- 09 Apr, 2012 40 commits
-
-
Jiri Slaby authored
All the needed stuff is moved to 68328serial.c now. Signed-off-by:
Jiri Slaby <jslaby@suse.cz> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: linux-m68k@lists.linux-m68k.org Acked-by:
Greg Ungerer <gerg@uclinux.org> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Jiri Slaby authored
flags passed to local_irq_save/restore should be ulong. Switch tehem to that. Otherwise we get compilation warnings: .../68328serial.c:248:9: warning: comparison of distinct pointer types lacks a cast [enabled by default] .../68328serial.c:257:9: warning: comparison of distinct pointer types lacks a cast [enabled by default] Signed-off-by:
-
Jiri Slaby authored
- empty functions - unused global variables Signed-off-by:
-
Jiri Slaby authored
Not everything from struct m68k_serial is really used. So remove unused or only-set members of that structure. Next step is to move it to 68328serial.c and remove 68328serial.h completely. This change also takes status_handle and batten_down_hatches away since they use break_abort but do nothing. Signed-off-by:
-
Jiri Slaby authored
serial_state in 68328serial.h is a duplicated structure. One is defined in linux/serial.h. So let us use that instead. And since the serial flags are identical, use ones from there too. Signed-off-by:
-
Jiri Slaby authored
It does not make the driver less racy though. Close and hangup should be rewritten and tty refcounting used properly. Signed-off-by:
David Sterba <dsterba@suse.cz> Signed-off-by:
-
Jiri Slaby authored
And use count from that. Signed-off-by:
-
Jiri Slaby authored
There are two functions which only print a status. Let us do that directly at places where they are called. Signed-off-by:
-
Jiri Slaby authored
Do not touch internal workqueue. Call tty_vhangup instead. Note that finished hangup does not necessarily mean that all processes are dead. Especially when the tty is a console. The code assumes that right now. Signed-off-by:
-
Jiri Slaby authored
Now, we switch to the refcounted model and do not need hp->lock to protect hp->tty anymore. Signed-off-by:
-
Jiri Slaby authored
- use tty, not hp->tty wherever possible - pass tty down to some functions and go to step one - do not defer tty_hangup calls -- it is as simple as schedule_work, so might be called with hp->lock held - do not defer tty buffer flips -- since the driver does not use low_latency (it cannot actually), the flip is a simple tail move plus schedule_work. It will make our life easier in the next patch. Signed-off-by:
-
Jiri Slaby authored
And use count from there. Signed-off-by:
-
Jiri Slaby authored
It is in termios cflags. So change the test in hvsi_recv_control to do the right thing. Previously it was actually testing TTY_LDISC_OPEN bit, i.e. whether an ldisc is active. And yes, it is most of the time. Signed-off-by:
-
Jiri Slaby authored
No refcounting, just a switch. The locking in the driver prevents races, so in fact the refcounting is not needed. But while we have a tty in tty_port, don't duplicate that and remove the one from hvcs_struct. Signed-off-by:
-
Jiri Slaby authored
A simple switch. Except we convert destroy_hvcs_struct to be tty_port_operations->destruct... Signed-off-by:
-
Jiri Slaby authored
And use count from there. Signed-off-by:
-
Jiri Slaby authored
Now, count is used from tty_port and protected by tty_port->lock. n_outbuf is left unprotected in hvc_hangup now, because there is no point to hold any lock, since other uses are unprotected too. Signed-off-by:
-
Jiri Slaby authored
The driver already used refcounting. So we just switch it to tty_port helpers. And switch to tty_port->lock for tty. Signed-off-by:
-
Jiri Slaby authored
And use kref from that. This means we need tty_port->ops->destruct to properly free the structure. This is what destroy_hvc_struct used to do so we leverage that. Signed-off-by:
-
Jiri Slaby authored
Switch from mutex to tty_port->lock and to tty refcounting. This needs a 'continue' to be added to re-grab a tty after schedule returns. And since tty is not protected by bfin_jc_tty_mutex remove it as well. But this needs tty_port->count to be protected by tty_port->lock now. Signed-off-by:
-
Jiri Slaby authored
And use open count from there. Switch to tty from there will follow. Signed-off-by:
-
Jiri Slaby authored
And use tty from that. This means, we convert most of the users to accept tty_port instead. This is not racy and ensures the tty to be properly refcounted. Signed-off-by:
-
Jiri Slaby authored
So that we do not need to access tp->tty there. It is going away. Signed-off-by:
-
Jiri Slaby authored
Blah, do not assume that raw3270_view is at the beginning of tty3270. Use proper types and container_of wherever needed. Signed-off-by:
-
Jiri Slaby authored
Let us initialize all the tty3270's members at the place where the structure is allocated. It cleans up tty->ops->open a bit. Signed-off-by:
-
Jiri Slaby authored
80 bytes which are allocated in tty->ops->open and assigned to tty->driver_data are never used. Remove that. Signed-off-by:
-
Jiri Slaby authored
tty_port will hold tty buffers in the future. So we need to have it even here. The only needed member here is tty, so let us store it in the structure now. Signed-off-by:
-
Jiri Slaby authored
tty_port will hold tty buffers in the future. So we need to have it even here. The only needed member here is tty, so let us store it in the structure now. Signed-off-by:
-
Jiri Slaby authored
There are two copies of allocations of device information. One of them is totally broken. See: raw->cdev = cdev; raw->inbuf = (char *) raw + sizeof(struct raw3215_info); memset(raw, 0, sizeof(struct raw3215_info)); It suggests that this path was never executed. The code uses both raw->cdev and raw->inbuf all over. And it is NULL due to the memset here, so it would panic immediately. I believe nobody used that driver without being a system console. Either way, let us fix it by moving the allocations (and initializations) to a single place. This will save us some double initializations later too. And while at it, initialize the timer properly -- once, at the allocation. Signed-off-by:
Heiko Carstens <heiko.carstens@de.ibm.com> Cc: linux390@de.ibm.com Cc: linux-s390@vger.kernel.org Signed-off-by:
-
Jiri Slaby authored
We switched tty refcounting there to the one provided by tty_port helpers. So tty_port->tty is now protected by tty_port->lock, not by hso_serial->serial_lock. Side note: tty->driver_data does not need the lock, so it is needed neither in open, nor in close paths. Signed-off-by:
-
Jiri Slaby authored
tty is never NULL in tty->ops->* while the device is open. (And they are not called otherwise.) So remove pointless checks and use tty->driver_data directly. Signed-off-by:
-
Jiri Slaby authored
And use open count from there. Other members will follow. Remark: port.count is (and never was) properly protected. Only a mutex is held, so ISR and all the functions it calls may see an invalid state. Signed-off-by:
-
Jiri Slaby authored
Do not leak tty_driver structure on each module removal. Also do proper frees in fail paths of module_init. Signed-off-by:
-
Jiri Slaby authored
It is set in alloc_tty_driver already. No need to re-set. Signed-off-by:
-
Jiri Slaby authored
This removes a bunch of duplicated code which does the same as tty_port_block_til_ready does. Signed-off-by:
-
Jiri Slaby authored
Add there .carrier_raised. It is taken from current block_til_ready. We will need tty_port->ops->carrier_raised for tty_port_block_til_ready helper. Signed-off-by:
-
Jiri Slaby authored
The code does almost the same, so there we can leverage the helper's code. The only difference is locking. The helper protects counts by a spinlock. This never hurts and should be added to other code parts too. Signed-off-by:
-
Jiri Slaby authored
In some functions we use tty_port much. So let us have a local pointer to that variable instead of having info->port all over the code. Signed-off-by:
-
Jiri Slaby authored
Signed-off-by:
-
Jiri Slaby authored
No recounting this time, just a plain switch. Signed-off-by:
-
