers/cdrom/viocd.c
index b88fdeb..c081e54 100644
--- a/drivers/cdrom/viocd.c
+++ b/drivers/cdrom/viocd.c
@@ -220,7 +220,7 @@ static void __init get_viocd_info(void)
struct cdrom_info *viocd_unitinfo;
dma_addr_t unitinfo_dmaaddr;
- viocd_unitinfo = dma_alloc_coherent(iSeries_vio_dev,
+ viocd_unitinfo = iseries_hv_alloc(
sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
&unitinfo_dmaaddr, GFP_ATOMIC);
if (viocd_unitinfo == NULL) {
@@ -262,8 +262,7 @@ static void __init get_viocd_info(void)
}
error_ret:
- dma_free_coherent(iSeries_vio_dev,
- sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
+ iseries_hv_free(sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
viocd_unitinfo, unitinfo_dmaaddr);
}
diff --git a/drivers/char/viotape.c b/drivers/char/viotape.c
index e12275d..064c091 100644
--- a/drivers/char/viotape.c
+++ b/drivers/char/viotape.c
@@ -392,8 +392,8 @@ static int get_viotape_info(void)
if (op == NULL)
return -ENOMEM;
- viotape_unitinfo = dma_alloc_coherent(iSeries_vio_dev, len,
- &viotape_unitinfo_token, GFP_ATOMIC);
+ viotape_unitinfo = iseries_hv_alloc(len, &viotape_unitinfo_token,
+ GFP_ATOMIC);
if (viotape_unitinfo == NULL) {
free_op_struct(op);
return -ENOMEM;
@@ -1103,8 +1103,7 @@ static void __exit viotap_exit(void)
class_destroy(tape_class);
unregister_chrdev(VIOTAPE_MAJOR, "viotape");
if (viotape_unitinfo)
- dma_free_coherent(iSeries_vio_dev,
- sizeof(viotape_unitinfo[0]) * VIOTAPE_MAX_TAPE,
+ iseries_hv_free(sizeof(viotape_unitinfo[0]) * VIOTAPE_MAX_TAPE,
viotape_unitinfo, viotape_unitinfo_token);
viopath_close(viopath_hostLp, viomajorsubtype_tape, VIOTAPE_MAXREQ + 2);
vio_clearHandler(viomajorsubtype_tape);
diff --git a/include/asm-powerpc/iseries/hv_call_event.h b/include/asm-powerpc/iseries/hv_call_event.h
index 4cec476..cc029d3 100644
--- a/include/asm-powerpc/iseries/hv_call_event.h
+++ b/include/asm-powerpc/iseries/hv_call_event.h
@@ -21,6 +21,9 @@
#ifndef _ASM_POWERPC_ISERIES_HV_CALL_EVENT_H
#define _ASM_POWERPC_ISERIES_HV_CALL_EVENT_H
+#include <linux/types.h>
+#include <linux/dma-mapping.h>
+
#include <asm/iseries/hv_call_sc.h>
#include <asm/iseries/hv_types.h>
#include <asm/abs_addr.h>
@@ -113,6 +116,13 @@ static inline HvLpEvent_Rc HvCallEvent_signalLpEventFast(HvLpIndex targetLp,
eventData3, eventData4, eventData5);
}
+extern void *iseries_hv_alloc(size_t size, dma_addr_t *dma_handle, gfp_t flag);
+extern void iseries_hv_free(size_t size, void *vaddr, dma_addr_t dma_handle);
+extern dma_addr_t iseries_hv_map(void *vaddr, size_t size,
+ enum dma_data_direction direction);
+extern void iseries_hv_unmap(dma_addr_t dma_handle, size_t size,
+ enum dma_data_direction direction);
+
static inline HvLpEvent_Rc HvCallEvent_ackLpEvent(struct HvLpEvent *event)
{
return HvCall1(HvCallEventAckLpEvent, virt_to_abs(event));
diff --git a/include/asm-powerpc/iseries/vio.h b/include/asm-powerpc/iseries/vio.h
index 7a95d29..5a5cd0f 100644
--- a/include/asm-powerpc/iseries/vio.h
+++ b/include/asm-powerpc/iseries/vio.h
@@ -150,8 +150,4 @@ enum viochar_rc {
viochar_rc_ebusy = 1
};
-struct device;
-
-extern struct device *iSeries_vio_dev;
-
#endif /* _ASM_POWERPC_ISERIES_VIO_H */
--
1.5.3.2
^ permalink raw reply related
* [PATCH 04/11] [POWERPC] iSeries: simplify viocd initialisation
From: Stephen Rothwell @ 2007-09-21 4:33 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143205.64271521.sfr@canb.auug.org.au>
We don't need to keep a lump of dma coherent memory around for the life
of the module.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
drivers/cdrom/viocd.c | 37 +++++++++++++------------------------
1 files changed, 13 insertions(+), 24 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/drivers/cdrom/viocd.c b/drivers/cdrom/viocd.c
index e51550d..b88fdeb 100644
--- a/drivers/cdrom/viocd.c
+++ b/drivers/cdrom/viocd.c
@@ -136,17 +136,12 @@ struct cdrom_info {
char type[4];
char model[3];
};
-/*
- * This needs to be allocated since it is passed to the
- * Hypervisor and we may be a module.
- */
-static struct cdrom_info *viocd_unitinfo;
-static dma_addr_t unitinfo_dmaaddr;
struct disk_info {
struct gendisk *viocd_disk;
struct cdrom_device_info viocd_info;
struct device *dev;
+ struct cdrom_info unitinfo;
};
static struct disk_info viocd_diskinfo[VIOCD_MAX_CD];
@@ -164,9 +159,9 @@ static int proc_viocd_show(struct seq_file *m, void *v)
for (i = 0; i < viocd_numdev; i++) {
seq_printf(m, "viocd device %d is iSeries resource %10.10s"
"type %4.4s, model %3.3s\n",
- i, viocd_unitinfo[i].rsrcname,
- viocd_unitinfo[i].type,
- viocd_unitinfo[i].model);
+ i, viocd_diskinfo[i].unitinfo.rsrcname,
+ viocd_diskinfo[i].unitinfo.type,
+ viocd_diskinfo[i].unitinfo.model);
}
return 0;
}
@@ -222,6 +217,8 @@ static void __init get_viocd_info(void)
HvLpEvent_Rc hvrc;
int i;
struct viocd_waitevent we;
+ struct cdrom_info *viocd_unitinfo;
+ dma_addr_t unitinfo_dmaaddr;
viocd_unitinfo = dma_alloc_coherent(iSeries_vio_dev,
sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
@@ -259,16 +256,15 @@ static void __init get_viocd_info(void)
goto error_ret;
}
- for (i = 0; (i < VIOCD_MAX_CD) && viocd_unitinfo[i].rsrcname[0]; i++)
+ for (i = 0; (i < VIOCD_MAX_CD) && viocd_unitinfo[i].rsrcname[0]; i++) {
+ viocd_diskinfo[viocd_numdev].unitinfo = viocd_unitinfo[i];
viocd_numdev++;
+ }
error_ret:
- if (viocd_numdev == 0) {
- dma_free_coherent(iSeries_vio_dev,
- sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
- viocd_unitinfo, unitinfo_dmaaddr);
- viocd_unitinfo = NULL;
- }
+ dma_free_coherent(iSeries_vio_dev,
+ sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
+ viocd_unitinfo, unitinfo_dmaaddr);
}
static int viocd_open(struct cdrom_device_info *cdi, int purpose)
@@ -674,7 +670,7 @@ static int viocd_probe(struct vio_dev *vdev, const struct vio_device_id *id)
d = &viocd_diskinfo[deviceno];
c = &d->viocd_info;
- ci = &viocd_unitinfo[deviceno];
+ ci = &d->unitinfo;
c->ops = &viocd_dops;
c->speed = 4;
@@ -816,9 +812,6 @@ static int __init viocd_init(void)
return 0;
out_free_info:
- dma_free_coherent(iSeries_vio_dev,
- sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
- viocd_unitinfo, unitinfo_dmaaddr);
vio_clearHandler(viomajorsubtype_cdio);
viopath_close(viopath_hostLp, viomajorsubtype_cdio, MAX_CD_REQ + 2);
out_unregister:
@@ -830,10 +823,6 @@ static void __exit viocd_exit(void)
{
remove_proc_entry("iSeries/viocd", NULL);
vio_unregister_driver(&viocd_driver);
- if (viocd_unitinfo != NULL)
- dma_free_coherent(iSeries_vio_dev,
- sizeof(*viocd_unitinfo) * VIOCD_MAX_CD,
- viocd_unitinfo, unitinfo_dmaaddr);
viopath_close(viopath_hostLp, viomajorsubtype_cdio, MAX_CD_REQ + 2);
vio_clearHandler(viomajorsubtype_cdio);
unregister_blkdev(VIOCD_MAJOR, VIOCD_DEVICE);
--
1.5.3.2
^ permalink raw reply related
* [PATCH 03/11] [POWERPC] Make vio_bus_type static
From: Stephen Rothwell @ 2007-09-21 4:32 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921143102.eb30d770.sfr@canb.auug.org.au>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/kernel/vio.c | 4 +++-
include/asm-powerpc/vio.h | 1 -
2 files changed, 3 insertions(+), 2 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index ee15c22..1d7b272 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -39,6 +39,8 @@
extern struct kset devices_subsys; /* needed for vio_find_name() */
+static struct bus_type vio_bus_type;
+
static struct vio_dev vio_bus_device = { /* fake "parent" device */
.name = vio_bus_device.dev.bus_id,
.type = "",
@@ -388,7 +390,7 @@ static int vio_hotplug(struct device *dev, char **envp, int num_envp,
return 0;
}
-struct bus_type vio_bus_type = {
+static struct bus_type vio_bus_type = {
.name = "vio",
.dev_attrs = vio_dev_attrs,
.uevent = vio_hotplug,
diff --git a/include/asm-powerpc/vio.h b/include/asm-powerpc/vio.h
index 3a0975e..598d111 100644
--- a/include/asm-powerpc/vio.h
+++ b/include/asm-powerpc/vio.h
@@ -63,7 +63,6 @@ struct vio_driver {
};
extern struct dma_mapping_ops vio_dma_ops;
-extern struct bus_type vio_bus_type;
extern int vio_register_driver(struct vio_driver *drv);
extern void vio_unregister_driver(struct vio_driver *drv);
--
1.5.3.2
^ permalink raw reply related
* [PATCH 02/11] [POWERPC] Simplify vio_bus_init a little for legacy iSeries
From: Stephen Rothwell @ 2007-09-21 4:31 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921142928.64451ad8.sfr@canb.auug.org.au>
iSeries_vio_dev was already statically initialised and we can remove
one set of #ifdef CONFIG_PPC_ISERIES guards.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/kernel/vio.c | 14 +++++++-------
1 files changed, 7 insertions(+), 7 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index 54645ba..ee15c22 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -64,6 +64,12 @@ static void __init iommu_vio_init(void)
printk("Virtual Bus VETH TCE table failed.\n");
if (!iommu_init_table(&vio_iommu_table, -1))
printk("Virtual Bus VIO TCE table failed.\n");
+ vio_bus_device.dev.archdata.dma_ops = &dma_iommu_ops;
+ vio_bus_device.dev.archdata.dma_data = &vio_iommu_table;
+}
+#else
+static void __init iommu_vio_init(void)
+{
}
#endif
@@ -282,14 +288,8 @@ static int __init vio_bus_init(void)
int err;
struct device_node *node_vroot;
-#ifdef CONFIG_PPC_ISERIES
- if (firmware_has_feature(FW_FEATURE_ISERIES)) {
+ if (firmware_has_feature(FW_FEATURE_ISERIES))
iommu_vio_init();
- vio_bus_device.dev.archdata.dma_ops = &dma_iommu_ops;
- vio_bus_device.dev.archdata.dma_data = &vio_iommu_table;
- iSeries_vio_dev = &vio_bus_device.dev;
- }
-#endif /* CONFIG_PPC_ISERIES */
err = bus_register(&vio_bus_type);
if (err) {
--
1.5.3.2
^ permalink raw reply related
* [PATCH 01/11] [POWERPC] Remove debug printk from vio_bus_init
From: Stephen Rothwell @ 2007-09-21 4:29 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
In-Reply-To: <20070921135533.461eecdf.sfr@canb.auug.org.au>
As it just adds noise to the boot messages.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/powerpc/kernel/vio.c | 5 +----
1 files changed, 1 insertions(+), 4 deletions(-)
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
index 62c1bc1..54645ba 100644
--- a/arch/powerpc/kernel/vio.c
+++ b/arch/powerpc/kernel/vio.c
@@ -317,11 +317,8 @@ static int __init vio_bus_init(void)
* the device tree. Drivers will associate with them later.
*/
for (of_node = node_vroot->child; of_node != NULL;
- of_node = of_node->sibling) {
- printk(KERN_DEBUG "%s: processing %p\n",
- __FUNCTION__, of_node);
+ of_node = of_node->sibling)
vio_register_device_node(of_node);
- }
of_node_put(node_vroot);
}
--
1.5.3.2
^ permalink raw reply related
* [PATCH 0/11][POWERPC] vio cleanups
From: Stephen Rothwell @ 2007-09-21 3:55 UTC (permalink / raw)
To: paulus; +Cc: ppc-dev
Hi all,
This series of patches does some cleanups to PowerPC vio especially the
legacy iSeries parts of that. The significant change is to do the
probing of the virtual cdroms, tapes and disks earlier and create device
tree entries for them dynamically so that we only have entries for those
that exist (rather than the theoretical maximum number).
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
^ permalink raw reply
* Re: [PATCH v2 2/4] Implement generic time of day clocksource for powerpc machines.
From: Daniel Walker @ 2007-09-21 4:05 UTC (permalink / raw)
To: Tony Breeds
Cc: linuxppc-dev, Thomas Gleixner, Paul Mackerras, Realtime Kernel
In-Reply-To: <20070921032603.07AF232C889@thor>
On Fri, 2007-09-21 at 13:26 +1000, Tony Breeds wrote:
> +
> + if (__USE_RTC())
> + clock = &clocksource_rtc;
> + else
> + clock = &clocksource_timebase;
> +
> + clock->mult = clocksource_hz2mult(tb_ticks_per_sec,
> clock->shift);
I don't think the RTC frequency isn't the same as the timebase? Seems
like the RTC only case about seconds at the lowest level. If that's the
case then the jiffies clock might be better to use .. The other thing I
wonder is if the __USE_RTC boards might have lower level clocks that
could be used instead ..
Daniel
^ permalink raw reply
* 44x bug: funny TLB writes?
From: Hollis Blanchard @ 2007-09-21 3:34 UTC (permalink / raw)
To: linuxppc-dev
I seem to have come across a strange bug while doing KVM development. It
seems that the final tlbwe in finish_tlb (head_44x.S) is actually
leaking RPN bits into the "attribute" word.
When I set a breakpoint there and press enter on the serial console, I
see r12=ef600703, which is the physical address of the UART on this chip
(440EP), plus the correct permission bits at the bottom.
Am I crazy? I'm not really looking to step through that assembly right
now... Clearly (current) hardware is just ignoring these errant writes,
but it should be fixed.
--
Hollis Blanchard
IBM Linux Technology Center
^ permalink raw reply
* [PATCH v2 3/4] Implement clockevents driver for powerpc
From: Tony Breeds @ 2007-09-21 3:26 UTC (permalink / raw)
To: linuxppc-dev, Paul Mackerras; +Cc: Thomas Gleixner, Realtime Kernel
In-Reply-To: <1190345162.620000.305760830507.qpush@thor>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
---
Updated to remove a trivial FIXME left behind
arch/powerpc/Kconfig | 3
arch/powerpc/kernel/smp.c | 3
arch/powerpc/kernel/time.c | 134 +++++++++++++++++++++++++------------
include/asm-powerpc/time.h | 1
4 files changed, 98 insertions(+), 43 deletions(-)
Index: working/arch/powerpc/Kconfig
===================================================================
--- working.orig/arch/powerpc/Kconfig
+++ working/arch/powerpc/Kconfig
@@ -30,6 +30,9 @@ config GENERIC_TIME
config GENERIC_TIME_VSYSCALL
def_bool y
+config GENERIC_CLOCKEVENTS
+ def_bool y
+
config GENERIC_HARDIRQS
bool
default y
Index: working/arch/powerpc/kernel/smp.c
===================================================================
--- working.orig/arch/powerpc/kernel/smp.c
+++ working/arch/powerpc/kernel/smp.c
@@ -560,6 +560,8 @@ int __devinit start_secondary(void *unus
if (system_state > SYSTEM_BOOTING)
snapshot_timebase();
+ secondary_cpu_time_init();
+
spin_lock(&call_lock);
cpu_set(cpu, cpu_online_map);
spin_unlock(&call_lock);
Index: working/arch/powerpc/kernel/time.c
===================================================================
--- working.orig/arch/powerpc/kernel/time.c
+++ working/arch/powerpc/kernel/time.c
@@ -75,6 +75,7 @@
/* powerpc clocksource/clockevent code */
+#include <linux/clockchips.h>
#include <linux/clocksource.h>
static cycle_t rtc_read(void);
@@ -99,6 +100,27 @@ static struct clocksource clocksource_ti
.read = timebase_read,
};
+#define DECREMENTER_MAX 0x7fffffff
+
+static int decrementer_set_next_event(unsigned long evt,
+ struct clock_event_device *dev);
+static void decrementer_set_mode(enum clock_event_mode mode,
+ struct clock_event_device *dev);
+
+static struct clock_event_device decrementer_clockevent = {
+ .name = "decrementer",
+ .rating = 200,
+ .shift = 32,
+ .mult = 0, /* To be filled in */
+ .irq = -1,
+ .set_next_event = decrementer_set_next_event,
+ .set_mode = decrementer_set_mode,
+ .features = CLOCK_EVT_FEAT_ONESHOT,
+};
+
+static DEFINE_PER_CPU(struct clock_event_device, decrementers);
+void init_decrementer_clockevent(void);
+
#ifdef CONFIG_PPC_ISERIES
static unsigned long __initdata iSeries_recal_titan;
static signed long __initdata iSeries_recal_tb;
@@ -519,10 +541,12 @@ void __init iSeries_time_init_early(void
void timer_interrupt(struct pt_regs * regs)
{
struct pt_regs *old_regs;
- int next_dec;
int cpu = smp_processor_id();
- unsigned long ticks;
- u64 tb_next_jiffy;
+ struct clock_event_device *evt = &per_cpu(decrementers, cpu);
+
+ /* Ensure a positive value is written to the decrementer, or else
+ * some CPUs will continuue to take decrementer exceptions */
+ set_dec(DECREMENTER_MAX);
#ifdef CONFIG_PPC32
if (atomic_read(&ppc_n_lost_interrupts) != 0)
@@ -532,7 +556,6 @@ void timer_interrupt(struct pt_regs * re
old_regs = set_irq_regs(regs);
irq_enter();
- profile_tick(CPU_PROFILING);
calculate_steal_time();
#ifdef CONFIG_PPC_ISERIES
@@ -540,44 +563,20 @@ void timer_interrupt(struct pt_regs * re
get_lppaca()->int_dword.fields.decr_int = 0;
#endif
- while ((ticks = tb_ticks_since(per_cpu(last_jiffy, cpu)))
- >= tb_ticks_per_jiffy) {
- /* Update last_jiffy */
- per_cpu(last_jiffy, cpu) += tb_ticks_per_jiffy;
- /* Handle RTCL overflow on 601 */
- if (__USE_RTC() && per_cpu(last_jiffy, cpu) >= 1000000000)
- per_cpu(last_jiffy, cpu) -= 1000000000;
-
- /*
- * We cannot disable the decrementer, so in the period
- * between this cpu's being marked offline in cpu_online_map
- * and calling stop-self, it is taking timer interrupts.
- * Avoid calling into the scheduler rebalancing code if this
- * is the case.
- */
- if (!cpu_is_offline(cpu))
- account_process_time(regs);
-
- /*
- * No need to check whether cpu is offline here; boot_cpuid
- * should have been fixed up by now.
- */
- if (cpu != boot_cpuid)
- continue;
+ /*
+ * We cannot disable the decrementer, so in the period
+ * between this cpu's being marked offline in cpu_online_map
+ * and calling stop-self, it is taking timer interrupts.
+ * Avoid calling into the scheduler rebalancing code if this
+ * is the case.
+ */
+ if (!cpu_is_offline(cpu))
+ account_process_time(regs);
- write_seqlock(&xtime_lock);
- tb_next_jiffy = tb_last_jiffy + tb_ticks_per_jiffy;
- if (__USE_RTC() && tb_next_jiffy >= 1000000000)
- tb_next_jiffy -= 1000000000;
- if (per_cpu(last_jiffy, cpu) >= tb_next_jiffy) {
- tb_last_jiffy = tb_next_jiffy;
- do_timer(1);
- }
- write_sequnlock(&xtime_lock);
- }
-
- next_dec = tb_ticks_per_jiffy - ticks;
- set_dec(next_dec);
+ if (evt->event_handler)
+ evt->event_handler(evt);
+ else
+ evt->set_next_event(DECREMENTER_MAX, evt);
#ifdef CONFIG_PPC_ISERIES
if (firmware_has_feature(FW_FEATURE_ISERIES) && hvlpevent_is_pending())
@@ -797,6 +796,53 @@ void __init clocksource_init(void)
clock->name, clock->mult, clock->shift);
}
+static int decrementer_set_next_event(unsigned long evt,
+ struct clock_event_device *dev)
+{
+ set_dec(evt);
+ return 0;
+}
+
+static void decrementer_set_mode(enum clock_event_mode mode,
+ struct clock_event_device *dev)
+{
+ if (mode != CLOCK_EVT_MODE_ONESHOT)
+ decrementer_set_next_event(DECREMENTER_MAX, dev);
+}
+
+static void register_decrementer_clockevent(int cpu)
+{
+ struct clock_event_device *dec = &per_cpu(decrementers, cpu);
+
+ *dec = decrementer_clockevent;
+ dec->cpumask = cpumask_of_cpu(cpu);
+
+ printk(KERN_ERR "clockevent: %s mult[%lx] shift[%d] cpu[%d]\n",
+ dec->name, dec->mult, dec->shift, cpu);
+
+ clockevents_register_device(dec);
+}
+
+void init_decrementer_clockevent(void)
+{
+ int cpu = smp_processor_id();
+
+ decrementer_clockevent.mult = div_sc(ppc_tb_freq, NSEC_PER_SEC,
+ decrementer_clockevent.shift);
+ decrementer_clockevent.max_delta_ns =
+ clockevent_delta2ns(DECREMENTER_MAX, &decrementer_clockevent);
+ decrementer_clockevent.min_delta_ns = 1000;
+
+ register_decrementer_clockevent(cpu);
+}
+
+void secondary_cpu_time_init(void)
+{
+ /* FIME: Should make unrelatred change to move snapshot_timebase
+ * call here ! */
+ register_decrementer_clockevent(smp_processor_id());
+}
+
/* This function is only called on the boot processor */
void __init time_init(void)
{
@@ -912,8 +958,7 @@ void __init time_init(void)
#endif
clocksource_init();
- /* Not exact, but the timer interrupt takes care of this */
- set_dec(tb_ticks_per_jiffy);
+ init_decrementer_clockevent();
}
Index: working/include/asm-powerpc/time.h
===================================================================
--- working.orig/include/asm-powerpc/time.h
+++ working/include/asm-powerpc/time.h
@@ -245,6 +245,7 @@ extern void snapshot_timebases(void);
#define snapshot_timebases() do { } while (0)
#endif
+extern void secondary_cpu_time_init(void);
extern void iSeries_time_init_early(void);
#endif /* __KERNEL__ */
^ permalink raw reply
* [PATCH v2 4/4] Enable tickless idle and high res timers for powerpc
From: Tony Breeds @ 2007-09-21 3:26 UTC (permalink / raw)
To: linuxppc-dev, Paul Mackerras; +Cc: Thomas Gleixner, Realtime Kernel
In-Reply-To: <1190345162.620000.305760830507.qpush@thor>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
---
Updated to remove the need for patching generic code to include hrtimer.h
arch/powerpc/Kconfig | 1 +
arch/powerpc/kernel/idle.c | 3 +++
arch/powerpc/platforms/iseries/setup.c | 5 +++++
3 files changed, 9 insertions(+)
Index: working/arch/powerpc/Kconfig
===================================================================
--- working.orig/arch/powerpc/Kconfig
+++ working/arch/powerpc/Kconfig
@@ -168,6 +168,7 @@ config HIGHMEM
bool "High memory support"
depends on PPC32
+source kernel/time/Kconfig
source kernel/Kconfig.hz
source kernel/Kconfig.preempt
source "fs/Kconfig.binfmt"
Index: working/arch/powerpc/kernel/idle.c
===================================================================
--- working.orig/arch/powerpc/kernel/idle.c
+++ working/arch/powerpc/kernel/idle.c
@@ -24,6 +24,7 @@
#include <linux/smp.h>
#include <linux/cpu.h>
#include <linux/sysctl.h>
+#include <linux/tick.h>
#include <asm/system.h>
#include <asm/processor.h>
@@ -59,6 +60,7 @@ void cpu_idle(void)
set_thread_flag(TIF_POLLING_NRFLAG);
while (1) {
+ tick_nohz_stop_sched_tick();
while (!need_resched() && !cpu_should_die()) {
ppc64_runlatch_off();
@@ -90,6 +92,7 @@ void cpu_idle(void)
HMT_medium();
ppc64_runlatch_on();
+ tick_nohz_restart_sched_tick();
if (cpu_should_die())
cpu_die();
preempt_enable_no_resched();
Index: working/arch/powerpc/platforms/iseries/setup.c
===================================================================
--- working.orig/arch/powerpc/platforms/iseries/setup.c
+++ working/arch/powerpc/platforms/iseries/setup.c
@@ -26,6 +26,8 @@
#include <linux/major.h>
#include <linux/root_dev.h>
#include <linux/kernel.h>
+#include <linux/hrtimer.h>
+#include <linux/tick.h>
#include <asm/processor.h>
#include <asm/machdep.h>
@@ -561,6 +563,7 @@ static void yield_shared_processor(void)
static void iseries_shared_idle(void)
{
while (1) {
+ tick_nohz_stop_sched_tick();
while (!need_resched() && !hvlpevent_is_pending()) {
local_irq_disable();
ppc64_runlatch_off();
@@ -574,6 +577,7 @@ static void iseries_shared_idle(void)
}
ppc64_runlatch_on();
+ tick_nohz_restart_sched_tick();
if (hvlpevent_is_pending())
process_iSeries_events();
@@ -589,6 +593,7 @@ static void iseries_dedicated_idle(void)
set_thread_flag(TIF_POLLING_NRFLAG);
while (1) {
+ tick_nohz_stop_sched_tick();
if (!need_resched()) {
while (!need_resched()) {
ppc64_runlatch_off();
@@ -605,6 +610,7 @@ static void iseries_dedicated_idle(void)
}
ppc64_runlatch_on();
+ tick_nohz_restart_sched_tick();
preempt_enable_no_resched();
schedule();
preempt_disable();
^ permalink raw reply
* [PATCH v2 2/4] Implement generic time of day clocksource for powerpc machines.
From: Tony Breeds @ 2007-09-21 3:26 UTC (permalink / raw)
To: linuxppc-dev, Paul Mackerras; +Cc: Thomas Gleixner, Realtime Kernel
In-Reply-To: <1190345162.620000.305760830507.qpush@thor>
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
---
Updated to create a clocksource driver specific to the 601
arch/powerpc/Kconfig | 6
arch/powerpc/kernel/time.c | 255 ++++++++++++-------------------------
2 files changed, 91 insertions(+), 170 deletions(-)
Index: working/arch/powerpc/Kconfig
===================================================================
--- working.orig/arch/powerpc/Kconfig
+++ working/arch/powerpc/Kconfig
@@ -24,6 +24,12 @@ config MMU
config GENERIC_CMOS_UPDATE
def_bool y
+config GENERIC_TIME
+ def_bool y
+
+config GENERIC_TIME_VSYSCALL
+ def_bool y
+
config GENERIC_HARDIRQS
bool
default y
Index: working/arch/powerpc/kernel/time.c
===================================================================
--- working.orig/arch/powerpc/kernel/time.c
+++ working/arch/powerpc/kernel/time.c
@@ -73,9 +73,38 @@
#include <asm/iseries/hv_call_xm.h>
#endif
+/* powerpc clocksource/clockevent code */
+
+#include <linux/clocksource.h>
+
+static cycle_t rtc_read(void);
+static struct clocksource clocksource_rtc = {
+ .name = "rtc",
+ .rating = 400,
+ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
+ .mask = CLOCKSOURCE_MASK(64),
+ .shift = 22,
+ .mult = 0, /* To be filled in */
+ .read = rtc_read,
+};
+
+static cycle_t timebase_read(void);
+static struct clocksource clocksource_timebase = {
+ .name = "timebase",
+ .rating = 400,
+ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
+ .mask = CLOCKSOURCE_MASK(64),
+ .shift = 22,
+ .mult = 0, /* To be filled in */
+ .read = timebase_read,
+};
+
#ifdef CONFIG_PPC_ISERIES
static unsigned long __initdata iSeries_recal_titan;
static signed long __initdata iSeries_recal_tb;
+
+/* Forward declaration is only needed for iSereis compiles */
+void __init clocksource_init(void);
#endif
#define XSEC_PER_SEC (1024*1024)
@@ -343,65 +372,6 @@ void udelay(unsigned long usecs)
}
EXPORT_SYMBOL(udelay);
-/*
- * This version of gettimeofday has microsecond resolution.
- */
-static inline void __do_gettimeofday(struct timeval *tv)
-{
- unsigned long sec, usec;
- u64 tb_ticks, xsec;
- struct gettimeofday_vars *temp_varp;
- u64 temp_tb_to_xs, temp_stamp_xsec;
-
- /*
- * These calculations are faster (gets rid of divides)
- * if done in units of 1/2^20 rather than microseconds.
- * The conversion to microseconds at the end is done
- * without a divide (and in fact, without a multiply)
- */
- temp_varp = do_gtod.varp;
-
- /* Sampling the time base must be done after loading
- * do_gtod.varp in order to avoid racing with update_gtod.
- */
- data_barrier(temp_varp);
- tb_ticks = get_tb() - temp_varp->tb_orig_stamp;
- temp_tb_to_xs = temp_varp->tb_to_xs;
- temp_stamp_xsec = temp_varp->stamp_xsec;
- xsec = temp_stamp_xsec + mulhdu(tb_ticks, temp_tb_to_xs);
- sec = xsec / XSEC_PER_SEC;
- usec = (unsigned long)xsec & (XSEC_PER_SEC - 1);
- usec = SCALE_XSEC(usec, 1000000);
-
- tv->tv_sec = sec;
- tv->tv_usec = usec;
-}
-
-void do_gettimeofday(struct timeval *tv)
-{
- if (__USE_RTC()) {
- /* do this the old way */
- unsigned long flags, seq;
- unsigned int sec, nsec, usec;
-
- do {
- seq = read_seqbegin_irqsave(&xtime_lock, flags);
- sec = xtime.tv_sec;
- nsec = xtime.tv_nsec + tb_ticks_since(tb_last_jiffy);
- } while (read_seqretry_irqrestore(&xtime_lock, seq, flags));
- usec = nsec / 1000;
- while (usec >= 1000000) {
- usec -= 1000000;
- ++sec;
- }
- tv->tv_sec = sec;
- tv->tv_usec = usec;
- return;
- }
- __do_gettimeofday(tv);
-}
-
-EXPORT_SYMBOL(do_gettimeofday);
/*
* There are two copies of tb_to_xs and stamp_xsec so that no
@@ -447,56 +417,6 @@ static inline void update_gtod(u64 new_t
++(vdso_data->tb_update_count);
}
-/*
- * When the timebase - tb_orig_stamp gets too big, we do a manipulation
- * between tb_orig_stamp and stamp_xsec. The goal here is to keep the
- * difference tb - tb_orig_stamp small enough to always fit inside a
- * 32 bits number. This is a requirement of our fast 32 bits userland
- * implementation in the vdso. If we "miss" a call to this function
- * (interrupt latency, CPU locked in a spinlock, ...) and we end up
- * with a too big difference, then the vdso will fallback to calling
- * the syscall
- */
-static __inline__ void timer_recalc_offset(u64 cur_tb)
-{
- unsigned long offset;
- u64 new_stamp_xsec;
- u64 tlen, t2x;
- u64 tb, xsec_old, xsec_new;
- struct gettimeofday_vars *varp;
-
- if (__USE_RTC())
- return;
- tlen = current_tick_length();
- offset = cur_tb - do_gtod.varp->tb_orig_stamp;
- if (tlen == last_tick_len && offset < 0x80000000u)
- return;
- if (tlen != last_tick_len) {
- t2x = mulhdu(tlen << TICKLEN_SHIFT, ticklen_to_xs);
- last_tick_len = tlen;
- } else
- t2x = do_gtod.varp->tb_to_xs;
- new_stamp_xsec = (u64) xtime.tv_nsec * XSEC_PER_SEC;
- do_div(new_stamp_xsec, 1000000000);
- new_stamp_xsec += (u64) xtime.tv_sec * XSEC_PER_SEC;
-
- ++vdso_data->tb_update_count;
- smp_mb();
-
- /*
- * Make sure time doesn't go backwards for userspace gettimeofday.
- */
- tb = get_tb();
- varp = do_gtod.varp;
- xsec_old = mulhdu(tb - varp->tb_orig_stamp, varp->tb_to_xs)
- + varp->stamp_xsec;
- xsec_new = mulhdu(tb - cur_tb, t2x) + new_stamp_xsec;
- if (xsec_new < xsec_old)
- new_stamp_xsec += xsec_old - xsec_new;
-
- update_gtod(cur_tb, new_stamp_xsec, t2x);
-}
-
#ifdef CONFIG_SMP
unsigned long profile_pc(struct pt_regs *regs)
{
@@ -568,6 +488,8 @@ static int __init iSeries_tb_recal(void)
iSeries_recal_titan = titan;
iSeries_recal_tb = tb;
+ /* Called here as now we know accurate values for the timebase */
+ clocksource_init();
return 0;
}
late_initcall(iSeries_tb_recal);
@@ -650,7 +572,6 @@ void timer_interrupt(struct pt_regs * re
if (per_cpu(last_jiffy, cpu) >= tb_next_jiffy) {
tb_last_jiffy = tb_next_jiffy;
do_timer(1);
- timer_recalc_offset(tb_last_jiffy);
}
write_sequnlock(&xtime_lock);
}
@@ -722,66 +643,6 @@ unsigned long long sched_clock(void)
return mulhdu(get_tb() - boot_tb, tb_to_ns_scale) << tb_to_ns_shift;
}
-int do_settimeofday(struct timespec *tv)
-{
- time_t wtm_sec, new_sec = tv->tv_sec;
- long wtm_nsec, new_nsec = tv->tv_nsec;
- unsigned long flags;
- u64 new_xsec;
- unsigned long tb_delta;
-
- if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
- return -EINVAL;
-
- write_seqlock_irqsave(&xtime_lock, flags);
-
- /*
- * Updating the RTC is not the job of this code. If the time is
- * stepped under NTP, the RTC will be updated after STA_UNSYNC
- * is cleared. Tools like clock/hwclock either copy the RTC
- * to the system time, in which case there is no point in writing
- * to the RTC again, or write to the RTC but then they don't call
- * settimeofday to perform this operation.
- */
-
- /* Make userspace gettimeofday spin until we're done. */
- ++vdso_data->tb_update_count;
- smp_mb();
-
- /*
- * Subtract off the number of nanoseconds since the
- * beginning of the last tick.
- */
- tb_delta = tb_ticks_since(tb_last_jiffy);
- tb_delta = mulhdu(tb_delta, do_gtod.varp->tb_to_xs); /* in xsec */
- new_nsec -= SCALE_XSEC(tb_delta, 1000000000);
-
- wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - new_sec);
- wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - new_nsec);
-
- set_normalized_timespec(&xtime, new_sec, new_nsec);
- set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
-
- ntp_clear();
-
- new_xsec = xtime.tv_nsec;
- if (new_xsec != 0) {
- new_xsec *= XSEC_PER_SEC;
- do_div(new_xsec, NSEC_PER_SEC);
- }
- new_xsec += (u64)xtime.tv_sec * XSEC_PER_SEC;
- update_gtod(tb_last_jiffy, new_xsec, do_gtod.varp->tb_to_xs);
-
- vdso_data->tz_minuteswest = sys_tz.tz_minuteswest;
- vdso_data->tz_dsttime = sys_tz.tz_dsttime;
-
- write_sequnlock_irqrestore(&xtime_lock, flags);
- clock_was_set();
- return 0;
-}
-
-EXPORT_SYMBOL(do_settimeofday);
-
static int __init get_freq(char *name, int cells, unsigned long *val)
{
struct device_node *cpu;
@@ -873,6 +734,69 @@ unsigned long read_persistent_clock(void
tm.tm_hour, tm.tm_min, tm.tm_sec);
}
+/* clocksource code */
+static cycle_t rtc_read(void)
+{
+ return (cycle_t)get_rtc();
+}
+
+static cycle_t timebase_read(void)
+{
+ return (cycle_t)get_tb();
+}
+
+void update_vsyscall(struct timespec *wall_time, struct clocksource *clock)
+{
+ u64 t2x, stamp_xsec;
+
+ if (__USE_RTC() || clock != &clocksource_timebase)
+ return;
+
+ /* Make userspace gettimeofday spin until we're done. */
+ ++vdso_data->tb_update_count;
+ smp_mb();
+
+ /* XXX this assumes clock->shift == 22 */
+ /* 4611686018 ~= 2^(20+64-22) / 1e9 */
+ t2x = (u64) clock->mult * 4611686018ULL;
+ stamp_xsec = (u64) xtime.tv_nsec * XSEC_PER_SEC;
+ do_div(stamp_xsec, 1000000000);
+ stamp_xsec += (u64) xtime.tv_sec * XSEC_PER_SEC;
+ update_gtod(clock->cycle_last, stamp_xsec, t2x);
+}
+
+void update_vsyscall_tz(void)
+{
+ /* Make userspace gettimeofday spin until we're done. */
+ ++vdso_data->tb_update_count;
+ smp_mb();
+ vdso_data->tz_minuteswest = sys_tz.tz_minuteswest;
+ vdso_data->tz_dsttime = sys_tz.tz_dsttime;
+ smp_mb();
+ ++vdso_data->tb_update_count;
+}
+
+void __init clocksource_init(void)
+{
+ struct clocksource *clock;
+
+ if (__USE_RTC())
+ clock = &clocksource_rtc;
+ else
+ clock = &clocksource_timebase;
+
+ clock->mult = clocksource_hz2mult(tb_ticks_per_sec, clock->shift);
+
+ if (clocksource_register(clock)) {
+ printk(KERN_ERR "clocksource: %s is already registered\n",
+ clock->name);
+ return;
+ }
+
+ printk(KERN_INFO "clocksource: %s mult[%x] shift[%d] registered\n",
+ clock->name, clock->mult, clock->shift);
+}
+
/* This function is only called on the boot processor */
void __init time_init(void)
{
@@ -982,6 +906,12 @@ void __init time_init(void)
write_sequnlock_irqrestore(&xtime_lock, flags);
+ /* Register the clocksource, if we're not running on iSeries */
+#ifdef CONFIG_ISERIES
+ if (!firmware_has_feature(FW_FEATURE_ISERIES))
+#endif
+ clocksource_init();
+
/* Not exact, but the timer interrupt takes care of this */
set_dec(tb_ticks_per_jiffy);
}
^ permalink raw reply
* [PATCH v2 1/4] Implement {read,update}_persistent_clock.
From: Tony Breeds @ 2007-09-21 3:26 UTC (permalink / raw)
To: linuxppc-dev, Paul Mackerras; +Cc: Thomas Gleixner, Realtime Kernel
With these functions implemented we cooperate better with the generic
timekeeping code. This obsoletes the need for the timer sysdev as a bonus.
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
---
Patch set updated to powerpc/for-2.6.24
* Compile tested for arch/powerpc/configs/*_defconfig
* Booted on pSeries, iSeries, Cell and PS3
arch/powerpc/Kconfig | 3 +
arch/powerpc/kernel/time.c | 85 ++++++++++-------------------------
arch/powerpc/sysdev/Makefile | 5 --
arch/powerpc/sysdev/timer.c | 81 ---------------------------------
4 files changed, 29 insertions(+), 145 deletions(-)
Index: working/arch/powerpc/Kconfig
===================================================================
--- working.orig/arch/powerpc/Kconfig
+++ working/arch/powerpc/Kconfig
@@ -21,6 +21,9 @@ config MMU
bool
default y
+config GENERIC_CMOS_UPDATE
+ def_bool y
+
config GENERIC_HARDIRQS
bool
default y
Index: working/arch/powerpc/kernel/time.c
===================================================================
--- working.orig/arch/powerpc/kernel/time.c
+++ working/arch/powerpc/kernel/time.c
@@ -73,16 +73,11 @@
#include <asm/iseries/hv_call_xm.h>
#endif
-/* keep track of when we need to update the rtc */
-time_t last_rtc_update;
#ifdef CONFIG_PPC_ISERIES
static unsigned long __initdata iSeries_recal_titan;
static signed long __initdata iSeries_recal_tb;
#endif
-/* The decrementer counts down by 128 every 128ns on a 601. */
-#define DECREMENTER_COUNT_601 (1000000000 / HZ)
-
#define XSEC_PER_SEC (1024*1024)
#ifdef CONFIG_PPC64
@@ -348,39 +343,6 @@ void udelay(unsigned long usecs)
}
EXPORT_SYMBOL(udelay);
-static __inline__ void timer_check_rtc(void)
-{
- /*
- * update the rtc when needed, this should be performed on the
- * right fraction of a second. Half or full second ?
- * Full second works on mk48t59 clocks, others need testing.
- * Note that this update is basically only used through
- * the adjtimex system calls. Setting the HW clock in
- * any other way is a /dev/rtc and userland business.
- * This is still wrong by -0.5/+1.5 jiffies because of the
- * timer interrupt resolution and possible delay, but here we
- * hit a quantization limit which can only be solved by higher
- * resolution timers and decoupling time management from timer
- * interrupts. This is also wrong on the clocks
- * which require being written at the half second boundary.
- * We should have an rtc call that only sets the minutes and
- * seconds like on Intel to avoid problems with non UTC clocks.
- */
- if (ppc_md.set_rtc_time && ntp_synced() &&
- xtime.tv_sec - last_rtc_update >= 659 &&
- abs((xtime.tv_nsec/1000) - (1000000-1000000/HZ)) < 500000/HZ) {
- struct rtc_time tm;
- to_tm(xtime.tv_sec + 1 + timezone_offset, &tm);
- tm.tm_year -= 1900;
- tm.tm_mon -= 1;
- if (ppc_md.set_rtc_time(&tm) == 0)
- last_rtc_update = xtime.tv_sec + 1;
- else
- /* Try again one minute later */
- last_rtc_update += 60;
- }
-}
-
/*
* This version of gettimeofday has microsecond resolution.
*/
@@ -689,7 +651,6 @@ void timer_interrupt(struct pt_regs * re
tb_last_jiffy = tb_next_jiffy;
do_timer(1);
timer_recalc_offset(tb_last_jiffy);
- timer_check_rtc();
}
write_sequnlock(&xtime_lock);
}
@@ -801,11 +762,6 @@ int do_settimeofday(struct timespec *tv)
set_normalized_timespec(&xtime, new_sec, new_nsec);
set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec);
- /* In case of a large backwards jump in time with NTP, we want the
- * clock to be updated as soon as the PLL is again in lock.
- */
- last_rtc_update = new_sec - 658;
-
ntp_clear();
new_xsec = xtime.tv_nsec;
@@ -881,12 +837,35 @@ void __init generic_calibrate_decr(void)
#endif
}
-unsigned long get_boot_time(void)
+int update_persistent_clock(struct timespec now)
{
struct rtc_time tm;
- if (ppc_md.get_boot_time)
- return ppc_md.get_boot_time();
+ if (!ppc_md.set_rtc_time)
+ return 0;
+
+ to_tm(now.tv_sec + 1 + timezone_offset, &tm);
+ tm.tm_year -= 1900;
+ tm.tm_mon -= 1;
+
+ return ppc_md.set_rtc_time(&tm);
+}
+
+unsigned long read_persistent_clock(void)
+{
+ struct rtc_time tm;
+ static int first = 1;
+
+ /* XXX this is a litle fragile but will work okay in the short term */
+ if (first) {
+ first = 0;
+ if (ppc_md.time_init)
+ timezone_offset = ppc_md.time_init();
+
+ /* get_boot_time() isn't guaranteed to be safe to call late */
+ if (ppc_md.get_boot_time)
+ return ppc_md.get_boot_time() -timezone_offset;
+ }
if (!ppc_md.get_rtc_time)
return 0;
ppc_md.get_rtc_time(&tm);
@@ -898,14 +877,10 @@ unsigned long get_boot_time(void)
void __init time_init(void)
{
unsigned long flags;
- unsigned long tm = 0;
struct div_result res;
u64 scale, x;
unsigned shift;
- if (ppc_md.time_init != NULL)
- timezone_offset = ppc_md.time_init();
-
if (__USE_RTC()) {
/* 601 processor: dec counts down by 128 every 128ns */
ppc_tb_freq = 1000000000;
@@ -980,19 +955,14 @@ void __init time_init(void)
/* Save the current timebase to pretty up CONFIG_PRINTK_TIME */
boot_tb = get_tb_or_rtc();
- tm = get_boot_time();
-
write_seqlock_irqsave(&xtime_lock, flags);
/* If platform provided a timezone (pmac), we correct the time */
if (timezone_offset) {
sys_tz.tz_minuteswest = -timezone_offset / 60;
sys_tz.tz_dsttime = 0;
- tm -= timezone_offset;
}
- xtime.tv_sec = tm;
- xtime.tv_nsec = 0;
do_gtod.varp = &do_gtod.vars[0];
do_gtod.var_idx = 0;
do_gtod.varp->tb_orig_stamp = tb_last_jiffy;
@@ -1010,9 +980,6 @@ void __init time_init(void)
time_freq = 0;
- last_rtc_update = xtime.tv_sec;
- set_normalized_timespec(&wall_to_monotonic,
- -xtime.tv_sec, -xtime.tv_nsec);
write_sequnlock_irqrestore(&xtime_lock, flags);
/* Not exact, but the timer interrupt takes care of this */
Index: working/arch/powerpc/sysdev/Makefile
===================================================================
--- working.orig/arch/powerpc/sysdev/Makefile
+++ working/arch/powerpc/sysdev/Makefile
@@ -20,11 +20,6 @@ obj-$(CONFIG_MV64X60) += $(mv64x60-y) m
obj-$(CONFIG_RTC_DRV_CMOS) += rtc_cmos_setup.o
obj-$(CONFIG_AXON_RAM) += axonram.o
-# contains only the suspend handler for time
-ifeq ($(CONFIG_RTC_CLASS),)
-obj-$(CONFIG_PM) += timer.o
-endif
-
ifeq ($(CONFIG_PPC_MERGE),y)
obj-$(CONFIG_PPC_INDIRECT_PCI) += indirect_pci.o
obj-$(CONFIG_PPC_I8259) += i8259.o
Index: working/arch/powerpc/sysdev/timer.c
===================================================================
--- working.orig/arch/powerpc/sysdev/timer.c
+++ /dev/null
@@ -1,81 +0,0 @@
-/*
- * Common code to keep time when machine suspends.
- *
- * Copyright 2007 Johannes Berg <johannes@sipsolutions.net>
- *
- * GPLv2
- */
-
-#include <linux/time.h>
-#include <linux/sysdev.h>
-#include <asm/rtc.h>
-
-static unsigned long suspend_rtc_time;
-
-/*
- * Reset the time after a sleep.
- */
-static int timer_resume(struct sys_device *dev)
-{
- struct timeval tv;
- struct timespec ts;
- struct rtc_time cur_rtc_tm;
- unsigned long cur_rtc_time, diff;
-
- /* get current RTC time and convert to seconds */
- get_rtc_time(&cur_rtc_tm);
- cur_rtc_time = mktime(cur_rtc_tm.tm_year + 1900,
- cur_rtc_tm.tm_mon + 1,
- cur_rtc_tm.tm_mday,
- cur_rtc_tm.tm_hour,
- cur_rtc_tm.tm_min,
- cur_rtc_tm.tm_sec);
-
- diff = cur_rtc_time - suspend_rtc_time;
-
- /* adjust time of day by seconds that elapsed while
- * we were suspended */
- do_gettimeofday(&tv);
- ts.tv_sec = tv.tv_sec + diff;
- ts.tv_nsec = tv.tv_usec * NSEC_PER_USEC;
- do_settimeofday(&ts);
-
- return 0;
-}
-
-static int timer_suspend(struct sys_device *dev, pm_message_t state)
-{
- struct rtc_time suspend_rtc_tm;
- WARN_ON(!ppc_md.get_rtc_time);
-
- get_rtc_time(&suspend_rtc_tm);
- suspend_rtc_time = mktime(suspend_rtc_tm.tm_year + 1900,
- suspend_rtc_tm.tm_mon + 1,
- suspend_rtc_tm.tm_mday,
- suspend_rtc_tm.tm_hour,
- suspend_rtc_tm.tm_min,
- suspend_rtc_tm.tm_sec);
-
- return 0;
-}
-
-static struct sysdev_class timer_sysclass = {
- .resume = timer_resume,
- .suspend = timer_suspend,
- set_kset_name("timer"),
-};
-
-static struct sys_device device_timer = {
- .id = 0,
- .cls = &timer_sysclass,
-};
-
-static int time_init_device(void)
-{
- int error = sysdev_class_register(&timer_sysclass);
- if (!error)
- error = sysdev_register(&device_timer);
- return error;
-}
-
-device_initcall(time_init_device);
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox