[git pull] FireWire updates

From: Stefan Richter
Date: Sat Jul 04 2009 - 16:50:13 EST

Next message: Yinghai Lu: "Re: [tip:x86/cleanups] x86: Clean up mtrr/cleanup.c"
Previous message: Felix von Leitner: "Core i7 not scaling frequency down"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Linus, please pull from the for-linus branch at

git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6.git for-linus

to receive the following IEEE 1394/ FireWire subsystem updates.
These are recent but straight-forward fixes.

Stefan Richter (3):
firewire: core: do not DMA-map stack addresses
firewire: sbp2: add support for disks >2 TB (and 16 bytes long CDBs)
ieee1394: sbp2: add support for disks >2 TB (and 16 bytes long CDBs)

drivers/firewire/core-card.c | 14 +++++++-------
drivers/firewire/core-cdev.c | 4 +++-
drivers/firewire/core-iso.c | 24 +++++++++++++-----------
drivers/firewire/core.h | 3 ++-
drivers/firewire/sbp2.c | 10 +++++++++-
drivers/ieee1394/sbp2.c | 1 +
drivers/ieee1394/sbp2.h | 8 +++++++-
include/linux/firewire.h | 1 +
8 files changed, 43 insertions(+), 22 deletions(-)

commit ebbb16bffa646f853899ef3fdc0ac7abab888703
Author: Stefan Richter <stefanr@xxxxxxxxxxxxxxxxx>
Date: Tue Jun 30 20:28:31 2009 +0200

ieee1394: sbp2: add support for disks >2 TB (and 16 bytes long CDBs)

Increase the command ORB data structure to transport up to 16 bytes long
CDBs (instead of 12 bytes), and tell the SCSI mid layer about it. This
is notably necessary for READ CAPACITY(16) and friends, i.e. support of
large disks.

Signed-off-by: Stefan Richter <stefanr@xxxxxxxxxxxxxxxxx>

diff --git a/drivers/ieee1394/sbp2.c b/drivers/ieee1394/sbp2.c
index a51ab23..f599f49 100644
--- a/drivers/ieee1394/sbp2.c
+++ b/drivers/ieee1394/sbp2.c
@@ -880,6 +880,7 @@ static struct sbp2_lu *sbp2_alloc_device(struct unit_directory *ud)
}

shost->hostdata[0] = (unsigned long)lu;
+ shost->max_cmd_len = SBP2_MAX_CDB_SIZE;

if (!scsi_add_host(shost, &ud->device)) {
lu->shost = shost;
diff --git a/drivers/ieee1394/sbp2.h b/drivers/ieee1394/sbp2.h
index c5036f1..64a3a66 100644
--- a/drivers/ieee1394/sbp2.h
+++ b/drivers/ieee1394/sbp2.h
@@ -25,6 +25,12 @@
#define SBP2_DEVICE_NAME "sbp2"

/*
+ * There is no transport protocol limit to the CDB length, but we implement
+ * a fixed length only. 16 bytes is enough for disks larger than 2 TB.
+ */
+#define SBP2_MAX_CDB_SIZE 16
+
+/*
* SBP-2 specific definitions
*/

@@ -51,7 +57,7 @@ struct sbp2_command_orb {
u32 data_descriptor_hi;
u32 data_descriptor_lo;
u32 misc;
- u8 cdb[12];
+ u8 cdb[SBP2_MAX_CDB_SIZE];
} __attribute__((packed));

#define SBP2_LOGIN_REQUEST 0x0

commit af2719415a5ceae06f2a6d33e78b555e64697fc8
Author: Stefan Richter <stefanr@xxxxxxxxxxxxxxxxx>
Date: Tue Jun 30 20:27:59 2009 +0200

firewire: sbp2: add support for disks >2 TB (and 16 bytes long CDBs)

Increase the command ORB data structure to transport up to 16 bytes long
CDBs (instead of 12 bytes), and tell the SCSI mid layer about it. This
is notably necessary for READ CAPACITY(16) and friends, i.e. support of
large disks.

Signed-off-by: Stefan Richter <stefanr@xxxxxxxxxxxxxxxxx>

diff --git a/drivers/firewire/sbp2.c b/drivers/firewire/sbp2.c
index 2353643..d27cb05 100644
--- a/drivers/firewire/sbp2.c
+++ b/drivers/firewire/sbp2.c
@@ -201,6 +201,12 @@ static struct fw_device *target_device(struct sbp2_target *tgt)
#define SBP2_CYCLE_LIMIT (0xc8 << 12) /* 200 125us cycles */

/*
+ * There is no transport protocol limit to the CDB length, but we implement
+ * a fixed length only. 16 bytes is enough for disks larger than 2 TB.
+ */
+#define SBP2_MAX_CDB_SIZE 16
+
+/*
* The default maximum s/g segment size of a FireWire controller is
* usually 0x10000, but SBP-2 only allows 0xffff. Since buffers have to
* be quadlet-aligned, we set the length limit to 0xffff & ~3.
@@ -312,7 +318,7 @@ struct sbp2_command_orb {
struct sbp2_pointer next;
struct sbp2_pointer data_descriptor;
__be32 misc;
- u8 command_block[12];
+ u8 command_block[SBP2_MAX_CDB_SIZE];
} request;
struct scsi_cmnd *cmd;
scsi_done_fn_t done;
@@ -1146,6 +1152,8 @@ static int sbp2_probe(struct device *dev)
if (fw_device_enable_phys_dma(device) < 0)
goto fail_shost_put;

+ shost->max_cmd_len = SBP2_MAX_CDB_SIZE;
+
if (scsi_add_host(shost, &unit->device) < 0)
goto fail_shost_put;

commit 6fdc03709433ccc2005f0f593ae9d9dd04f7b485
Author: Stefan Richter <stefanr@xxxxxxxxxxxxxxxxx>
Date: Sat Jun 20 13:23:59 2009 +0200

firewire: core: do not DMA-map stack addresses

The DMA mapping API cannot map on-stack addresses, as explained in
Documentation/DMA-mapping.txt. Convert the two cases of on-stack packet
payload buffers in firewire-core (payload of lock requests in the bus
manager work and in iso resource management) to slab-allocated memory.

There are a number on-stack buffers for quadlet write or quadlet read
requests in firewire-core and firewire-sbp2. These are harmless; they
are copied to/ from card driver internal DMA buffers since quadlet
payloads are inlined with packet headers.

Signed-off-by: Stefan Richter <stefanr@xxxxxxxxxxxxxxxxx>

diff --git a/drivers/firewire/core-card.c b/drivers/firewire/core-card.c
index 543fcca..f74edae 100644
--- a/drivers/firewire/core-card.c
+++ b/drivers/firewire/core-card.c
@@ -196,8 +196,8 @@ static void allocate_broadcast_channel(struct fw_card *card, int generation)
{
int channel, bandwidth = 0;

- fw_iso_resource_manage(card, generation, 1ULL << 31,
- &channel, &bandwidth, true);
+ fw_iso_resource_manage(card, generation, 1ULL << 31, &channel,
+ &bandwidth, true, card->bm_transaction_data);
if (channel == 31) {
card->broadcast_channel_allocated = true;
device_for_each_child(card->device, (void *)(long)generation,
@@ -230,7 +230,6 @@ static void fw_card_bm_work(struct work_struct *work)
bool do_reset = false;
bool root_device_is_running;
bool root_device_is_cmc;
- __be32 lock_data[2];

spin_lock_irqsave(&card->lock, flags);

@@ -273,22 +272,23 @@ static void fw_card_bm_work(struct work_struct *work)
goto pick_me;
}

- lock_data[0] = cpu_to_be32(0x3f);
- lock_data[1] = cpu_to_be32(local_id);
+ card->bm_transaction_data[0] = cpu_to_be32(0x3f);
+ card->bm_transaction_data[1] = cpu_to_be32(local_id);

spin_unlock_irqrestore(&card->lock, flags);

rcode = fw_run_transaction(card, TCODE_LOCK_COMPARE_SWAP,
irm_id, generation, SCODE_100,
CSR_REGISTER_BASE + CSR_BUS_MANAGER_ID,
- lock_data, sizeof(lock_data));
+ card->bm_transaction_data,
+ sizeof(card->bm_transaction_data));

if (rcode == RCODE_GENERATION)
/* Another bus reset, BM work has been rescheduled. */
goto out;

if (rcode == RCODE_COMPLETE &&
- lock_data[0] != cpu_to_be32(0x3f)) {
+ card->bm_transaction_data[0] != cpu_to_be32(0x3f)) {

/* Somebody else is BM. Only act as IRM. */
if (local_id == irm_id)
diff --git a/drivers/firewire/core-cdev.c b/drivers/firewire/core-cdev.c
index d1d30c6..ced186d 100644
--- a/drivers/firewire/core-cdev.c
+++ b/drivers/firewire/core-cdev.c
@@ -125,6 +125,7 @@ struct iso_resource {
int generation;
u64 channels;
s32 bandwidth;
+ __be32 transaction_data[2];
struct iso_resource_event *e_alloc, *e_dealloc;
};

@@ -1049,7 +1050,8 @@ static void iso_resource_work(struct work_struct *work)
r->channels, &channel, &bandwidth,
todo == ISO_RES_ALLOC ||
todo == ISO_RES_REALLOC ||
- todo == ISO_RES_ALLOC_ONCE);
+ todo == ISO_RES_ALLOC_ONCE,
+ r->transaction_data);
/*
* Is this generation outdated already? As long as this resource sticks
* in the idr, it will be scheduled again for a newer generation or at
diff --git a/drivers/firewire/core-iso.c b/drivers/firewire/core-iso.c
index 166f19c..110e731 100644
--- a/drivers/firewire/core-iso.c
+++ b/drivers/firewire/core-iso.c
@@ -177,9 +177,8 @@ EXPORT_SYMBOL(fw_iso_context_stop);
*/

static int manage_bandwidth(struct fw_card *card, int irm_id, int generation,
- int bandwidth, bool allocate)
+ int bandwidth, bool allocate, __be32 data[2])
{
- __be32 data[2];
int try, new, old = allocate ? BANDWIDTH_AVAILABLE_INITIAL : 0;

/*
@@ -215,9 +214,9 @@ static int manage_bandwidth(struct fw_card *card, int irm_id, int generation,
}

static int manage_channel(struct fw_card *card, int irm_id, int generation,
- u32 channels_mask, u64 offset, bool allocate)
+ u32 channels_mask, u64 offset, bool allocate, __be32 data[2])
{
- __be32 data[2], c, all, old;
+ __be32 c, all, old;
int i, retry = 5;

old = all = allocate ? cpu_to_be32(~0) : 0;
@@ -260,7 +259,7 @@ static int manage_channel(struct fw_card *card, int irm_id, int generation,
}

static void deallocate_channel(struct fw_card *card, int irm_id,
- int generation, int channel)
+ int generation, int channel, __be32 buffer[2])
{
u32 mask;
u64 offset;
@@ -269,7 +268,7 @@ static void deallocate_channel(struct fw_card *card, int irm_id,
offset = channel < 32 ? CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_HI :
CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_LO;

- manage_channel(card, irm_id, generation, mask, offset, false);
+ manage_channel(card, irm_id, generation, mask, offset, false, buffer);
}

/**
@@ -298,7 +297,7 @@ static void deallocate_channel(struct fw_card *card, int irm_id,
*/
void fw_iso_resource_manage(struct fw_card *card, int generation,
u64 channels_mask, int *channel, int *bandwidth,
- bool allocate)
+ bool allocate, __be32 buffer[2])
{
u32 channels_hi = channels_mask; /* channels 31...0 */
u32 channels_lo = channels_mask >> 32; /* channels 63...32 */
@@ -310,10 +309,12 @@ void fw_iso_resource_manage(struct fw_card *card, int generation,

if (channels_hi)
c = manage_channel(card, irm_id, generation, channels_hi,
- CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_HI, allocate);
+ CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_HI,
+ allocate, buffer);
if (channels_lo && c < 0) {
c = manage_channel(card, irm_id, generation, channels_lo,
- CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_LO, allocate);
+ CSR_REGISTER_BASE + CSR_CHANNELS_AVAILABLE_LO,
+ allocate, buffer);
if (c >= 0)
c += 32;
}
@@ -325,12 +326,13 @@ void fw_iso_resource_manage(struct fw_card *card, int generation,
if (*bandwidth == 0)
return;

- ret = manage_bandwidth(card, irm_id, generation, *bandwidth, allocate);
+ ret = manage_bandwidth(card, irm_id, generation, *bandwidth,
+ allocate, buffer);
if (ret < 0)
*bandwidth = 0;

if (allocate && ret < 0 && c >= 0) {
- deallocate_channel(card, irm_id, generation, c);
+ deallocate_channel(card, irm_id, generation, c, buffer);
*channel = ret;
}
}
diff --git a/drivers/firewire/core.h b/drivers/firewire/core.h
index c3cfc64..6052816 100644
--- a/drivers/firewire/core.h
+++ b/drivers/firewire/core.h
@@ -120,7 +120,8 @@ void fw_node_event(struct fw_card *card, struct fw_node *node, int event);

int fw_iso_buffer_map(struct fw_iso_buffer *buffer, struct vm_area_struct *vma);
void fw_iso_resource_manage(struct fw_card *card, int generation,
- u64 channels_mask, int *channel, int *bandwidth, bool allocate);
+ u64 channels_mask, int *channel, int *bandwidth,
+ bool allocate, __be32 buffer[2]);

/* -topology */
diff --git a/include/linux/firewire.h b/include/linux/firewire.h
index 9823946..192d1e4 100644
--- a/include/linux/firewire.h
+++ b/include/linux/firewire.h
@@ -127,6 +127,7 @@ struct fw_card {
struct delayed_work work;
int bm_retries;
int bm_generation;
+ __be32 bm_transaction_data[2];

bool broadcast_channel_allocated;
u32 broadcast_channel;

Thanks,
--
Stefan Richter
-=====-==--= -=== --=--
http://arcgraph.de/sr/

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Next message: Yinghai Lu: "Re: [tip:x86/cleanups] x86: Clean up mtrr/cleanup.c"
Previous message: Felix von Leitner: "Core i7 not scaling frequency down"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]