summaryrefslogtreecommitdiff
path: root/include/linux
AgeCommit message (Collapse)AuthorFilesLines
2006-12-02PCI: switch pci_{enable,disable}_device() to be nestableInaky Perez-Gonzalez1-1/+2
Changes the pci_{enable,disable}_device() functions to work in a nested basis, so that eg, three calls to enable_device() require three calls to disable_device(). The reason for this is to simplify PCI drivers for multi-interface/capability devices. These are devices that cram more than one interface in a single function. A relevant example of that is the Wireless [USB] Host Controller Interface (similar to EHCI) [see http://www.intel.com/technology/comms/wusb/whci.htm]. In these kind of devices, multiple interfaces are accessed through a single bar and IRQ line. For that, the drivers map only the smallest area of the bar to access their register banks and use shared IRQ handlers. However, because the order at which those drivers load cannot be known ahead of time, the sequence in which the calls to pci_enable_device() and pci_disable_device() cannot be predicted. Thus: 1. driverA starts pci_enable_device() 2. driverB starts pci_enable_device() 3. driverA shutdown pci_disable_device() 4. driverB shutdown pci_disable_device() between steps 3 and 4, driver B would loose access to it's device, even if it didn't intend to. By using this modification, the device won't be disabled until all the callers to enable() have called disable(). This is implemented by replacing 'struct pci_dev->is_enabled' from a bitfield to an atomic use count. Each caller to enable increments it, each caller to disable decrements it. When the count increments from 0 to 1, __pci_enable_device() is called to actually enable the device. When it drops to zero, pci_disable_device() actually does the disabling. We keep the backend __pci_enable_device() for pci_default_resume() to use and also change the sysfs method implementation, so that userspace enabling/disabling the device doesn't disable it one time too much. Signed-off-by: Inaky Perez-Gonzalez <inaky@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-02Altix: Initial ACPI support - ROM shadowing.John Keller1-0/+1
Support a shadowed ROM when running with an ACPI capable PROM. Define a new dev.resource flag IORESOURCE_ROM_BIOS_COPY to describe the case of a BIOS shadowed ROM, which can then be used to avoid pci_map_rom() making an unneeded call to pci_enable_rom(). Signed-off-by: John Keller <jpk@sgi.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-02PCI: Make some MSI-X #defines genericMichael Ellerman1-0/+6
Move some MSI-X #defines into pci_regs.h so they can be used outside of drivers/pci. Signed-off-by: Michael Ellerman <michael@ellerman.id.au> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-02USB: keep count of unsuspended childrenAlan Stern1-0/+1
This patch (as818b) simplifies autosuspend processing by keeping track of the number of unsuspended children of each USB hub. This will permit us to avoid a good deal of unnecessary work all the time; we will no longer have to create a bunch of workqueue entries to carry out autosuspend requests, only to have them fail because one of the hub's children isn't suspended. The basic idea is simple. There already is a usage counter in the usb_device structure for preventing autosuspends. The patch just increments that counter for every unsuspended child. There's only one tricky part: When a device disconnects we need to remember whether it was suspended at the time (leave the counter alone) or not (decrement the counter). Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-02USB: struct usb_device: change flag to bitflagAlan Stern1-1/+1
This patch (as816) changes an existing flag in the usb_device structure to a bitflag, preparing the way for more bitflags to come in the future. Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-02USB: expand autosuspend/autoresume APIAlan Stern1-3/+26
This patch (as814) adds usb_autopm_set_interface() to the autosuspend API. It also provides convenient wrapper routines, usb_autopm_enable() and usb_autopm_disable(), for drivers that want to specify directly whether autosuspend should be allowed. Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-02USB: makes usb_endpoint_* functions inline.Luiz Fernando N. Capitulino1-11/+131
We have no benefits of having the usb_endpoint_* functions as functions, but making them inline saves text and data segment sizes: text data bss dec hex filename 14893634 3108770 1108840 19111244 1239d4c vmlinux.func 14893185 3108566 1108840 19110591 1239abf vmlinux.inline This is the result of a 2.6.19-rc3 kernel compiled with GCC 4.1.1 without CONFIG_MODULES, CONFIG_CC_OPTIMIZE_FOR_SIZE, CONFIG_REGPARM options set. USB support is fully enabled (while most of the other drivers are not), and that kernel has most of the USB code ported to use the endpoint functions. That happens because a call to those functions are expensive (in terms of bytes), while the function's size is smaller or have the same 'size' of the call. Signed-off-by: Luiz Fernando N. Capitulino <lcapitulino@mandriva.com.br> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-02usb/hub: allow hubs up to 31 childreninaky@linux.intel.com1-1/+6
Current Wireless USB host hardware (Intel i1480 for example) allows up to 22 devices to connect, thus bringing up the max number of children in the WUSB Host Controller to 22 'fake' ports. Upcoming hardware might raise that limit. Makes almost no difference to go to 31, as the bit arrays are byte-aligned (plus an extra bit in general), so 22 bits fit in 4 bytes as 31 do. As well, the only other array that depends on USB_MAXCHILDREN is 'struct usb_hub->indicator'. By declaring it 'u8' instead of 'enum hub_led_mode', we reduce the size of each entry from 4 bytes (in i386) to 1, which will add as we when are doubling USB_MAXCHILDREN (with 16 the size of that array is 64 bytes, with 31 would be 128; by using u8 that goes down to 31 bytes). Signed-off-by: Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-12-01mmc: Support for high speed SD cardsPierre Ossman2-1/+28
Modern SD cards support a clock speed of 50 MHz. Make sure we test for this capability and do the song and dance required to activate it. Activating high speed support actually modifies the TRAN_SPEED field of the CSD. But as the spec says that the cards must report 50 MHz, we might as well skip re-reading the CSD. Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
2006-12-01mmc: Add support for mmc v4 wide-bus modesPhilip Langdale1-0/+5
This change adds support for the mmc4 4-bit wide-bus mode. The mmc4 spec defines 8-bit and 4-bit transfer modes. As we do not support any 8-bit hardware, this patch only adds support for the 4-bit mode, but it can easily be built upon when the time comes. The 4-bit mode is electrically compatible with SD's 4-bit mode but the procedure for turning it on is different. This patch implements only the essential parts of the procedure as defined by the spec. Two additional steps are recommended but not compulsory. I am documenting them here so that there's a record. 1) A bus-test mechanism is implemented using dedicated mmc commands which allow for testing the functionality of the data bus at the electrical level. This is pretty paranoid and they way the commands work is not compatible with the mmc subsystem (they don't set valid CRC values). 2) MMC v4 cards can indicate they would like to draw more than the default amount of current in wide-bus modes. We currently will never switch the card into a higher draw mode. Supposedly, allowing the card to draw more current will let it perform better, but the specs seem to indicate that the card will function correctly without the mode change. Empirical testing supports this interpretation. Signed-off-by: Philip Langdale <philipl@overt.org> Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
2006-12-01[PATCH] mmc: Add support for mmc v4 high speed modePhilip Langdale2-3/+52
This adds support for the high-speed modes defined by mmc v4 (assuming the host controller is up to it). On a TI sdhci controller, it improves read speed from 1.3MBps to 2.3MBps. The TI controller can only go up to 24MHz, but everything helps. Another person has taken this basic patch and used it on a Nokia 770 to get a bigger boost because that controller can run at 48MHZ. Signed-off-by: Philip Langdale <philipl@overt.org> Signed-off-by: Pierre Ossman <drzeus@drzeus.cx>
2006-12-01Merge branch 'master' of ↵David Woodhouse43-171/+276
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6
2006-12-01[BLOCK] Cleanup unused variable passingJens Axboe1-2/+2
- ->init_queue() does not need the elevator passed in - ->put_request() is a hot path and need not have the queue passed in - cfq_update_io_seektime() does not need cfqd passed in Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2006-12-01[PATCH] block: support larger block pc requestsMike Christie1-3/+4
This patch modifies blk_rq_map/unmap_user() and the cdrom and scsi_ioctl.c users so that it supports requests larger than bio by chaining them together. Signed-off-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2006-12-01[PATCH] blktrace: add timestamp messageOlaf Kirch1-0/+12
This adds a new timestamp message to blktrace, giving the timeofday when we starting tracing. This helps user space correlate block trace events with eg an application strace. This requires a (compatible) update to blkparse. The changed blkparse is still able to process traces generated by older kernels, and older versions of blkparse should silently ignore the new records (because they have a pid of 0). Signed-off-by: Olaf Kirch <okir@suse.de> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2006-11-30[GFS2] Don't copy meta_header for rgrp in and outSteven Whitehouse1-2/+0
The meta_header for an ondisk rgrp never changes, so there is no point copying it in and back out to disk. Also there is no reason to keep a copy for each rgrp in memory. The code already checks to ensure that the header is correct before it calls the routine to copy the data in, so that we don't even need to check whether its correct on disk in the functions in ondisk.c Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Shrink gfs2_inode (7) - di_payload_formatSteven Whitehouse1-1/+0
This is almost never used. Its there for backward compatibility with GFS1. It doesn't need its own field since it can always be calculated from the inode mode & flags. This saves a bit more space in the gfs2_inode. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Shrink gfs2_inode (6) - di_atime/di_mtime/di_ctimeSteven Whitehouse1-3/+0
Remove the di_[amc]time fields and use inode->i_[amc]time fields instead. This saves 24 bytes from the gfs2_inode. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Shrink gfs2_inode (5) - di_nlinkSteven Whitehouse1-1/+0
Remove the di_nlink field in favour of inode->i_nlink and update the nlink handling to use the proper macros. This saves 4 bytes. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Shrink gfs2_inode (4) - di_uid/di_gidSteven Whitehouse1-2/+0
Remove duplicate di_uid/di_gid fields in favour of using inode->i_uid/inode->i_gid instead. This saves 8 bytes. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Shrink gfs2_inode (3) - di_modeSteven Whitehouse1-1/+0
This removes the duplicate di_mode field in favour of using the inode->i_mode field. This saves 4 bytes. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Shrink gfs2_inode (2) - di_major/di_minorSteven Whitehouse1-2/+0
This removes the device numbers from this structure by using inode->i_rdev instead. It also cleans up the code in gfs2_mknod. It results in shrinking the gfs2_inode by 8 bytes. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Shrink gfs2_inode (1) - di_header/di_numSteven Whitehouse1-5/+0
The metadata header doesn't need to be stored in the incore struct gfs2_inode since its constant, and this patch removes it. Also, there is already a field for the inode's number in the struct gfs2_inode, so we don't need one in struct gfs2_dinode_host as well. This saves 28 bytes of space in the struct gfs2_inode. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Change argument to gfs2_dinode_printSteven Whitehouse1-1/+1
Change argument for gfs2_dinode_print in order to prepare for removal of duplicate fields between struct inode and struct gfs2_dinode_host. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Move gfs2_dinode_in to inode.cSteven Whitehouse1-1/+1
gfs2_dinode_in() is only ever called from one place, so move it to that place (in inode.c) and make it static. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Change argument to gfs2_dinode_inSteven Whitehouse1-1/+1
This is a preliminary patch to enable the removal of fields in gfs2_dinode_host which are duplicated in struct inode. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] Change argument of gfs2_dinode_outSteven Whitehouse1-1/+2
Everywhere this was called, a struct gfs2_inode was available, but despite that, it was always called with a struct gfs2_dinode as an argument. By making this change it paves the way to start eliminating fields duplicated between the kernel's struct inode and the struct gfs2_dinode. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs2_quota_changeAl Viro1-1/+7
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs2_statfs_changeAl Viro1-2/+8
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs2_quotaAl Viro1-2/+7
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs2_inumAl Viro1-7/+12
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs_rindexAl Viro1-3/+11
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs2_meta_headerAl Viro1-4/+10
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs2_log_headAl Viro1-1/+11
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs2_inum_rangeAl Viro1-2/+7
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split and annotate gfs2_rgrpAl Viro1-2/+11
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] fields of gfs2_sb_host are host-endianAl Viro1-7/+4
... and several could be killed, but that's another story. Annotate scalar ones, kill completely unused. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split gfs2_sbAl Viro1-1/+21
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] gfs2_dinode_host fields are host-endianAl Viro1-27/+20
Annotated scalar fields, dropped unused ones. Note that it's not at all obvious that we want to convert all of them to host-endian... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[GFS2] split gfs2_dinode into on-disk and host variantsAl Viro1-3/+45
The latter is used as part of gfs2-private part of struct inode. It actually stores a lot of fields differently; for now the declaration is just cloned, inode field is swtiched and changes propagated. Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
2006-11-30[ARM] Remove EEPROM slave emulation from i2c-pxa driver.Russell King1-31/+0
The i2c-pxa driver should not contain EEPROM slave-mode emulation; this is something the platform should provide where required. Remove it. Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
2006-11-30Fix typos in doc and commentsJan Engelhardt1-2/+2
Changes persistant -> persistent. www.dictionary.com does not know persistant (with an A), but should it be one of those things you can spell in more than one correct way, let me know. Signed-off-by: Jan Engelhardt <jengelh@gmx.de> Signed-off-by: Adrian Bunk <bunk@stusta.de>
2006-11-30Fix misc .c/.h comment typosMatt LaPlante2-2/+2
Fix various .c/.h typos in comments (no code changes). Signed-off-by: Matt LaPlante <kernel1@cyberdogtech.com> Signed-off-by: Adrian Bunk <bunk@stusta.de>
2006-11-30Fix jiffies.h commentChase Venters1-1/+1
jiffies.h includes a comment informing that jiffies_64 must be read with the assistance of the xtime_lock seqlock. The comment text, however, calls jiffies_64 "not volatile", which should probably read "not atomic". Signed-off-by: Chase Venters <chase.venters@clientec.com> Signed-off-by: Adrian Bunk <bunk@stusta.de>
2006-11-30fix spelling error in include/linux/kernel.hJim Cromie1-1/+1
Signed-off-by: Adrian Bunk <bunk@stusta.de>
2006-11-30mqueue.h: don't include linux/types.hAlexey Dobriyan1-2/+0
This #include is not required. Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by: Adrian Bunk <bunk@stusta.de>
2006-11-30[MIPS] Add support for kexecNicolas Schichan1-0/+2
A tiny userland application loading the kernel and invoking kexec_load for mips is available here: http://chac.le-poulpe.net/~nico/kexec/kexec-2006-10-18.tar.gz Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2006-11-29[MTD] add get and put methodsArtem Bityutskiy1-0/+7
This patch adds get_device() and put_device() methods to the MTD description structure (struct mtd_info). These methods are called by MTD whenever the MTD device is get or put. They are needed when the underlying driver is something smarter then just flash chip driver, for example UBI. Signed-off-by: Artem Bityutskiy <dedekind@infradead.org>
2006-11-29[MTD] add get_mtd_device_nm() functionArtem Bityutskiy1-0/+1
This patch adds one more function to the MTD interface to make it possible to open MTD devices by their names, not only numbers. This is very handy in many situations. Also, MTD device number depend on load order and may vary, while names are fixed. Signed-off-by: Artem Bityutskiy <dedekind@infradead.org>
2006-11-29[MTD] NAND: add subpage write supportThomas Gleixner2-0/+11
Many SLC NANDs support up to 4 writes at one NAND page. Add support of this feature. Signed-off-by: Artem Bityutskiy <dedekind@infradead.org>