tty: Implement exclusive mode

Implemented exclusive mode on Haiku and added the related `ioctl`
operations (`TIOCEXCL` and `TIOCNXCL`).

Change-Id: Iaa201ea20eec0e45d02dd5db9ba6aa35fd27dfb2
Reviewed-on: https://review.haiku-os.org/c/haiku/+/6387
Tested-by: Commit checker robot <no-reply+buildbot@haiku-os.org>
Reviewed-by: Jessica Hamilton <jessica.l.hamilton@gmail.com>

generic/tty: restore lock sharing between master and slave TTYs

The code in this module was derived from the one in driver/tty. However,
the driver uses a shared lock between the master and slave side of a
TTY, and this was changed to use two separate locks. The approach with
two locks does not work. It seems the change was unfinished and the
second TTY was never locked. But attempting to lock it will result in
lock inversion problems, unless we do complicated things (try to find
which of the two TTY is the master side, and lock that first, for
example). It is simpler to restore the shared lock as used in the
driver.

To set up the shared lock, I modified the tty_create function to take a
pointer to the master TTY when creating the slave. Maybe it makes more
sense to create both sides in the same call, create_tty_pair?

However, this does not work as easily as I wanted, because there is some
recursion going on: at least in one case, the tty_control function is
calling the driver's tty_service function, which in turns attempts to
call back into tty_control for the "other side" TTY. To handle this
case, replace the mutex with a recursive_lock.

Fixes #17091, where the root problem was access to
other_tty->select_pool without locking. This was also made unconvenient
to debug because select_pool objects are self-deleting, when the last
item in the pool is removed. As a result, the code accessing it without
log would suddenly find out that the data it was accessing had been
freed and erased.

This also makes the TTY code in driver/tty and generic/tty a bit more
similar than it was before, and brings us one step closer to merging the
two together. There are still two main differences and I don't know
enough about TTY to decide if they are important, and which version
should be kept:
- The driver has extra code for "background" read and write. I don't
know what this is used for.
- The driver has a single "settings" instance shared by a master and
slave TTY, while the module has two separate instances, but seems to
copy one to the other. I'm not sure which approach is correct.

Change-Id: Ie2daddd027859ce32ba395af76b4f109f8b984b0
Reviewed-on: https://review.haiku-os.org/c/haiku/+/4604
Reviewed-by: waddlesplash <waddlesplash@gmail.com>

tty: implement TCFLSH to flush buffers

This is probably incomplete. Is locking needed? Should we notify the
next writer (if any) that the port is writable when flushing the output?

Change-Id: I2566e2d036a61af4819894a44f57603179aa27df
Reviewed-on: https://review.haiku-os.org/c/haiku/+/2516
Reviewed-by: Adrien Destugues <pulkomandy@gmail.com>

pc_serial: mask THRE again when we're out of data

pc_serial: wait for writes to complete on a sem

This avoids close being called too early and discarding output.

pc_serial: toggle THRE in IER to make VirtualBox happy

Not sure it's needed on real hardware yet.

pc_serial: cache IER and temporarily mask IRQs until dpc done

No need to be flooded by IRQs while filling in the FIFO...

pc_serial: apply lcr parameters even when unmasking latch reg

No need to change line params even more if it was already the same before.

pc_serial: try to enable 64 byte FIFO

We should probably detect the UART type properly first,
but the IRQ handler also checks for the available FIFO length anyway.

pc_serial: skip waiting for pending DPC in free hook

Since it's possible to fail queuing them, the count doesn't go down to 0,
and the process just hangs forever on exit. It might not be necessary anyway.

pc_serial: break instead of continue in irq handler

We don't want to busy loop, do we?

pc_serial: add a termios member to store config across open cycles

pc_serial: fix KDL on close

It's working \o/
Just need to figure out why it waits for \n to transmit.

We now only destroy the cookies in the free hook.

Mention pending DPC as soon as when inside the irq handler.

Also limit the number of loops in the IRQ handler to avoid busy looping.
It only happened when testing with normal priority but you never know...

pc_serial: use the tty module directly from the deferred IRQ

It still crashes on close sometimes but writing seems to work.

pc_serial: manually probe the default ISA ports

Since config_manager is basically a stub, and even VirtualBox
doesn't publish the serial ports as PCI devices, we must probe for them.

Ideally config_manager would find them in the device manager tree
which would have been populated from the PnP BIOS or even ACPI tables...

pc_serial: convert irq handler to use DPC

Let's hope we won't loose data because it.

We have to cache the value of IIR read from IsInterruptPending(),
because some conditions are acknowledged only by reading it...

pc_serial: WIP: kinda supports writes through tty module

It's still not fully working, and the tty module doesn't help
since it's not callable from an irq handler.

pc_serial: replace contiguous area by arrays as buffers

We don't need to DMA unlike usb_serial.

pc_serial: Remove read and write lock

As with hrev42089 for usb_serial.

pc_serial: Drop BeOS stuff

Even I don't use it anymore...

pc_serial: Use two tty cookies

pc_serial: Fix TTYSETRTS

Same fix as hrev42151 for usb_serial:
When determining the bit to be set for RTS/DTR there was a missing comparison
to op which would've caused a wrong line state to be set.

Port pc_serial to the new tty module.
Builds but won't work because the tty module lacks essential calls to read and write data :^)
Lots of ifdef since I need the R5 code around for now.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@39791 a95241bf-73f2-0310-859d-f6bbb57e9c96

Update copyrights. Mention TuneTracker Systems as a sponsor.
Add several ways to restrict UART probing when matching devices:
- max port count,
- PCI subsystem ID mask (some cards encode the port count there),
- Base Address Register mask (some cards only have UART in the first 2 ranges).
Use those to match 2 specific NetMos chips, and avoid screwing up when input_server tries to access UARTs at boggus ranges.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35965 a95241bf-73f2-0310-859d-f6bbb57e9c96

Less tracing by default...


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35936 a95241bf-73f2-0310-859d-f6bbb57e9c96

Change dprintfs to TRACE(), and make always on tracing less verbose.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34571 a95241bf-73f2-0310-859d-f6bbb57e9c96

Working version of the driver. At least it does seem to work with my onboard com port, but not the OxfordSemi PCI card I have here. Still requires tty_manager, and config_manager too, so won't work in Haiku yet. Needs cleanup.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@29328 a95241bf-73f2-0310-859d-f6bbb57e9c96

Start of a PC-style serial port driver. For now it's mostly just copied parts of usb_serial, not yet usable.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@27111 a95241bf-73f2-0310-859d-f6bbb57e9c96

pc_serial: mask THRE again when we're out of data

pc_serial: wait for writes to complete on a sem

This avoids close being called too early and discarding output.

pc_serial: toggle THRE in IER to make VirtualBox happy

Not sure it's needed on real hardware yet.

pc_serial: cache IER and temporarily mask IRQs until dpc done

No need to be flooded by IRQs while filling in the FIFO...

pc_serial: apply lcr parameters even when unmasking latch reg

No need to change line params even more if it was already the same before.

pc_serial: try to enable 64 byte FIFO

We should probably detect the UART type properly first,
but the IRQ handler also checks for the available FIFO length anyway.

pc_serial: skip waiting for pending DPC in free hook

Since it's possible to fail queuing them, the count doesn't go down to 0,
and the process just hangs forever on exit. It might not be necessary anyway.

pc_serial: break instead of continue in irq handler

We don't want to busy loop, do we?

pc_serial: add a termios member to store config across open cycles

pc_serial: fix KDL on close

It's working \o/
Just need to figure out why it waits for \n to transmit.

We now only destroy the cookies in the free hook.

Mention pending DPC as soon as when inside the irq handler.

Also limit the number of loops in the IRQ handler to avoid busy looping.
It only happened when testing with normal priority but you never know...

pc_serial: use the tty module directly from the deferred IRQ

It still crashes on close sometimes but writing seems to work.

pc_serial: manually probe the default ISA ports

Since config_manager is basically a stub, and even VirtualBox
doesn't publish the serial ports as PCI devices, we must probe for them.

Ideally config_manager would find them in the device manager tree
which would have been populated from the PnP BIOS or even ACPI tables...

pc_serial: convert irq handler to use DPC

Let's hope we won't loose data because it.

We have to cache the value of IIR read from IsInterruptPending(),
because some conditions are acknowledged only by reading it...

pc_serial: WIP: kinda supports writes through tty module

It's still not fully working, and the tty module doesn't help
since it's not callable from an irq handler.

pc_serial: replace contiguous area by arrays as buffers

We don't need to DMA unlike usb_serial.

pc_serial: Remove read and write lock

As with hrev42089 for usb_serial.

pc_serial: Drop BeOS stuff

Even I don't use it anymore...

pc_serial: Use two tty cookies

pc_serial: Fix TTYSETRTS

Same fix as hrev42151 for usb_serial:
When determining the bit to be set for RTS/DTR there was a missing comparison
to op which would've caused a wrong line state to be set.

Port pc_serial to the new tty module.
Builds but won't work because the tty module lacks essential calls to read and write data :^)
Lots of ifdef since I need the R5 code around for now.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@39791 a95241bf-73f2-0310-859d-f6bbb57e9c96

Update copyrights. Mention TuneTracker Systems as a sponsor.
Add several ways to restrict UART probing when matching devices:
- max port count,
- PCI subsystem ID mask (some cards encode the port count there),
- Base Address Register mask (some cards only have UART in the first 2 ranges).
Use those to match 2 specific NetMos chips, and avoid screwing up when input_server tries to access UARTs at boggus ranges.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35965 a95241bf-73f2-0310-859d-f6bbb57e9c96

Less tracing by default...


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@35936 a95241bf-73f2-0310-859d-f6bbb57e9c96

Change dprintfs to TRACE(), and make always on tracing less verbose.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34571 a95241bf-73f2-0310-859d-f6bbb57e9c96

Working version of the driver. At least it does seem to work with my onboard com port, but not the OxfordSemi PCI card I have here. Still requires tty_manager, and config_manager too, so won't work in Haiku yet. Needs cleanup.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@29328 a95241bf-73f2-0310-859d-f6bbb57e9c96

Start of a PC-style serial port driver. For now it's mostly just copied parts of usb_serial, not yet usable.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@27111 a95241bf-73f2-0310-859d-f6bbb57e9c96