linux-toradex.git/drivers/serial/8250.c, branch v2.6.19.2

IRQ: Maintain regs pointer globally rather than passing to IRQ handlers

2006-10-05T14:10:12+00:00

Maintain a per-CPU global "struct pt_regs *" variable which can be used instead
of passing regs around manually through all ~1800 interrupt handlers in the
Linux kernel.

The regs pointer is used in few places, but it potentially costs both stack
space and code to pass it around.  On the FRV arch, removing the regs parameter
from all the genirq function results in a 20% speed up of the IRQ exit path
(ie: from leaving timer_interrupt() to leaving do_IRQ()).

Where appropriate, an arch may override the generic storage facility and do
something different with the variable.  On FRV, for instance, the address is
maintained in GR28 at all times inside the kernel as part of general exception
handling.

Having looked over the code, it appears that the parameter may be handed down
through up to twenty or so layers of functions.  Consider a USB character
device attached to a USB hub, attached to a USB controller that posts its
interrupts through a cascaded auxiliary interrupt controller.  A character
device driver may want to pass regs to the sysrq handler through the input
layer which adds another few layers of parameter passing.

I've build this code with allyesconfig for x86_64 and i386.  I've runtested the
main part of the code on FRV and i386, though I can't test most of the drivers.
I've also done partial conversion for powerpc and MIPS - these at least compile
with minimal configurations.

This will affect all archs.  Mostly the changes should be relatively easy.
Take do_IRQ(), store the regs pointer at the beginning, saving the old one:

	struct pt_regs *old_regs = set_irq_regs(regs);

And put the old one back at the end:

	set_irq_regs(old_regs);

Don't pass regs through to generic_handle_irq() or __do_IRQ().

In timer_interrupt(), this sort of change will be necessary:

	-	update_process_times(user_mode(regs));
	-	profile_tick(CPU_PROFILING, regs);
	+	update_process_times(user_mode(get_irq_regs()));
	+	profile_tick(CPU_PROFILING);

I'd like to move update_process_times()'s use of get_irq_regs() into itself,
except that i386, alone of the archs, uses something other than user_mode().

Some notes on the interrupt handling in the drivers:

 (*) input_dev() is now gone entirely.  The regs pointer is no longer stored in
     the input_dev struct.

 (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking.  It does
     something different depending on whether it's been supplied with a regs
     pointer or not.

 (*) Various IRQ handler function pointers have been moved to type
     irq_handler_t.

Signed-Off-By: David Howells 
(cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)

Merge branch 'devel' of master.kernel.org:/home/rmk/linux-2.6-serial

2006-10-03T16:13:29+00:00

* 'devel' of master.kernel.org:/home/rmk/linux-2.6-serial: (21 commits)
  [SERIAL] add PNP IDs for FPI based touchscreens
  [SERIAL] Magic SysRq SAK does nothing on serial consoles
  [SERIAL] tickle NMI watchdog on serial output.
  [SERIAL] Fix oops when removing suspended serial port
  [SERIAL] Fix resume handling bug
  [SERIAL] Remove wrong asm/serial.h inclusions
  [SERIAL] CONFIG_PM=n slim: drivers/serial/8250_pci.c
  [SERIAL] OMAP1510 serial fix for 115200 baud
  [SERIAL] returning proper error from serial core driver
  [SERIAL] Make uart_line_info() correctly tell MMIO from I/O port
  [SERIAL] suspend/resume handlers don't have level arg anymore
  [SERIAL] 8250 resourse management fixes
  [SERIAL] serial_cs: Add quirk for brainboxes 2-port RS232 card
  [SERIAL] serial_cs: handle Nokia multi->single port bodge via config quirk
  [SERIAL] serial_cs: add configuration quirk
  [SERIAL] serial_cs: Convert Oxford 950 / Possio GCC wakeup quirk
  [SERIAL] serial_cs: convert IBM post-init handling to a quirk
  [SERIAL] serial_cs: allow wildcarded quirks
  [SERIAL] serial_cs: convert multi-port table to quirk table
  [SERIAL] serial_cs: Use clean up multiport card detection
  ...

[SERIAL] tickle NMI watchdog on serial output.

2006-10-01T19:03:19+00:00

Serial is _slow_ sometimes. So slow, that the NMI watchdog kicks in.

NMI Watchdog detected LOCKUP on CPU2CPU 2
Modules linked in: loop usb_storage md5 ipv6 parport_pc lp parport autofs4 i2c_dev i2c_core rfcomm l2cap bluetooth sunrpc pcdPid: 3138, comm: gpm Not tainted 2.6.11-1.1290_FC4smp
RIP: 0010:[] {serial_in+106}
RSP: 0018:ffff81003afc3d50  EFLAGS: 00000002
RAX: 0000000000000020 RBX: 0000000000000020 RCX: 0000000000000000
RDX: 00000000000003fd RSI: 0000000000000005 RDI: ffffffff804dcd60
RBP: 00000000000024fc R08: 000000000000000a R09: 0000000000000033
R10: ffff81001beb7c20 R11: 0000000000000020 R12: ffffffff804dcd60
R13: ffffffff804ade76 R14: 000000000000002b R15: 000000000000002c
FS:  00002aaaaaac4920(0000) GS:ffffffff804fca00(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00002aaaaabcb000 CR3: 000000003c0d0000 CR4: 00000000000006e0
Process gpm (pid: 3138, threadinfo ffff81003afc2000, task ffff81003eb63780)
Stack: ffffffff80275f2e 0000000000000000 ffffffff80448380 0000000000007d6b
       000000000000002c fffffffffffffbbf 0000000000000292 0000000000008000
       ffffffff80138e8c 0000000000007d97
Call Trace:{serial8250_console_write+270} {__call_console_drivers+76}
       {release_console_sem+315} {con_open+149}
       {tty_open+537} {chrdev_open+387}
       {dentry_open+260} {filp_open+68}
       {get_unused_fd+227} {sys_open+76}
       {tracesys+209}

Code: 0f b6 c0 c3 66 90 41 57 49 89 f7 41 56 41 be 00 01 00 00 41
console shuts up ...

I initially did the patch below a year ago for the Fedora kernel, and have
been keeping it up to date since.  I recently got the same thing happening
on a vanilla kernel, so figured it was time to repost this.

Signed-off-by: Dave Jones 
Acked-by: Alan Cox 
Signed-off-by: Andrew Morton 
Signed-off-by: Russell King

[SERIAL] OMAP1510 serial fix for 115200 baud

2006-10-01T16:07:06+00:00

The patch below is necessary for 115200 baud on an OMAP1510 internal UART.
It's been in the linux-omap tree for some time and with it applied to a
vanilla Linus git tree the serial console on the Amstrad Delta (which is
OMAP1510 based and whose initial bootloader runs at 115200) works fine (it
doesn't without it).

Signed-off-by: Jonathan McDowell 
Signed-off-by: Andrew Morton 
Signed-off-by: Russell King

[SERIAL] suspend/resume handlers don't have level arg anymore

2006-10-01T16:06:54+00:00

8250.c and serial_txx9.c port suspend/resume handler still have this obsolete
argument documented...

Signed-off-by: Sergei Shtylyov 
Signed-off-by: Russell King

[SERIAL] 8250 resourse management fixes

2006-10-01T16:06:49+00:00

I think register ranges obviously need to be claimed/released for all UARTs
including those with UPIO_MEM32 and UPIO_TSI iotype.

Also, serial8250_request_rsa_resources() returns false positives with
UPIO_MEM32, UPIO_AU, and UPIO_TSI iotype -- I don't think this makes any sense.

Signed-off-by: Sergei Shtylyov 
Signed-off-by: Russell King

[POWERPC] UPIO_TSI cleanup

2006-09-26T05:41:02+00:00

(le32_to_cpu(x) >> 8) & 0xff is a very odd way to spell (x >> 16) & 0xff,
even if that code is hit only on ppc.  The value is host-endian - we've
got it from readl(), after all...

Signed-off-by: Al Viro 
Signed-off-by: Paul Mackerras

[SERIAL] 8250: sysrq deadlock fix

2006-07-09T20:11:10+00:00

Fix http://bugzilla.kernel.org/show_bug.cgi?id=6716

Doing a sysrq over a serial line into an SMP machine presently deadlocks.

Signed-off-by: Andrew Morton 
Signed-off-by: Russell King

[SERIAL] 8250: add tsi108 serial support

2006-07-09T20:11:09+00:00

The following patch gets rid of CONFIG_TSI108_BRIDGE.  I add UPIO_TSI to
handle IIR and IER register in serial_in and serial_out.

(1) the reason to rewrite serial_in:

    TSI108 rev Z1 version ERRATA.  Reading the UART's Interrupt
    Identification Register (IIR) clears the Transmit Holding Register
    Empty (THRE) and Transmit buffer Empty (TEMP) interrupts even if they
    are not enabled in the Interrupt Enable Register (IER).  This leads to
    loss of the interrupts.  Interrupts are not cleared when reading UART
    registers as 32-bit word.

(2) the reason to rewrite serial_out:

    Check for UART_IER_UUE bit in the autoconfig routine.  This section
    of autoconfig is excluded for Tsi108/109 because bits 7 and 6 are
    reserved for internal use.  They are R/W bits.  In addition to
    incorrect identification, changing these bits (from 00) will make
    Tsi108/109 UART non-functional.

Signed-off-by: Roy Zang	
Signed-off-by: Andrew Morton 
Signed-off-by: Russell King

[PATCH] irq-flags: serial: Use the new IRQF_ constants

2006-07-02T20:58:53+00:00

Signed-off-by: Thomas Gleixner 
Cc: Ingo Molnar 
Cc: "David S. Miller" 
Cc: Benjamin Herrenschmidt 
Cc: Russell King 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds