* [PATCH 02/11] staging/fwserial: Release port regardless of unplug response code
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-29 3:34 ` [PATCH 03/11] staging/fwserial: Fix sparse build warnings Peter Hurley
` (8 subsequent siblings)
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
After sending the unplug response, release the port even if an
error occurred.
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index b1482ef..055085d 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -2675,10 +2675,9 @@ static void fwserial_handle_unplug_req(struct work_struct *work)
spin_lock_bh(&peer->lock);
if (peer->state == FWPS_UNPLUG_RESPONDING) {
- if (rcode == RCODE_COMPLETE)
- port = peer_revert_state(peer);
- else
+ if (rcode != RCODE_COMPLETE)
fwtty_err(&peer->unit, "UNPLUG_RSP error (%d)", rcode);
+ port = peer_revert_state(peer);
}
cleanup:
spin_unlock_bh(&peer->lock);
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* [PATCH 03/11] staging/fwserial: Fix sparse build warnings
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
2013-01-29 3:34 ` [PATCH 02/11] staging/fwserial: Release port regardless of unplug response code Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-29 3:34 ` [PATCH 04/11] staging/fwserial: Create loop device the 'tty' way Peter Hurley
` (7 subsequent siblings)
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
Reported-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.c | 4 +++-
drivers/staging/fwserial/fwserial.h | 2 +-
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index 055085d..473a2db 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -78,6 +78,8 @@ static int num_ports;
/* slab used as pool for struct fwtty_transactions */
static struct kmem_cache *fwtty_txn_cache;
+struct tty_driver *fwtty_driver;
+
struct fwtty_transaction;
typedef void (*fwtty_transaction_cb)(struct fw_card *card, int rcode,
void *data, size_t length,
@@ -2543,7 +2545,7 @@ static struct fw_descriptor fwserial_unit_directory = {
* The management address is in the unit space region but above other known
* address users (to keep wild writes from causing havoc)
*/
-const struct fw_address_region fwserial_mgmt_addr_region = {
+static const struct fw_address_region fwserial_mgmt_addr_region = {
.start = CSR_REGISTER_BASE + 0x1e0000ULL,
.end = 0x1000000000000ULL,
};
diff --git a/drivers/staging/fwserial/fwserial.h b/drivers/staging/fwserial/fwserial.h
index 953ece6..3602809 100644
--- a/drivers/staging/fwserial/fwserial.h
+++ b/drivers/staging/fwserial/fwserial.h
@@ -352,7 +352,7 @@ struct fw_serial {
static const char tty_dev_name[] = TTY_DEV_NAME;
static const char loop_dev_name[] = "fwloop";
-struct tty_driver *fwtty_driver;
+extern struct tty_driver *fwtty_driver;
#define driver_err(s, v...) pr_err(KBUILD_MODNAME ": " s, ##v)
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* [PATCH 04/11] staging/fwserial: Create loop device the 'tty' way
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
2013-01-29 3:34 ` [PATCH 02/11] staging/fwserial: Release port regardless of unplug response code Peter Hurley
2013-01-29 3:34 ` [PATCH 03/11] staging/fwserial: Fix sparse build warnings Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-29 3:34 ` [PATCH 05/11] staging/fwserial: Cleanup /proc/tty/driver/ file Peter Hurley
` (6 subsequent siblings)
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
Register a second tty driver to create loopback devices for
each firewire node. Note that the loopback devices are numbered
from 0; the tty->index is transformed when used to index the
port table.
Remove the hack that previously enabled this.
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/TODO | 13 ---
drivers/staging/fwserial/fwserial.c | 176 +++++++++++++++++++-----------------
2 files changed, 93 insertions(+), 96 deletions(-)
diff --git a/drivers/staging/fwserial/TODO b/drivers/staging/fwserial/TODO
index dc61d97..382a795 100644
--- a/drivers/staging/fwserial/TODO
+++ b/drivers/staging/fwserial/TODO
@@ -12,16 +12,3 @@ TODOs prior to this driver moving out of staging
1. This driver uses the same unregistered vendor id that the firewire core does
(0xd00d1e). Perhaps this could be exposed as a define in
firewire.h?
-
--- Issues with TTY core --
- 1. Hack for alternate device name scheme
- - because udev no longer allows device renaming, devices should have
- their proper names on creation. This is an issue for creating the
- fwloop<n> device with the fwtty<n> devices because although duplicating
- roughly the same operations as tty_port_register_device() isn't difficult,
- access to the tty_class & tty_fops is restricted in scope.
-
- This is currently being worked around in create_loop_device() by
- extracting the tty_class ptr and tty_fops ptr from the previously created
- tty devices. Perhaps an add'l api can be added -- eg.,
- tty_{port_}register_named_device().
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index 473a2db..e143b7a 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -72,6 +72,9 @@ static DEFINE_MUTEX(port_table_lock);
static bool port_table_corrupt;
#define FWTTY_INVALID_INDEX MAX_TOTAL_PORTS
+#define loop_idx(port) (((port)->index) / num_ports)
+#define table_idx(loop) ((loop) * num_ports + num_ttys)
+
/* total # of tty ports created per fw_card */
static int num_ports;
@@ -79,6 +82,7 @@ static int num_ports;
static struct kmem_cache *fwtty_txn_cache;
struct tty_driver *fwtty_driver;
+static struct tty_driver *fwloop_driver;
struct fwtty_transaction;
typedef void (*fwtty_transaction_cb)(struct fw_card *card, int rcode,
@@ -1156,6 +1160,19 @@ static int fwtty_install(struct tty_driver *driver, struct tty_struct *tty)
return err;
}
+static int fwloop_install(struct tty_driver *driver, struct tty_struct *tty)
+{
+ struct fwtty_port *port = fwtty_port_get(table_idx(tty->index));
+ int err;
+
+ err = tty_standard_install(driver, tty);
+ if (!err)
+ tty->driver_data = port;
+ else
+ fwtty_port_put(port);
+ return err;
+}
+
static int fwtty_write(struct tty_struct *tty, const unsigned char *buf, int c)
{
struct fwtty_port *port = tty->driver_data;
@@ -1592,6 +1609,26 @@ static const struct tty_operations fwtty_ops = {
.proc_fops = &fwtty_proc_fops,
};
+static const struct tty_operations fwloop_ops = {
+ .open = fwtty_open,
+ .close = fwtty_close,
+ .hangup = fwtty_hangup,
+ .cleanup = fwtty_cleanup,
+ .install = fwloop_install,
+ .write = fwtty_write,
+ .write_room = fwtty_write_room,
+ .chars_in_buffer = fwtty_chars_in_buffer,
+ .send_xchar = fwtty_send_xchar,
+ .throttle = fwtty_throttle,
+ .unthrottle = fwtty_unthrottle,
+ .ioctl = fwtty_ioctl,
+ .set_termios = fwtty_set_termios,
+ .break_ctl = fwtty_break_ctl,
+ .tiocmget = fwtty_tiocmget,
+ .tiocmset = fwtty_tiocmset,
+ .get_icount = fwtty_get_icount,
+};
+
static inline int mgmt_pkt_expected_len(__be16 code)
{
static const struct fwserial_mgmt_pkt pkt;
@@ -1888,7 +1925,8 @@ free_pkt:
* The port reference is put by fwtty_cleanup (if a reference was
* ever taken).
*/
-static void fwserial_close_port(struct fwtty_port *port)
+static void fwserial_close_port(struct tty_driver *driver,
+ struct fwtty_port *port)
{
struct tty_struct *tty;
@@ -1900,7 +1938,10 @@ static void fwserial_close_port(struct fwtty_port *port)
}
mutex_unlock(&port->port.mutex);
- tty_unregister_device(fwtty_driver, port->index);
+ if (driver == fwloop_driver)
+ tty_unregister_device(driver, loop_idx(port));
+ else
+ tty_unregister_device(driver, port->index);
}
/**
@@ -2158,78 +2199,6 @@ static void fwserial_remove_peer(struct fwtty_peer *peer)
}
/**
- * create_loop_device - create a loopback tty device
- * @tty_driver: tty_driver to own loopback device
- * @prototype: ptr to already-assigned 'prototype' tty port
- * @index: index to associate this device with the tty port
- * @parent: device to child to
- *
- * HACK - this is basically tty_port_register_device() with an
- * alternate naming scheme. Suggest tty_port_register_named_device()
- * helper api.
- *
- * Creates a loopback tty device named 'fwloop<n>' which is attached to
- * the local unit in fwserial_add_peer(). Note that <n> in the device
- * name advances in increments of port allocation blocks, ie., for port
- * indices 0..3, the device name will be 'fwloop0'; for 4..7, 'fwloop1',
- * and so on.
- *
- * Only one loopback device should be created per fw_card.
- */
-static void release_loop_device(struct device *dev)
-{
- kfree(dev);
-}
-
-static struct device *create_loop_device(struct tty_driver *driver,
- struct fwtty_port *prototype,
- struct fwtty_port *port,
- struct device *parent)
-{
- char name[64];
- int index = port->index;
- dev_t devt = MKDEV(driver->major, driver->minor_start) + index;
- struct device *dev = NULL;
- int err;
-
- if (index >= fwtty_driver->num)
- return ERR_PTR(-EINVAL);
-
- snprintf(name, 64, "%s%d", loop_dev_name, index / num_ports);
-
- tty_port_link_device(&port->port, driver, index);
-
- cdev_init(&driver->cdevs[index], driver->cdevs[prototype->index].ops);
- driver->cdevs[index].owner = driver->owner;
- err = cdev_add(&driver->cdevs[index], devt, 1);
- if (err)
- return ERR_PTR(err);
-
- dev = kzalloc(sizeof(*dev), GFP_KERNEL);
- if (!dev) {
- cdev_del(&driver->cdevs[index]);
- return ERR_PTR(-ENOMEM);
- }
-
- dev->devt = devt;
- dev->class = prototype->device->class;
- dev->parent = parent;
- dev->release = release_loop_device;
- dev_set_name(dev, "%s", name);
- dev->groups = NULL;
- dev_set_drvdata(dev, NULL);
-
- err = device_register(dev);
- if (err) {
- put_device(dev);
- cdev_del(&driver->cdevs[index]);
- return ERR_PTR(err);
- }
-
- return dev;
-}
-
-/**
* fwserial_create - init everything to create TTYs for a specific fw_card
* @unit: fw_unit for first 'serial' unit device probed for this fw_card
*
@@ -2327,17 +2296,17 @@ static int fwserial_create(struct fw_unit *unit)
if (create_loop_dev) {
struct device *loop_dev;
- loop_dev = create_loop_device(fwtty_driver,
- serial->ports[0],
- serial->ports[num_ttys],
- card->device);
+ loop_dev = tty_port_register_device(&serial->ports[j]->port,
+ fwloop_driver,
+ loop_idx(serial->ports[j]),
+ card->device);
if (IS_ERR(loop_dev)) {
err = PTR_ERR(loop_dev);
fwtty_err(&unit, "create loop device failed (%d)", err);
goto unregister_ttys;
}
- serial->ports[num_ttys]->device = loop_dev;
- serial->ports[num_ttys]->loopback = true;
+ serial->ports[j]->device = loop_dev;
+ serial->ports[j]->loopback = true;
}
list_add_rcu(&serial->list, &fwserial_list);
@@ -2353,6 +2322,8 @@ static int fwserial_create(struct fw_unit *unit)
/* fall-through to error processing */
list_del_rcu(&serial->list);
+ if (create_loop_dev)
+ tty_unregister_device(fwloop_driver, loop_idx(serial->ports[j]));
unregister_ttys:
for (--j; j >= 0; --j)
tty_unregister_device(fwtty_driver, serial->ports[j]->index);
@@ -2441,8 +2412,10 @@ static int fwserial_remove(struct device *dev)
/* unlink from the fwserial_list here */
list_del_rcu(&serial->list);
- for (i = 0; i < num_ports; ++i)
- fwserial_close_port(serial->ports[i]);
+ for (i = 0; i < num_ttys; ++i)
+ fwserial_close_port(fwtty_driver, serial->ports[i]);
+ if (create_loop_dev)
+ fwserial_close_port(fwloop_driver, serial->ports[i]);
kref_put(&serial->kref, fwserial_destroy);
}
mutex_unlock(&fwserial_list_mutex);
@@ -2863,12 +2836,39 @@ static int __init fwserial_init(void)
goto put_tty;
}
+ if (create_loop_dev) {
+ fwloop_driver = alloc_tty_driver(MAX_TOTAL_PORTS / num_ports);
+ if (!fwloop_driver) {
+ err = -ENOMEM;
+ goto unregister_driver;
+ }
+
+ fwloop_driver->driver_name = KBUILD_MODNAME "_loop";
+ fwloop_driver->name = loop_dev_name;
+ fwloop_driver->major = 0;
+ fwloop_driver->minor_start = 0;
+ fwloop_driver->type = TTY_DRIVER_TYPE_SERIAL;
+ fwloop_driver->subtype = SERIAL_TYPE_NORMAL;
+ fwloop_driver->flags = TTY_DRIVER_REAL_RAW |
+ TTY_DRIVER_DYNAMIC_DEV;
+
+ fwloop_driver->init_termios = tty_std_termios;
+ fwloop_driver->init_termios.c_cflag |= CLOCAL;
+ tty_set_operations(fwloop_driver, &fwloop_ops);
+
+ err = tty_register_driver(fwloop_driver);
+ if (err) {
+ driver_err("register loop driver failed (%d)", err);
+ goto put_loop;
+ }
+ }
+
fwtty_txn_cache = kmem_cache_create("fwtty_txn_cache",
sizeof(struct fwtty_transaction),
0, 0, fwtty_txn_constructor);
if (!fwtty_txn_cache) {
err = -ENOMEM;
- goto unregister_driver;
+ goto unregister_loop;
}
/*
@@ -2910,6 +2910,12 @@ remove_handler:
fw_core_remove_address_handler(&fwserial_mgmt_addr_handler);
destroy_cache:
kmem_cache_destroy(fwtty_txn_cache);
+unregister_loop:
+ if (create_loop_dev)
+ tty_unregister_driver(fwloop_driver);
+put_loop:
+ if (create_loop_dev)
+ put_tty_driver(fwloop_driver);
unregister_driver:
tty_unregister_driver(fwtty_driver);
put_tty:
@@ -2923,6 +2929,10 @@ static void __exit fwserial_exit(void)
fw_core_remove_descriptor(&fwserial_unit_directory);
fw_core_remove_address_handler(&fwserial_mgmt_addr_handler);
kmem_cache_destroy(fwtty_txn_cache);
+ if (create_loop_dev) {
+ tty_unregister_driver(fwloop_driver);
+ put_tty_driver(fwloop_driver);
+ }
tty_unregister_driver(fwtty_driver);
put_tty_driver(fwtty_driver);
}
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* [PATCH 05/11] staging/fwserial: Cleanup /proc/tty/driver/ file
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
` (2 preceding siblings ...)
2013-01-29 3:34 ` [PATCH 04/11] staging/fwserial: Create loop device the 'tty' way Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-29 3:34 ` [PATCH 06/11] staging/fwserial: Factor unstable stats/debug/status info to debugfs Peter Hurley
` (5 subsequent siblings)
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
Factor out extra stats, data profiles, debugging info and peer info
from procfs file in preparation for using debugfs instead.
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.c | 39 ++++++++++++++-----------------------
1 file changed, 15 insertions(+), 24 deletions(-)
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index e143b7a..ba8ffa2 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -1500,17 +1500,26 @@ static void fwtty_proc_show_port(struct seq_file *m, struct fwtty_port *port)
if (port->port.console)
(*port->fwcon_ops->stats)(&stats, port->con_data);
- seq_printf(m, " tx:%d rx:%d", port->icount.tx + stats.xchars,
- port->icount.rx);
+ seq_printf(m, " addr:%012llx tx:%d rx:%d", port->rx_handler.offset,
+ port->icount.tx + stats.xchars, port->icount.rx);
seq_printf(m, " cts:%d dsr:%d rng:%d dcd:%d", port->icount.cts,
port->icount.dsr, port->icount.rng, port->icount.dcd);
seq_printf(m, " fe:%d oe:%d pe:%d brk:%d", port->icount.frame,
port->icount.overrun, port->icount.parity, port->icount.brk);
+}
+
+static void fwtty_debugfs_show_port(struct seq_file *m, struct fwtty_port *port)
+{
+ struct stats stats;
+
+ memcpy(&stats, &port->stats, sizeof(stats));
+ if (port->port.console)
+ (*port->fwcon_ops->stats)(&stats, port->con_data);
+
seq_printf(m, " dr:%d st:%d err:%d lost:%d", stats.dropped,
stats.tx_stall, stats.fifo_errs, stats.lost);
seq_printf(m, " pkts:%d thr:%d wtrmk:%d", stats.sent, stats.throttled,
stats.watermark);
- seq_printf(m, " addr:%012llx", port->rx_handler.offset);
if (port->port.console) {
seq_printf(m, "\n ");
@@ -1520,7 +1529,7 @@ static void fwtty_proc_show_port(struct seq_file *m, struct fwtty_port *port)
dump_profile(m, &port->stats);
}
-static void fwtty_proc_show_peer(struct seq_file *m, struct fwtty_peer *peer)
+static void fwtty_debugfs_show_peer(struct seq_file *m, struct fwtty_peer *peer)
{
int generation = peer->generation;
@@ -1529,21 +1538,14 @@ static void fwtty_proc_show_peer(struct seq_file *m, struct fwtty_peer *peer)
seq_printf(m, " node:%04x gen:%d", peer->node_id, generation);
seq_printf(m, " sp:%d max:%d guid:%016llx", peer->speed,
peer->max_payload, (unsigned long long) peer->guid);
-
- if (capable(CAP_SYS_ADMIN)) {
- seq_printf(m, " mgmt:%012llx",
- (unsigned long long) peer->mgmt_addr);
- seq_printf(m, " addr:%012llx",
- (unsigned long long) peer->status_addr);
- }
+ seq_printf(m, " mgmt:%012llx", (unsigned long long) peer->mgmt_addr);
+ seq_printf(m, " addr:%012llx", (unsigned long long) peer->status_addr);
seq_putc(m, '\n');
}
static int fwtty_proc_show(struct seq_file *m, void *v)
{
struct fwtty_port *port;
- struct fw_serial *serial;
- struct fwtty_peer *peer;
int i;
seq_puts(m, "fwserinfo: 1.0 driver: 1.0\n");
@@ -1554,17 +1556,6 @@ static int fwtty_proc_show(struct seq_file *m, void *v)
fwtty_port_put(port);
seq_printf(m, "\n");
}
- seq_putc(m, '\n');
-
- rcu_read_lock();
- list_for_each_entry_rcu(serial, &fwserial_list, list) {
- seq_printf(m, "card: %s guid: %016llx\n",
- dev_name(serial->card->device),
- (unsigned long long) serial->card->guid);
- list_for_each_entry_rcu(peer, &serial->peer_list, list)
- fwtty_proc_show_peer(m, peer);
- }
- rcu_read_unlock();
return 0;
}
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* [PATCH 06/11] staging/fwserial: Factor unstable stats/debug/status info to debugfs
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
` (3 preceding siblings ...)
2013-01-29 3:34 ` [PATCH 05/11] staging/fwserial: Cleanup /proc/tty/driver/ file Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-29 3:34 ` [PATCH 07/11] staging/fwserial: Don't use deprecated alloc_tty_driver() Peter Hurley
` (4 subsequent siblings)
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
Add the following file hierarchy to debugfs:
<debugfs>-+
+- firewire_serial -+- <unit> -+- peers
| +- stats
|
+- <unit> -+- peers
+- stats
The 'peers' file (read-only) contains status and configuration
info for attached peers for the given fwserial unit.
The 'stats' file (read-only) contains statistics and data profiling
information for each tty port for the given fwserial unit.
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.c | 82 +++++++++++++++++++++++++++++++++++++
drivers/staging/fwserial/fwserial.h | 2 +
2 files changed, 84 insertions(+)
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index ba8ffa2..66463d1 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -84,6 +84,8 @@ static struct kmem_cache *fwtty_txn_cache;
struct tty_driver *fwtty_driver;
static struct tty_driver *fwloop_driver;
+static struct dentry *fwserial_debugfs;
+
struct fwtty_transaction;
typedef void (*fwtty_transaction_cb)(struct fw_card *card, int rcode,
void *data, size_t length,
@@ -1559,11 +1561,71 @@ static int fwtty_proc_show(struct seq_file *m, void *v)
return 0;
}
+static int fwtty_debugfs_stats_show(struct seq_file *m, void *v)
+{
+ struct fw_serial *serial = m->private;
+ struct fwtty_port *port;
+ int i;
+
+ for (i = 0; i < num_ports; ++i) {
+ port = fwtty_port_get(serial->ports[i]->index);
+ if (port) {
+ seq_printf(m, "%2d:", port->index);
+ fwtty_proc_show_port(m, port);
+ fwtty_debugfs_show_port(m, port);
+ fwtty_port_put(port);
+ seq_printf(m, "\n");
+ }
+ }
+ return 0;
+}
+
+static int fwtty_debugfs_peers_show(struct seq_file *m, void *v)
+{
+ struct fw_serial *serial = m->private;
+ struct fwtty_peer *peer;
+
+ rcu_read_lock();
+ seq_printf(m, "card: %s guid: %016llx\n",
+ dev_name(serial->card->device),
+ (unsigned long long) serial->card->guid);
+ list_for_each_entry_rcu(peer, &serial->peer_list, list)
+ fwtty_debugfs_show_peer(m, peer);
+ rcu_read_unlock();
+ return 0;
+}
+
static int fwtty_proc_open(struct inode *inode, struct file *fp)
{
return single_open(fp, fwtty_proc_show, NULL);
}
+static int fwtty_stats_open(struct inode *inode, struct file *fp)
+{
+ return single_open(fp, fwtty_debugfs_stats_show, inode->i_private);
+}
+
+static int fwtty_peers_open(struct inode *inode, struct file *fp)
+{
+ return single_open(fp, fwtty_debugfs_peers_show, inode->i_private);
+}
+
+static const struct file_operations fwtty_stats_fops = {
+ .owner = THIS_MODULE,
+ .open = fwtty_stats_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = single_release,
+};
+
+static const struct file_operations fwtty_peers_fops = {
+ .owner = THIS_MODULE,
+ .open = fwtty_peers_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = single_release,
+};
+
static const struct file_operations fwtty_proc_fops = {
.owner = THIS_MODULE,
.open = fwtty_proc_open,
@@ -2300,6 +2362,17 @@ static int fwserial_create(struct fw_unit *unit)
serial->ports[j]->loopback = true;
}
+ if (!IS_ERR_OR_NULL(fwserial_debugfs)) {
+ serial->debugfs = debugfs_create_dir(dev_name(&unit->device),
+ fwserial_debugfs);
+ if (!IS_ERR_OR_NULL(serial->debugfs)) {
+ debugfs_create_file("peers", 0444, serial->debugfs,
+ serial, &fwtty_peers_fops);
+ debugfs_create_file("stats", 0444, serial->debugfs,
+ serial, &fwtty_stats_fops);
+ }
+ }
+
list_add_rcu(&serial->list, &fwserial_list);
fwtty_notice(&unit, "TTY over FireWire on device %s (guid %016llx)",
@@ -2312,6 +2385,8 @@ static int fwserial_create(struct fw_unit *unit)
fwtty_err(&unit, "unable to add peer unit device (%d)", err);
/* fall-through to error processing */
+ debugfs_remove_recursive(serial->debugfs);
+
list_del_rcu(&serial->list);
if (create_loop_dev)
tty_unregister_device(fwloop_driver, loop_idx(serial->ports[j]));
@@ -2403,6 +2478,8 @@ static int fwserial_remove(struct device *dev)
/* unlink from the fwserial_list here */
list_del_rcu(&serial->list);
+ debugfs_remove_recursive(serial->debugfs);
+
for (i = 0; i < num_ttys; ++i)
fwserial_close_port(fwtty_driver, serial->ports[i]);
if (create_loop_dev)
@@ -2797,6 +2874,9 @@ static int __init fwserial_init(void)
{
int err, num_loops = !!(create_loop_dev);
+ /* XXX: placeholder for a "firewire" debugfs node */
+ fwserial_debugfs = debugfs_create_dir(KBUILD_MODNAME, NULL);
+
/* num_ttys/num_ports must not be set above the static alloc avail */
if (num_ttys + num_loops > MAX_CARD_PORTS)
num_ttys = MAX_CARD_PORTS - num_loops;
@@ -2911,6 +2991,7 @@ unregister_driver:
tty_unregister_driver(fwtty_driver);
put_tty:
put_tty_driver(fwtty_driver);
+ debugfs_remove_recursive(fwserial_debugfs);
return err;
}
@@ -2926,6 +3007,7 @@ static void __exit fwserial_exit(void)
}
tty_unregister_driver(fwtty_driver);
put_tty_driver(fwtty_driver);
+ debugfs_remove_recursive(fwserial_debugfs);
}
module_init(fwserial_init);
diff --git a/drivers/staging/fwserial/fwserial.h b/drivers/staging/fwserial/fwserial.h
index 3602809..c768aad 100644
--- a/drivers/staging/fwserial/fwserial.h
+++ b/drivers/staging/fwserial/fwserial.h
@@ -15,6 +15,7 @@
#include <linux/serial_reg.h>
#include <linux/module.h>
#include <linux/seq_file.h>
+#include <linux/debugfs.h>
#include "dma_fifo.h"
@@ -338,6 +339,7 @@ struct fw_serial {
struct fw_card *card;
struct kref kref;
+ struct dentry *debugfs;
struct fwtty_peer *self;
struct list_head list;
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* [PATCH 07/11] staging/fwserial: Don't use deprecated alloc_tty_driver()
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
` (4 preceding siblings ...)
2013-01-29 3:34 ` [PATCH 06/11] staging/fwserial: Factor unstable stats/debug/status info to debugfs Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-29 3:34 ` [PATCH 08/11] staging/fwserial: Remove reference to removed constant Peter Hurley
` (3 subsequent siblings)
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
Use tty_alloc_driver() instead.
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.c | 21 +++++++++------------
1 file changed, 9 insertions(+), 12 deletions(-)
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index 66463d1..912b17a 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -2882,9 +2882,10 @@ static int __init fwserial_init(void)
num_ttys = MAX_CARD_PORTS - num_loops;
num_ports = num_ttys + num_loops;
- fwtty_driver = alloc_tty_driver(MAX_TOTAL_PORTS);
- if (!fwtty_driver) {
- err = -ENOMEM;
+ fwtty_driver = tty_alloc_driver(MAX_TOTAL_PORTS, TTY_DRIVER_REAL_RAW
+ | TTY_DRIVER_DYNAMIC_DEV);
+ if (IS_ERR(fwtty_driver)) {
+ err = PTR_ERR(fwtty_driver);
return err;
}
@@ -2894,9 +2895,6 @@ static int __init fwserial_init(void)
fwtty_driver->minor_start = 0;
fwtty_driver->type = TTY_DRIVER_TYPE_SERIAL;
fwtty_driver->subtype = SERIAL_TYPE_NORMAL;
- fwtty_driver->flags = TTY_DRIVER_REAL_RAW |
- TTY_DRIVER_DYNAMIC_DEV;
-
fwtty_driver->init_termios = tty_std_termios;
fwtty_driver->init_termios.c_cflag |= CLOCAL;
tty_set_operations(fwtty_driver, &fwtty_ops);
@@ -2908,9 +2906,11 @@ static int __init fwserial_init(void)
}
if (create_loop_dev) {
- fwloop_driver = alloc_tty_driver(MAX_TOTAL_PORTS / num_ports);
- if (!fwloop_driver) {
- err = -ENOMEM;
+ fwloop_driver = tty_alloc_driver(MAX_TOTAL_PORTS / num_ports,
+ TTY_DRIVER_REAL_RAW
+ | TTY_DRIVER_DYNAMIC_DEV);
+ if (IS_ERR(fwloop_driver)) {
+ err = PTR_ERR(fwloop_driver);
goto unregister_driver;
}
@@ -2920,9 +2920,6 @@ static int __init fwserial_init(void)
fwloop_driver->minor_start = 0;
fwloop_driver->type = TTY_DRIVER_TYPE_SERIAL;
fwloop_driver->subtype = SERIAL_TYPE_NORMAL;
- fwloop_driver->flags = TTY_DRIVER_REAL_RAW |
- TTY_DRIVER_DYNAMIC_DEV;
-
fwloop_driver->init_termios = tty_std_termios;
fwloop_driver->init_termios.c_cflag |= CLOCAL;
tty_set_operations(fwloop_driver, &fwloop_ops);
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* [PATCH 08/11] staging/fwserial: Remove reference to removed constant
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
` (5 preceding siblings ...)
2013-01-29 3:34 ` [PATCH 07/11] staging/fwserial: Don't use deprecated alloc_tty_driver() Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-29 3:34 ` [PATCH 09/11] staging/fwserial: add diagnostic for buffer overflow Peter Hurley
` (2 subsequent siblings)
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
FWSERIAL_TTY_START_MINOR was removed. The minor_start is allocated
by tty_alloc_driver().
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/staging/fwserial/fwserial.h b/drivers/staging/fwserial/fwserial.h
index c768aad..33a3a53 100644
--- a/drivers/staging/fwserial/fwserial.h
+++ b/drivers/staging/fwserial/fwserial.h
@@ -194,7 +194,7 @@ struct buffered_rx {
* @port: underlying tty_port
* @device: tty device
* @index: index into port_table for this particular port
- * note: minor = index + FWSERIAL_TTY_START_MINOR
+ * note: minor = index + minor_start assigned by tty_alloc_driver()
* @serial: back pointer to the containing fw_serial
* @rx_handler: bus address handler for unique addr region used by remotes
* to communicate with this port. Every port uses
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* [PATCH 09/11] staging/fwserial: add diagnostic for buffer overflow
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
` (6 preceding siblings ...)
2013-01-29 3:34 ` [PATCH 08/11] staging/fwserial: Remove reference to removed constant Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-29 3:34 ` [PATCH 10/11] staging/fwserial: Fix premature unthrottle Peter Hurley
2013-01-29 3:34 ` [PATCH 11/11] staging/fwserial: Remove unneeded push work Peter Hurley
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index 912b17a..912ab77 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -569,8 +569,11 @@ static int fwtty_buffer_rx(struct fwtty_port *port, unsigned char *d, size_t n)
struct buffered_rx *buf;
size_t size = (n + sizeof(struct buffered_rx) + 0xFF) & ~0xFF;
- if (port->buffered + n > HIGH_WATERMARK)
+ if (port->buffered + n > HIGH_WATERMARK) {
+ fwtty_err_ratelimited(port, "overflowed rx buffer: buffered: %d new: %ld wtrmk: %d",
+ port->buffered, n, HIGH_WATERMARK);
return 0;
+ }
buf = kmalloc(size, GFP_ATOMIC);
if (!buf)
return 0;
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* [PATCH 10/11] staging/fwserial: Fix premature unthrottle
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
` (7 preceding siblings ...)
2013-01-29 3:34 ` [PATCH 09/11] staging/fwserial: add diagnostic for buffer overflow Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
2013-01-30 4:29 ` Greg Kroah-Hartman
2013-01-29 3:34 ` [PATCH 11/11] staging/fwserial: Remove unneeded push work Peter Hurley
9 siblings, 1 reply; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
The unthrottle may restart the writer before the rx push
has a chance to start emptying the rx buffer, resulting in an
overflowed rx buffer and lost data.
Perform the actual device unthrottle with the rx push instead.
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.c | 30 ++++++++++++++----------------
1 file changed, 14 insertions(+), 16 deletions(-)
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index 912ab77..121beff 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -463,6 +463,14 @@ static void __fwtty_throttle(struct fwtty_port *port, struct tty_struct *tty)
__fwtty_write_port_status(port);
}
+static void __fwtty_unthrottle(struct fwtty_port *port, struct tty_struct *tty)
+{
+ port->mctrl &= ~OOB_RX_THROTTLE;
+ if (C_CRTSCTS(tty))
+ port->mctrl |= TIOCM_RTS;
+ __fwtty_write_port_status(port);
+}
+
/**
* fwtty_do_hangup - wait for ldisc to deliver all pending rx; only then hangup
*
@@ -527,14 +535,15 @@ static void fwtty_emit_breaks(struct work_struct *work)
port->icount.brk += brk;
}
-static void fwtty_pushrx(struct work_struct *work)
+static void fwtty_pushrx(struct fwtty_port *port, struct tty_struct *tty)
{
- struct fwtty_port *port = to_port(work, push);
- struct tty_struct *tty;
struct buffered_rx *buf, *next;
int n, c = 0;
spin_lock_bh(&port->lock);
+
+ __fwtty_unthrottle(port, tty);
+
list_for_each_entry_safe(buf, next, &port->buf_list, list) {
n = tty_insert_flip_string_fixed_flag(&port->port, buf->data,
TTY_NORMAL, buf->n);
@@ -545,11 +554,7 @@ static void fwtty_pushrx(struct work_struct *work)
memmove(buf->data, buf->data + n, buf->n - n);
buf->n -= n;
}
- tty = tty_port_tty_get(&port->port);
- if (tty) {
- __fwtty_throttle(port, tty);
- tty_kref_put(tty);
- }
+ __fwtty_throttle(port, tty);
break;
} else {
list_del(&buf->list);
@@ -1264,14 +1269,7 @@ static void fwtty_unthrottle(struct tty_struct *tty)
profile_fifo_avail(port, port->stats.unthrottle);
- schedule_work(&port->push);
-
- spin_lock_bh(&port->lock);
- port->mctrl &= ~OOB_RX_THROTTLE;
- if (C_CRTSCTS(tty))
- port->mctrl |= TIOCM_RTS;
- __fwtty_write_port_status(port);
- spin_unlock_bh(&port->lock);
+ fwtty_pushrx(port, tty);
}
static int check_msr_delta(struct fwtty_port *port, unsigned long mask,
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread* Re: [PATCH 10/11] staging/fwserial: Fix premature unthrottle
2013-01-29 3:34 ` [PATCH 10/11] staging/fwserial: Fix premature unthrottle Peter Hurley
@ 2013-01-30 4:29 ` Greg Kroah-Hartman
2013-01-30 13:11 ` Peter Hurley
0 siblings, 1 reply; 16+ messages in thread
From: Greg Kroah-Hartman @ 2013-01-30 4:29 UTC (permalink / raw)
To: Peter Hurley; +Cc: linux1394-devel, linux-kernel, Stefan Richter
On Mon, Jan 28, 2013 at 10:34:44PM -0500, Peter Hurley wrote:
> The unthrottle may restart the writer before the rx push
> has a chance to start emptying the rx buffer, resulting in an
> overflowed rx buffer and lost data.
>
> Perform the actual device unthrottle with the rx push instead.
>
> Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
> ---
> drivers/staging/fwserial/fwserial.c | 30 ++++++++++++++----------------
> 1 file changed, 14 insertions(+), 16 deletions(-)
this patch fails to apply to my tree, but the 9 previous ones were fine,
so I'll not apply it, or ght 11/11 patch. Care to fix it up and resend
both of them?
thanks,
greg k-h
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH 10/11] staging/fwserial: Fix premature unthrottle
2013-01-30 4:29 ` Greg Kroah-Hartman
@ 2013-01-30 13:11 ` Peter Hurley
2013-01-30 13:32 ` Greg Kroah-Hartman
0 siblings, 1 reply; 16+ messages in thread
From: Peter Hurley @ 2013-01-30 13:11 UTC (permalink / raw)
To: Greg Kroah-Hartman; +Cc: linux1394-devel, linux-kernel, Stefan Richter
On Tue, 2013-01-29 at 23:29 -0500, Greg Kroah-Hartman wrote:
> On Mon, Jan 28, 2013 at 10:34:44PM -0500, Peter Hurley wrote:
> > The unthrottle may restart the writer before the rx push
> > has a chance to start emptying the rx buffer, resulting in an
> > overflowed rx buffer and lost data.
> >
> > Perform the actual device unthrottle with the rx push instead.
> >
> > Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
> > ---
> > drivers/staging/fwserial/fwserial.c | 30 ++++++++++++++----------------
> > 1 file changed, 14 insertions(+), 16 deletions(-)
>
> this patch fails to apply to my tree, but the 9 previous ones were fine,
> so I'll not apply it, or ght 11/11 patch. Care to fix it up and resend
> both of them?
Hi Greg,
I can rebase these against your staging-next, if you'd prefer, but it
will make the -next merge non-trivial.
Your tty-next tree is carrying patches to this driver from patch series
"TTY: switch flipping functions to tty_port"
which is already in -next.
Let me know what you'd like me to do.
Regards,
Peter Hurley
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH 10/11] staging/fwserial: Fix premature unthrottle
2013-01-30 13:11 ` Peter Hurley
@ 2013-01-30 13:32 ` Greg Kroah-Hartman
2013-01-30 13:37 ` Peter Hurley
0 siblings, 1 reply; 16+ messages in thread
From: Greg Kroah-Hartman @ 2013-01-30 13:32 UTC (permalink / raw)
To: Peter Hurley; +Cc: linux1394-devel, linux-kernel, Stefan Richter
On Wed, Jan 30, 2013 at 08:11:05AM -0500, Peter Hurley wrote:
> On Tue, 2013-01-29 at 23:29 -0500, Greg Kroah-Hartman wrote:
> > On Mon, Jan 28, 2013 at 10:34:44PM -0500, Peter Hurley wrote:
> > > The unthrottle may restart the writer before the rx push
> > > has a chance to start emptying the rx buffer, resulting in an
> > > overflowed rx buffer and lost data.
> > >
> > > Perform the actual device unthrottle with the rx push instead.
> > >
> > > Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
> > > ---
> > > drivers/staging/fwserial/fwserial.c | 30 ++++++++++++++----------------
> > > 1 file changed, 14 insertions(+), 16 deletions(-)
> >
> > this patch fails to apply to my tree, but the 9 previous ones were fine,
> > so I'll not apply it, or ght 11/11 patch. Care to fix it up and resend
> > both of them?
>
> Hi Greg,
>
> I can rebase these against your staging-next, if you'd prefer, but it
> will make the -next merge non-trivial.
>
> Your tty-next tree is carrying patches to this driver from patch series
> "TTY: switch flipping functions to tty_port"
> which is already in -next.
>
> Let me know what you'd like me to do.
Ugh, that's a mess.
Can these wait until after 3.9-rc1 is out? If so, things should get
synced up by then. Otherwise I'll have to figure out how to handle the
merge properly...
thanks,
greg k-h
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH 10/11] staging/fwserial: Fix premature unthrottle
2013-01-30 13:32 ` Greg Kroah-Hartman
@ 2013-01-30 13:37 ` Peter Hurley
2013-01-30 16:25 ` Greg Kroah-Hartman
0 siblings, 1 reply; 16+ messages in thread
From: Peter Hurley @ 2013-01-30 13:37 UTC (permalink / raw)
To: Greg Kroah-Hartman; +Cc: linux1394-devel, linux-kernel, Stefan Richter
On Wed, 2013-01-30 at 14:32 +0100, Greg Kroah-Hartman wrote:
> On Wed, Jan 30, 2013 at 08:11:05AM -0500, Peter Hurley wrote:
> > On Tue, 2013-01-29 at 23:29 -0500, Greg Kroah-Hartman wrote:
> > > On Mon, Jan 28, 2013 at 10:34:44PM -0500, Peter Hurley wrote:
> > > > The unthrottle may restart the writer before the rx push
> > > > has a chance to start emptying the rx buffer, resulting in an
> > > > overflowed rx buffer and lost data.
> > > >
> > > > Perform the actual device unthrottle with the rx push instead.
> > > >
> > > > Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
> > > > ---
> > > > drivers/staging/fwserial/fwserial.c | 30 ++++++++++++++----------------
> > > > 1 file changed, 14 insertions(+), 16 deletions(-)
> > >
> > > this patch fails to apply to my tree, but the 9 previous ones were fine,
> > > so I'll not apply it, or ght 11/11 patch. Care to fix it up and resend
> > > both of them?
> >
> > Hi Greg,
> >
> > I can rebase these against your staging-next, if you'd prefer, but it
> > will make the -next merge non-trivial.
> >
> > Your tty-next tree is carrying patches to this driver from patch series
> > "TTY: switch flipping functions to tty_port"
> > which is already in -next.
> >
> > Let me know what you'd like me to do.
>
> Ugh, that's a mess.
>
> Can these wait until after 3.9-rc1 is out?
Ok. Should I plan to resend patches 10 & 11 then (as 1 & 2)?
Regards,
Peter Hurley
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH 10/11] staging/fwserial: Fix premature unthrottle
2013-01-30 13:37 ` Peter Hurley
@ 2013-01-30 16:25 ` Greg Kroah-Hartman
0 siblings, 0 replies; 16+ messages in thread
From: Greg Kroah-Hartman @ 2013-01-30 16:25 UTC (permalink / raw)
To: Peter Hurley; +Cc: linux1394-devel, linux-kernel, Stefan Richter
On Wed, Jan 30, 2013 at 08:37:23AM -0500, Peter Hurley wrote:
> On Wed, 2013-01-30 at 14:32 +0100, Greg Kroah-Hartman wrote:
> > On Wed, Jan 30, 2013 at 08:11:05AM -0500, Peter Hurley wrote:
> > > On Tue, 2013-01-29 at 23:29 -0500, Greg Kroah-Hartman wrote:
> > > > On Mon, Jan 28, 2013 at 10:34:44PM -0500, Peter Hurley wrote:
> > > > > The unthrottle may restart the writer before the rx push
> > > > > has a chance to start emptying the rx buffer, resulting in an
> > > > > overflowed rx buffer and lost data.
> > > > >
> > > > > Perform the actual device unthrottle with the rx push instead.
> > > > >
> > > > > Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
> > > > > ---
> > > > > drivers/staging/fwserial/fwserial.c | 30 ++++++++++++++----------------
> > > > > 1 file changed, 14 insertions(+), 16 deletions(-)
> > > >
> > > > this patch fails to apply to my tree, but the 9 previous ones were fine,
> > > > so I'll not apply it, or ght 11/11 patch. Care to fix it up and resend
> > > > both of them?
> > >
> > > Hi Greg,
> > >
> > > I can rebase these against your staging-next, if you'd prefer, but it
> > > will make the -next merge non-trivial.
> > >
> > > Your tty-next tree is carrying patches to this driver from patch series
> > > "TTY: switch flipping functions to tty_port"
> > > which is already in -next.
> > >
> > > Let me know what you'd like me to do.
> >
> > Ugh, that's a mess.
> >
> > Can these wait until after 3.9-rc1 is out?
>
> Ok. Should I plan to resend patches 10 & 11 then (as 1 & 2)?
Yes, please do.
thanks,
greg k-h
^ permalink raw reply [flat|nested] 16+ messages in thread
* [PATCH 11/11] staging/fwserial: Remove unneeded push work
2013-01-29 3:34 [PATCH 01/11] staging/fwserial: Only reset port status for attached peers Peter Hurley
` (8 preceding siblings ...)
2013-01-29 3:34 ` [PATCH 10/11] staging/fwserial: Fix premature unthrottle Peter Hurley
@ 2013-01-29 3:34 ` Peter Hurley
9 siblings, 0 replies; 16+ messages in thread
From: Peter Hurley @ 2013-01-29 3:34 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: linux1394-devel, linux-kernel, Stefan Richter, Peter Hurley
Since rx push can be performed synchronously with unthrottle,
remove the push work struct and code references to it.
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
---
drivers/staging/fwserial/fwserial.c | 2 --
drivers/staging/fwserial/fwserial.h | 4 ----
2 files changed, 6 deletions(-)
diff --git a/drivers/staging/fwserial/fwserial.c b/drivers/staging/fwserial/fwserial.c
index 121beff..8983911 100644
--- a/drivers/staging/fwserial/fwserial.c
+++ b/drivers/staging/fwserial/fwserial.c
@@ -1112,7 +1112,6 @@ static void fwtty_port_shutdown(struct tty_port *tty_port)
cancel_delayed_work_sync(&port->emit_breaks);
cancel_delayed_work_sync(&port->drain);
- cancel_work_sync(&port->push);
spin_lock_bh(&port->lock);
list_for_each_entry_safe(buf, next, &port->buf_list, list) {
@@ -2300,7 +2299,6 @@ static int fwserial_create(struct fw_unit *unit)
INIT_DELAYED_WORK(&port->drain, fwtty_drain_tx);
INIT_DELAYED_WORK(&port->emit_breaks, fwtty_emit_breaks);
INIT_WORK(&port->hangup, fwtty_do_hangup);
- INIT_WORK(&port->push, fwtty_pushrx);
INIT_LIST_HEAD(&port->buf_list);
init_waitqueue_head(&port->wait_tx);
port->max_payload = link_speed_to_max_payload(SCODE_100);
diff --git a/drivers/staging/fwserial/fwserial.h b/drivers/staging/fwserial/fwserial.h
index 33a3a53..6c179f0 100644
--- a/drivers/staging/fwserial/fwserial.h
+++ b/drivers/staging/fwserial/fwserial.h
@@ -223,9 +223,6 @@ struct buffered_rx {
* The work can race with the writer but concurrent sending is
* prevented with the IN_TX flag. Scheduled under lock to
* limit scheduling when fifo has just been drained.
- * @push: work responsible for pushing buffered rx to the ldisc.
- * rx can become buffered if the tty buffer is filled before the
- * ldisc throttles the sender.
* @buf_list: list of buffered rx yet to be sent to ldisc
* @buffered: byte count of buffered rx
* @tx_fifo: fifo used to store & block-up writes for dma to remote
@@ -267,7 +264,6 @@ struct fwtty_port {
spinlock_t lock;
unsigned mctrl;
struct delayed_work drain;
- struct work_struct push;
struct list_head buf_list;
int buffered;
struct dma_fifo tx_fifo;
--
1.8.1.1
^ permalink raw reply related [flat|nested] 16+ messages in thread