- 08 Dec, 2006 40 commits
-
-
Jiri Slaby authored
Mxser tty driver name should be ttyMI, not ttyM. Correct this in mxser_new to avoid conflicts with isicom driver, which is ttyM. Acked-by:
Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by:
Jiri Slaby <jirislaby@gmail.com> Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
Linus Torvalds <torvalds@osdl.org>
-
Jiri Slaby authored
I can maintain moxa and isicom char drivers, because I've rewritten them to the new API. Signed-off-by:
-
Jiri Slaby authored
Celan redundant macros. Signed-off-by:
-
Jiri Slaby authored
Finally, the intention of the patch serie: PCI probing for this driver. add pci_driver structure, (un)register it and add some stuff, which this needs. Remove pdev pointer from board structure, because it's no longer needed. Signed-off-by:
-
Jiri Slaby authored
- Driver uses global tty_struct array, which tries to assign to the tty_driver's ttys. - the very same thing for termios - the same for termios_locked Kill such constructions. Signed-off-by:
-
Jiri Slaby authored
Prototype of the driver's interrupt handler is old-fashioned, past changes tell us, we won't get pt_regs as the 3rd argument. There are only 2 params. Signed-off-by:
-
Jiri Slaby authored
mxser_new, comments cleanup - Remove commented code, since we have version control. - Remove comments containing "following added by..." and "above added by,..". It's useless. - Align other comments. Signed-off-by:
-
Jiri Slaby authored
This patch was intorduced in mxser to 1.9.1 conversion, but causes endless sleep in tcdrain userspace call. Thanks to Sergei Organov <osv@javad.com> for testing and pointing this out. Signed-off-by:
-
Jiri Slaby authored
Too many structures are in the driver for each type of card. Make only one for all cards and appropriate alter the code. Signed-off-by:
-
Jiri Slaby authored
Register tty indexes only for real devices, udev then creates nodes for them (and only for them). Move tty_register_driver before probing, to be correct when calling tty_register_device. Also tell tty layer by tty_driver flags, that we are registering devices. Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by:
-
Jiri Slaby authored
ISA cards finding was too complex -- 2 (module params + predefined) absolutely same routines, join them together with one for loop, one if and one indent level. Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by:
-
Jiri Slaby authored
Like other drivers silently unregister_tty_driver and put_tty_driver. It shouldn't be busy when module release function is called, since we are not bsd, no refs shouldn't be held. Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by:
-
Jiri Slaby authored
Resources were not released in some fail paths. Correct this behaviour by implementing function and calling it when something fails. Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by:
-
Jiri Slaby authored
Use testbit like in tty subsystem for TTY_IO_ERROR testing. Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by:
-
Jiri Slaby authored
Use tty_ldisc_flush and tty_wakeup helpers for accessing ldisc internals. Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Signed-off-by:
-
Jiri Slaby authored
No need to redefine termbits in char tree. Signed-off-by:
-
Jiri Slaby authored
There is no need to have another (ifndeffed) definition of CMSPAR. It's defined in includes. Signed-off-by:
-
Jiri Slaby authored
Reorder functions upside down not to have too many prototypes of each function and have some order (similar to other drivers). Signed-off-by:
-
Jiri Slaby authored
- We don't need useless paragraphs in license terms. - Add me to copyright. Signed-off-by:
-
Jiri Slaby authored
Print some debug info only when DEBUG is enabled (use pr_debug macro). Eliminate verbose kernel parameter which takes care of this. Signed-off-by:
-
Jiri Slaby authored
Remove printk with info we are looking for a tester. Signed-off-by:
-
Jiri Slaby authored
Some spinlock changes were introduced in 1.9.1 original moxa driver. Revert them, since they cause machine not responding. Signed-off-by:
-
Jiri Slaby authored
mxser_new, kill unneeded memsets There is no need to re-zero static global variables' memory, hence memsets doing this are useless. alloc_tty_struct also zeroes allocated memory: another candidate for removing. This fixes also a bug -- global structures are cleaned up after initialization of some its parts. Signed-off-by:
-
Jiri Slaby authored
mxser_new, check request_region retvals Return values of (pci_)request_region should be checked and error should be returned if something is in bad state. Signed-off-by:
-
Jiri Slaby authored
mxser_new, pci_request_region for pci regions Use pci_request_region instead of standard request_region for pci device regions. More checking, simplier use. Signed-off-by:
-
Jiri Slaby authored
mxser_new, use __(dev)init macros Let kernel know what can be freed after init of the driver. Signed-off-by:
-
Jiri Slaby authored
This patch is preparation for further patches (pci probing) to allow allocated structures to be private data in pci_dev structure. Union two different structures used in the driver (hw_conf and port/board descriptor) to another 2: port and board not to initialize 2 different things and to have ports contained in board structure. Signed-off-by:
-
Jiri Slaby authored
Change cloned experimental driver according to original 1.9.1 moxa driver. Some int->ulong conversions, outb ~UART_IER_THRI constant. Remove commented stuff. I also added printk line with info, if somebody wants to test it, he may contact me as I can potentially debug the driver with him or just to confirm it works properly. Signed-off-by:
-
Jiri Slaby authored
include mxser_new.h instead of original mxser.h Signed-off-by:
-
Andrew Morton authored
Signed-off-by:
-
Jiri Slaby authored
Clone a new driver for moxa smartio devices by copying mxser.c to mxser_new.c and mxser.h to mxser_new.h. No other changes are made. This is for purposes of updating the driver to the latest vendor version. Signed-off-by:
-
Alan Cox authored
Switch to pci ref counts for mxser when handling PCI devices. Use pci_get_device and drop the reference when we finish and unload. Signed-off-by:
Alan Cox <alan@redhat.com> Cc: Greg KH <greg@kroah.com> Cc: <jirislaby@gmail.com> Signed-off-by:
-
Jiri Slaby authored
Mxser tty driver name should be ttyMI, not ttyM. Correct this in mxser to avoid conflicts with isicom driver, which is ttyM. Acked-by:
-
Haavard Skinnemoen authored
Convert SH64 to use generic ioremap_page_range() Signed-off-by:
Haavard Skinnemoen <hskinnemoen@atmel.com> Cc: Paul Mundt <lethal@linux-sh.org> Signed-off-by:
-
Haavard Skinnemoen authored
Convert SH to use generic ioremap_page_range() Signed-off-by:
Paul Mundt <lethal@linux-sh.org> Signed-off-by:
-
Haavard Skinnemoen authored
Convert S390 to use generic ioremap_page_range() Signed-off-by:
-
Haavard Skinnemoen authored
Convert parisc to use generic ioremap_page_range() Signed-off-by:
Kyle McMartin <kyle@parisc-linux.org> Signed-off-by:
-
Haavard Skinnemoen authored
Convert MIPS to use generic ioremap_page_range() [yoichi_yuasa@tripeaks.co.jp: build fix] Signed-off-by:
-
Oleg Nesterov authored
All members of the process group have the same sid and it can't be == 0. NOTE: this code (and a similar one in sys_setpgid) was needed because it was possibe to have ->session == 0. It's not possible any longer since [PATCH] pidhash: don't use zero pids Commit: c7c64641Signed-off-by:
Oleg Nesterov <oleg@tv-sign.ru> Signed-off-by:
-
Oleg Nesterov authored
All tasks in the process group have the same sid, we don't need to iterate them all to check that the caller of sys_setpgid() doesn't change its session. Signed-off-by:
-
