Summary of changes from v2.5.20 to v2.5.21 ============================================ NTFS: Remove unused source file. net/ipv4/af_inet.c - remove spurious spaces and tabs at end of lines - make sure if, while, for, switch has a space before the opening '(' - make sure no line has more than 80 chars - move initializations to the declaration line where possible - bitwise, logical and arithmetic operators have spaces before and after, improving readability of complex expressions - remove uneeded () in returns - other minor cleanups net/ipv4/devinet.c - remove spurious spaces and tabs at end of lines - make sure if, while, for, switch has a space before the opening '(' - make sure no line has more than 80 chars - move initializations to the declaration line where possible - bitwise, logical and arithmetic operators have spaces before and after, improving readability of complex expressions - remove uneeded () in returns - use member initializations - other minor cleanups net/ipv4/icmp.c - remove spurious spaces and tabs at end of lines - make sure if, while, for, switch has a space before the opening '(' - make sure no line has more than 80 chars - move initializations to the declaration line where possible - bitwise, logical and arithmetic operators have spaces before and after, improving readability of complex expressions - remove uneeded () in returns - use member initializations - other minor cleanups net/core/dev.c - remove spurious spaces and tabs at end of lines - make sure if, while, for, switch has a space before the opening '(' - make sure no line has more than 80 chars - move initializations to the declaration line where possible - bitwise, logical and arithmetic operators have spaces before and after, improving readability of complex expressions - remove uneeded () in returns - other minor cleanups Before: text data bss dec hex filename 15228 12376 2560 30164 75d4 net/core/dev.o After: text data bss dec hex filename 14840 12364 2624 29828 7484 net/core/dev.o net/core/dev.c: Do not cast pointers to int, use long. net/sched fixes: - include/net/pkt_sched.h: wrong detection of 32bit archs, Martin Devera - net/sched/sch_api.c: jiffies wrap bug, Martin Devera ipv6 raw fixes: - RFC2292 bis compliance: enable checksumming of ICMPv6 raw sockets by default, reject odd offsets - calculate checksum correctly when user forgets to uninitialize checksum word net/ipv4/tcp_ipv4.c - remove spurious spaces and tabs at end of lines - make sure if, while, for, switch has a space before the opening '(' - make sure no line has more than 80 chars - move initializations to the declaration line where possible - bitwise, logical and arithmetic operators have spaces before and after, improving readability of complex expressions - remove uneeded () in returns - use named member initializations in structs - other minor cleanups Sizes: Before: text data bss dec hex filename 19408 2560 8 21976 55d8 net/ipv4/tcp_ipv4.o After: text data bss dec hex filename 19312 2560 12 21884 557c net/ipv4/tcp_ipv4.o sk->num no longer exists in 2.5, use inet_sk(sk)->num NTFS: The beginning of 2.0.8. - Major updates for handling of case sensitivity. - Remove unused source file fs/ntfs/attraops.c. - Remove show_inodes mount option(s), thus dropping support for displaying of short file names. - Remove deprecated mount option posix. - Restore show_sys_files mount option. - Add new mount option case_sensitive, to determine if the driver treats file names as case sensitive or not. If case sensitive, create file names in the POSIX namespace. Otherwise create file names in the WIN32 namespace. By default, or when case_sensitive is set to FALSE, files remain accessible via their short file name, if it exists. - Add additional argument to ntfs_lookup_inode_by_name() in which we return information about the matching file name if the case is not matching or the match is a short file name. See comments above the function definition for details. NTFS: Fix really dumb logic bug in boot sector recovery. NTFS: Fix potential 1 byte overflow in fs/ntfs/unistr.c::ntfs_ucstonls(). Also, minor updates/fixes to docs and comments. NTFS: 2.0.8 release. Major updates for dcache aliasing issues wrt short/long file names. In order to handle the case insensitivity issues of NTFS with regards to the dcache and the dcache requiring only one dentry per directory, we deal with dentry aliases that only differ in case in ->ntfs_lookup() while maintining a case sensitive dcache. This means that we get the full benefit of dcache speed when the file/directory is looked up with the same case as returned by ->ntfs_readdir() but that a lookup for any other case (or for the short file name) will not find anything in dcache and will enter ->ntfs_lookup() instead, where we search the directory for a fully matching file name (including case) and if that is not found, we search for a file name that matches with different case and if that has non-POSIX semantics we return that. We actually do only one search (case sensitive) and keep tabs on whether we have found a case insensitive match in the process. To simplify matters for us, we do not treat the short vs long filenames as two hard links but instead if the lookup matches a short filename, we return the dentry for the corresponding long filename instead. In ->ntfs_lookup() we distinguish three cases: 1) @dent perfectly matches (i.e. including case) a directory entry with a file name in the WIN32 or POSIX namespaces. In this case ntfs_lookup_inode_by_name() will return with name set to NULL and we just d_add() @dent. 2) @dent matches (not including case) a directory entry with a file name in the WIN32 namespace. In this case ntfs_lookup_inode_by_name() will return with name set to point to a kmalloc()ed ntfs_name structure containing the properly cased little endian Unicode name. We convert the name to the current NLS code page, search if a dentry with this name already exists and if so return that instead of @dent. The VFS will then destroy the old @dent and use the one we returned. If a dentry is not found, we allocate a new one, d_add() it, and return it as above. 3) @dent matches either perfectly or not (i.e. we don't care about case) a directory entry with a file name in the DOS namespace. In this case ntfs_lookup_inode_by_name() will return with name set to point to a kmalloc()ed ntfs_name structure containing the mft reference (cpu endian) of the inode. We use the mft reference to read the inode and to find the file name in the WIN32 namespace corresponding to the matched short file name. We then convert the name to the current NLS code page, and proceed searching for a dentry with this name, etc, as in case 2), above. Reversed a mistake in cyber200fb.c Finshed porting the 3Dfx driver over to the new api. Ported over NeoMagic over to new api. Fixed a nasty bug in the 3Dfx driver and added the ahrdware routines for the NeoMagic chipset. Bug fixes for NeoMagic and 3Dfx driver More bug fixes. When will it end? Ported over the apollo framebuffer device. Updated the Permedia 2 fbdev driver to the latest code. Small bug fix for the NeoMagic driver. . arch/* drivers/cdrom/* drivers/char/* Fix some copy_{to,from}_user and {put,get}_user error handling, get rid of some verify_area, copy_{to,from}_user already checks for errors. Ported the Maxine framebuffer driver to the new api. Split up "iput()" and make it more readable. Add "drop_inode" VFS interface to make FS operations cleaner and race-free. Remove old force_delete interface, and update filesystems that used it to use the new infrastructure. [PATCH] 2.5.20 airo wireless - "I can't get no, compilation..." Fix namespace clash with proc stuff an compilation warnings. [PATCH] ]PATCH] 2.5.20 IDE 83 - Remove last parameter from ide_dump_status. This information is now permanently present in device->staus field, so there is not need to pass it around. - Patch for DVD read through ide-scsi. There is the possibility that we can get request structures passed down, which don't have the queue field set. At lest on the BIO code path this seems to be something worth further investigation. Found by Adam J. Richter. (Jens?) - Revert my change to the hostdata handling. I did get it wrong about the way host structures are allocated by the generic SCSI layer. It plays tricks there. - piix driver updates by Vojtech Pavlik. - We have a ata_out_regfile, so we should have ata_in_regfile too. [PATCH] remove wq_lock_t cruft This patch removes the whole wq_lock_t abstraction, forcing the behavior to be that of a standard spinlock and changes all the wq_lock code in the tree appropriately. Removes lots of code - always a Good Thing to me. New behavior is same as previous behavior (USE_RW_WAIT_QUEUE_SPINLOCK unset). [PATCH] capability.c cleanup I started looking into a couple FIXMEs in kernel/capability.c and I ended up with a fairly largish patch (although not quite so many changes to object code). First, it is unsafe to touch task->cap_* while not holding task_capability_lock. The most notable occurrence of this is sys_access which saves the current cap_* values, changes them, does its business, then restores them. In between all this they can change and then be restored to old values. Unfortunately we cannot just grab the lock here since the function can sleep - I marked this with a FIXME for now. Second, I formalized the locking rules with task_capability_lock. I declared the lock in include/linux/capability.h so other code can grab it. Finally, there is a whole boatload of code cleanup: - remove conditional locking/unlocking - that is just gross - don't pointlessly grab the read_lock twice - add/remove/edit comments - change some types (int -> pid_t, etc) - static inline two small functions that are called only once each - remove two FIXMEs - general code cleanup for readability and performance TODO: - fix sys_access and other cap_* accesses - do something about the annoying oddball 5-space indentation in kernel/capability.c !! Patch is against 2.5.20, please apply. Robert Love [PATCH] trivial misc. scheduler cleanups Resend of trivial bits from my scheduler tree...: - shift cpu by 1UL not 1 to match type - clarify various comments - remove the barrier from preempt_schedule. This was here because I used to check need_resched before returning from preempt_schedule but we do not now (although should). The barrier insured need_resched and preempt_count were in sync now and after an interrupt that could occur. [PATCH] maintainers update Might as well make it explicit... Patch is against 2.5.20, please apply. Robert Love [PATCH] make smp.c preempt-safe The attached patch cleans up some per-CPU code in arch/i386/kernel/smp.c that could be problematic under preemption. The first I solve with the new get_cpu interface, for the second two I explicitly disable preemption. I also changed 1 to 1UL in the shift to properly match the type. [PATCH] sys_sysinfo cleanup Looks like sys_sysinfo has not been touched in years. Among other things, it uses a global cli() for protection; I switched it to an existing rwlock. I also pulled it out of info.c and stuck it in timer.c (I choose timer.c because it shares dependencies there already). The details: - move sys_sysinfo to kernel/timer.c from kernel/info.c: why one small syscall got its own file is beyond me. - delete kernel/info.c - stop the global cli! now grab a read_lock on xtime_lock. this is safe as we moved the write_unlock on xtime_lock down one line to cover the calculating of avenrun. - trivial code cleanup [PATCH] ipchains_core netlink fix oom-loop fixes error handling after a netlink failure - it does not do a cleanup and it makes every next call to ip_fw_check to detect a loop and drop the packet. [PATCH] Fix for recent swap changes on 64 bit archs This fixes two shift warnings on 64 bit archs. [PATCH] unplugging fix Fix queue plug locking. [PATCH] Fix Oops due to use of incorrect km_type in RPC socket code... The following has been vetted with davem w.r.t. the change to KM_SKB_DATA. Apologies for the bug... include/asm-*/kmap_types.h: Replace the unused km_type slot KM_SKB_DATA with KM_SKB_SUNRPC_DATA. net/sunrpc/xdr.c: Replace the use of KM_USER0 with KM_SKB_SUNRPC_DATA for copying data from an skb into the page cache when in the sk->data_ready() callback. [PATCH] Missing include in mm/bootmem.c Missing include of asm_io.h in mm_bootmem.c: Submitted this trivial patch on May 14, but nothing has happened yet. Perhaps better chance if you took care of it. It's needed by phys_to_virt() but it happens to work on i386 etc. since dma.h includes io.h for that arch. [PATCH] ipchains_core GFP_KERNEL fix [ Made applicable with -p1 by Rusty ] nlma fixes a call to netlink_broadcast with GFP_KERNEL ( passed to skb_clone ) while we are in_interrupt() ( catched by a BUG() in slab.c:1109 ). [PATCH] "General options" - begone There is more of this janitorial work to come (eg General setup). I'm just piecemealing it [PATCH] update to pci_quirks.c Attached is a really trivial patch that changes an "unknown" PCI_DEVICE_ID 0x3112 to PCI_DEVICE_ID_VIA_8361 in drivers/pci/quirks.c . It applies cleanly to 2.5.18 and 2.4.19-pre8. [PATCH] fix NULL dereferencing in dcache.c Unrelated to my first dcache patch, this is something more crucial and should be applied first. fs/dcache.c: - handle d_alloc() returning NULL. [PATCH] fix for /proc operation: I found that 'max' pointer is not updated in proc_dointvec_minmax() and proc_doulongvec_minmax(), when I write smaller values than min to /proc/sys entry (and val<*min++ check becomes true.) This may lead to min/max checking of values with bogus maximum. [PATCH] bogus casts in ide-cd.c: This patch against 2.5.19 gets rid of some bogus casts in ide-cd.c In my opinion the casts as is are bugs waiting to happen. [PATCH] Spelling Rusty Russell : Oh, I feel so violated...: [PATCH] TAGS creation should go into arch dirs Playing with arch stuff does this to you. [PATCH] correct zone_table comment The comment describing the usage of zone_table[] assumes the existence of an unsigned char page->zone field from the original implementation of page->zone size reduction. This patch corrects the comment to accurately describe the lookup mechanism used by page_zone() and also to mention explicitly the sole user of the table, page_zone(). [PATCH] duplicate decl in sched_init() I found this one while trying to straighten out bootstrap ordering issues elsewhere. There appears to be a duplicate declaration of rq in sched_init(). This removes the nested declaration and otherwise leaves things alone. [PATCH] make memclass() an inline memclass is too large to be a #define; it overflows 80 columns and does not make use of facilities available only to macros. This patch convert memclass() to be an inline function. [PATCH] remove antiquated comment This comment, describing how to optimize for gcc-2.2.2, is so outdated it should be removed. It's also quite doubtful it should ever have been placed in this file at all (perhaps something under Documentation/ ?). This patch removes it. [PATCH] remove macros from page_alloc.c The memlist_* macros serve as nothing but an insulation layer from the Linux-native generic list operations. This patch removes them in favor of using generic list operations directly. [PATCH] static list init page_alloc.c inactive_list and active_list are global, yet they are repeatedly initialized using INIT_LIST_HEAD() in free_area_init_core(). This patch is originally due to Christoph Hellwig, and by some reports has been implementated before in 2.4-based trees by Andrea Arcangeli. [PATCH] remove space in cache names Most /proc/slabinfo cache_names are in the format: cache_name. There are a couple with spaces in the name, which is inconsistent and requires a special case when scripting. Changes "fasync cache" and "file lock cache" to have the usual underscore. [PATCH] must be __KERNEL__ for byteorder/generic.h Here's that patch again (MIME this time, so tabs don't get lost by my silly gui mailer); applies cleanly against against 2.4.19-pre8. Nobody commented on it last time I posted it, and it does make compiling gcc easier, so I guess that makes it trivial patch monkey fodder. Or am I making a silly mistake? [PATCH] bluesmoke merge This patch merges in all the currently outstanding bluesmoke bits from 2.5-dj to 2.5.20, it also has the pleasant side effect of fixing the compilation. Test compiled with and without MCE. [PATCH] Fix suspend-to-RAM in 2.5.20 I created arch/i386/suspend.c not to clash with ACPI people so much in future. (More stuff is going to move into it in the future, to clean up functions that really do not belong to the headers.) [PATCH] Cleanup swsusp in 2.5.20 This cleans up swsusp in 2.5.20. Killed sysrq-D support (it is too much trouble to support suspending from interrupt), kill unused define, fix compile-time warnings (thanks to Adam). [PATCH] Re: Fix suspend-to-RAM in 2.5.20 Here's followup patch that makes it work. Notice freeze_processes() -- if you don't do that you risk data corruption. A few small fixes from porting it over to the new api. ALso a few more drivers from the MIPS platform ported over. device model udpate: - make sure drv->devices is initialized on registration (from Peter Osterlund) - add remove_driver for forcing removal of driver There was a potential race with the module unload code. When a pci driver was unloaded, it would call pci_unregister_driver, which would simply call put_driver. If the driver's refcount wasn't 0, it wouldn't unbind it from devices, but the module unload would still continue. If something tried to access the driver later (since everyone thinks its still there), Bad Things would happen. This fixes it until there can be tighter integration between the device model and module unload code. Do manual traversing of drivers' devices list when unbinding the driver. driver_unbind was called when drv->refcount == 0. It would call driver_for_each_dev to do the unbinding The first thing that would do was get_device, which... BUG()'d if drv->refcount == 0. Duh. PCI driver mgmt: - Make sure proper pci id is passed to probe() - make sure pci_dev->driver is set and reset on driver registration/unregistration - call remove_driver to force unload of driver on unregistration net/core/dev.c: Make handle_diverter return 0 Fix extra parenthesis in the constant-address rwlock case. Noted by Alexey Vyskubov [PATCH] fix 2.5.20 ramdisk 2.5.20 seems to be incapable of executing binaries in a ramdisk-based root filesystem. The ramdisk in question is an ext2fs, with a 1K block size loaded via the compressed ramdisk loader in do_mounts(). It appears that, in the case of a 1K block sized filesystem, we attempt to read two 512-byte sectors into a BIO vector. The first one is copied into the first 512 bytes. The second sector, however, is copied over the first 512 bytes. Obviously not what we really want. Rev. 2, slightly cleaned up: [PATCH] 2.5.20 IDE 84 - Simplify ide_cmd_type_parse by removing the handling of commands which we never use. - Realize that pre_task_out_intr and pre_task_mulout_intr are semanticaly identical. Use only pre_task_out_intr(). This allowed us to eliminate the prehandler altogether. - Updated fix for misconfigured host chips by Vojtech Pavlik. - Be more permissive about ioctl handling to allow device type drivers to do they own checks. - ali14xx cleanups by Andrej Panin. - Unfold usage ide_cmd_type_parser in tcq.c code. This makes this operation local to ide-disk.c. Move it as well as the interrupt handlers used only for the handling of disk requests there too. - Guard against calling handler before the drive is ready for it in ata_taskfile()! Well this bug was there before, but right now we inform about it. - Unfold ide_cmd_type_praser in ide-disk.c. Merge the remaining bits of it with get_command. Well it's no more. - Move recal_intr to ide.c - the only place where it's used. This doesn't change the "mechanics" of the code but it makes it a lot more "obvious" what's going on. Ported over the TX3912 framebuffer to the new api. Fixed a NODEV error for the Permedia framebuffer driver and patched fbcmap.c to prevent another Oops. Bonding driver: Merge 2.4.x driver updates into 2.5. [PATCH] remove suser() Attached patch replaces the lone remaining suser() call with capable() and then removes suser() itself in a triumphant celebration of the glory of capable(). Or something. ;-) Small cleanup of capable() and some comments, too. [PATCH] remove fsuser() This patch removes fsuser(). Now both suser() and fsuser() are gone and all permission checks use an appropriate capable() call. Change unused_initcall to postcore_initcall for things that must be initialized early, but not too early (after the core is done) s/subsys_initcall/postcore_initcall/ for sys_bys and pci, so they get initialized early enough for arch and subsys initcalls to use them. Bug fixes for the fbdev layer. [ARM] Makefile and page.h ARM updates for 2.5.20 [ARM] Clean up map_desc structure - remove LAST_DESC in favour of passing the array size to iotable_init - replace domain and permission bits/cache attributes with a generic "type", which can be easily converted to the required domain and permission bits/cache attributes at run time. It also removes the possibility for getting such things wrong and (accidentally) allowing all user space to fiddle with devices directly. [ARM] Cast thread_saved_{pc,fp} to unsigned long. The kernel expects these to return values in unsigned long types. [ARM PATCH] 1166/1: further cleanup of SA1110 suspend/resume code (2.5) Same thing as patch #1165/1 but for 2.5.x kbuild: Split Makefile into needs / needs not .config The current top-level Makefile has a fundamental problem which makes "make oldconfig vmlinux" impossible: It includes .config, which is changed by "oldconfig". So after "oldconfig" .config has changed and the .config the Makefile had read is obsolete. make provides a mechanism to cope with this, it'll restart automatically if any of the files it included changed, if you let it know that you changed it, just using a normal rule which has .config as its target. However, once you tell make that "make oldconfig" changes .config, you have another problem: oldconfig always uses .config to be remade, there's no mechanism to tell if it's up to date. So makes notices that .config has changed, restarts, makes oldconfig again, notices that .config has changed, restarts, ... you get the picture. The way to solve this is to do a proper two-stage approach: If you just say "make oldconfig", there's no need for the Makefile to even read the .config. If it does not, it won't restart and recurse infintely. So we divide the Makefile into two sections: One for targets which don't need the variables from .config, like *config, clean, mrproper and one section which does the actual build, which needs to know the CONFIG_ options. If one of the "noconfig" targets is given, we handle those, without reading .config. From there, we call make again, filtering out the already handled targets, to do the main work. The fact that this actually works correctly can be seen by trying "make vmlinux oldconfig" which will execute things in the right order - and this is not just nitpicking, it means that "-j" will get this case right, too. The $(CONFIGURATION) hack used to start "make config" automatically can go away now, too. Since we don't know which of make *config the user prefers, we'll just ask him call "make whatever-config" himself, instead of forcing "make config" on him. [PATCH] 2.5.20 IDE 85 - Work a bit on the automatic CRC error recovery handling. System still hangs. But one thing for sure - we don't have to use any specialized irq handler for it. - Since ioctl don't any longer submit requests to the queue without rq->special set, we can safely remove args_error handling from start_request. - Make REQ_SPECIAL usage in ide-floppy obvious. - Use REQ_SPECIAL everywhere instead of REQ_DRIVE_ACB. This is actually a bit dangerous but should work as far as I can see. - Unfold the now not REQ_SPECIAL specific dequeing part of ide_end_drive_cmd(). Turns out that we can thereafter remove the calls to ide_end_drice_cmd() at places where the request type isn't REQ_SPECIAL all any longer. (tcq.c) - After the above operation it turns out that there are just two places where ide_end_drive_cmd remains, namely: ata_error, task_no_data_intr drive_cmd_intr. We can avoid it by changing the logics in ata_error a slightly. So now just to cases remain where ide_end_drive_cmd remains used: drive_cmd_intr and task_no_data_intr. - Now looking a bit closer we can realize that drive_cmd_intr and task_no_data_intr can be easly merged together, since the usage of task_no_data_intr implied taskfile.sector_number == 0. - Use one single ata_special_intr function for the handling of interrupts submitted through ide_raw_cmd. - Move the remaining artefacts of ide_end_drive_cmd directly to ata_special_intr. Oh we don't need to check for REQ_SPECIAL any longer there, since the context is already known. - Set the ata_special_intr handler and command type directly inside ide_raw_taskfile to save code. [PATCH] fs/inode.c list_del_init A few cases of list_del(x) + INIT_LIST_HEAD(x) crept in recently which can be replaced with list_del_init(x). [PATCH] fs/locks.c use list_del_init Trivial part of a patch by Matthew Wilcox [PATCH] fcntl_[sg]etlk() only need the file * Another simple part of a Matthew Wilcox patch. We do not need to pass the file descriptor to the fcntl_[sg]etlk functions, only the struct file * which we have already got from the file descriptor and verified. [PATCH] Allow mpage.c to build When trying to build mpage.c for ARM, I get errors from bio.h since kdev_t isn't defined. The following fixes this. (I fail to see how this can build for anyone as it currently stands; its probably something x86 specific buried in the asm-i386 includes.) [PATCH] Cleanup i386 abuses The following patch cleans up the i386 usage of . This remove from which did not need it, which only had it due to an extern using __init, which is not needed. This adds to which actually has numerous __init functions and adds to 9 files inside of arch/i386 which were indirectly including previously. kbuild: Fix make -s (silent) and add a quiet mode Suppress echoing of commands when using "make -s", so that make -s does indeed have the effect one would expect. Add a quiet mode, which will print not the entire command but only one line per rule. To turn it on, use make KBUILD_VERBOSE=0 vmlinux/whatever or set KBUILD_VERBOSE=0 in your environment. For now, the verbose mode is default, which gives you the old behavior of printing all commands. The output in quiet mode is based on what Keith Owens' kbuild-2.5 does, I like, I did not want to invent yet another output format. kbuild: Fix 'make some/dir/foo.lst' Just use 'make some/dir/foo.lst' to produce mixed source code and assembly for debugging. (If the object gets linked in and you have a System.map, it'll relocate appropriately) Apart from the needed Makefile bits, also clean up the script "makelst". kbuild: Fix calling of scripts We source some scripts, but still pass parameters to them, e.g. . mk_version_h $@ $(KERNELRELEASE) $(VERSION) ... This does not work for all kinds of /bin/sh (it does for bash, that's why I did not notice). The fix is easy: Just mark the scripts executable and call instead of source them. Unfortunately, patch(1) doesn't understand about propagating chmod. bk does, so changing the tree isn't hard, and we introduce an explicit chmod a+x executed during the build for propagating this change into those trees which get "traditionally" patched up. kbuild: Make dependencies at compile time Making dependencies once up front is not ideal. For one, you don't need them initially, since when you don't have the .o file, you bet you have to build it no matter what the dependencies say - dependencies are about deciding when to *re*build. There's more reasons, like: o you don't even know which files you'll build, so you have to go over all files (even over drivers/{sbus,s390,...} on i386) o generated files don't exist yet, so you cannot pick up dependencies on them o even if dependencies are right initially, they change when you work on your tree or patch it, and nobody will notice unless you run "make dep" explicitly again Anyway, gcc knows hows to emit a correct dependency list, so we just use that. Well, a little bit of hacking is necessary to remove the dependency on autoconf.h and put in individual CONFIG_WHAT_EVER dependencies instead, since otherwise changing one config option would cause everything to be rebuilt. I should add that I didn't come up with this all by myself, most work is actually done in gcc and there were discussions about using -MD on kbuild-devel way back, so I should mention Keith Owens and Michael Elizabeth Chastain, and probably others that I forgot, so I apologize just in case. kbuild: Use deep directory structure for include/linux/modules We used to force the obvious deep structure of all objects which export symbols into a flat list in include/linux/modules. This initially caused the restriction the no two exporting objects could have the same name (Ever wondered why there's ksyms.c and i386_ksyms.c?) With the ALSA merge this restriction was mostly lifted by some hack, but some cases still don't work right (Hi XFS). As it's much cleaner to just use a normal tree under include/linux/modules, reflecting the source tree, we now do just that. kbuild: Clean up descending into subdirs Rules.make now has three targets: o default (a.k.a first_rule): The actual build. Deciding whether to build built-in or modular or both is decided by $(KBUILD_MODULES) and $(KBUILD_BUILTIN) now, instead of using different targets o fastdep: doesn't actually dependencies anymore, only generates modversions o modules_install: Well, you guess what that does. Cleaned up descending, and no more differentiating between $(subdir-y) and $(subdir-m). That means $(mod-subdirs) can go away now. kbuild: Really stick with verbose output by default If KBUILD_VERBOSE is not set in the environment, have it default to 1 (i.e. on) kbuild: modversions improvements If an object was changed to not export symbols anymore, the corresponding stale .ver file would have been left lying around and been picked up when generating modversions.h. The obvious solution to remove include/linux/modules/* at the beginning of "make dep" is not really good, since that means that .ver files would be regenerated unconditionally, thus causing a lot of possibly unnecessary rebuilds. So, instead, we build a temporary shadow tree of all export-objs (as empty files) during the recursive "make fastdep" phase, and use that to generate modversions.h. Ensure that we touch include/linux/modversions.h if any of the .ver files changes, that's our marker to rebuild all modversions affected files. [PATCH] large x86 setup cleanup. Patrick Mochel did a great job here at splitting up some of the larger messy parts of arch/i386/kernel/setup.c, and introduced a nice abstraction which gives us a much nicer way to ensure we can add workarounds for vendor specific bugs / features without polluting other vendor code paths. Mark Haverkamp also brought this up to date for merging in my tree circa 2.5.14, and asides from 1-2 now fixed small thinkos, there haven't been any problems. This also features a workaround for an errata item on stepping C0 of the Intel Pentium 4 Xeon, which isn't in your tree yet, where we must disable the hardware prefetcher to ensure sane operation. kbuild: Build modules by default So here we are: make vmlinux/bzImage/whatever will now also build modules as it goes. Other than that, everything works as usual. "make modules" builds only the modules (but you shouldn't need it anymore). If you don't want the modules built, you can do "make KBUILD_MODULES= vmlinux/whatever" to only compile built-in objects. If people want it, I can also allow for "make vmlinux/whatever nomodules" to do the same. Also, add ' ' in Rules.make to properly align output in quiet mode. USB: emi26.c small fix to enable the driver to build properly [PATCH] ohci-hcd, fix urb unlink races This patch: - Fixes a longstanding urb unlink race, by switching to a single queue for EDs being unlinked. The previous two-queue scheme was sensitive to IRQ latencies: one extra millisecond would make it use the wrong queue. This updated scheme should handle latencies of up to 32K microseconds (Cthulu forfend :) and slightly shrinks object code size. - Related (mostly) cleanup. Some functions and one ED field renamed, ED layout is a smidgeon more compact (even given more data), driver version string doesn't reflect CVS, debug message only comes out in verbose mode. [PATCH] synchronous control/bulk messaging This one-liner fixes a problem in synchronous messaging with usb_bulk_msg(), usb_control_msg(), and everything that calls usb_control_msg(): you're not allowed to call blocking functions when you're already on a wait queue. A better fix would be to just stick the thread on the wait queue _after_ submitting the URB, but that should involve more testing than I have time for just now. [PATCH] UHCI bix for build error under unstable debian uhci_stop is __devexit but the pointer to it doesn't use __devexit_p. The fix is simple: USB: formatting cleanups for some USB drivers from the 2.5.20-dj3 tree Clean up and fix Makefile from x86 CPU split Fix x86 CPU merge dangling ends Fix sound compile error exposed by header file cleanups [PATCH] 2.5.20 swsusp: stop abusing headers with non-inlined functions swsusp kind-of abused include/asm-i386/suspend.h for stuff that belonged to arch/i386/kernel/suspend.c. This moves code where it belongs. Pavel device model update s/{driver,device}_bind/{driver,device}_attach/ and s/{driver,device}_unbind/{driver,device}_detach/ call bus's match callback instead of bind callback kbuild: Use a standard "update-if-changed" For some cases, we cannot decide if a target would change just by looking at its prequisites, i.e. it's quite likely that it remains the same even though a prequisite changed. The updated timestamp would cause a lot of unnecessary recompiles. In this case, we actually generate a temporary file, compare it to the old file and only the contents are different do overwrite the old file. The "update-if-changed" snippet remains always the same, so let's put it into a macro instead of duplicating it. After that change, scripts/mkversion_h is so small that we rather put the three remaining lines directly into the Makefile. Attempt to better locking in device model core: - remove device from driver's list on device_detach - set device's driver to NULL - decrement reference count on driver on device_detach - remove devices from driver's list in driver_detach - use a write_lock instead of read_lock - don't lock around initialization of device fields - assume we have a bus in __remove_driver (we enforce this in driver_register) - do put_bus last in __remove_driver - lock bus around atomic_set in remove_driver and atomic_dec_and_test in put_driver - remove from bus's list while we have it locked [PATCH] Have core/drivers.c include Currently, drivers/usb/core/drivers.c includes to get at . The following includes directly. Added usb-midi driver from NAGANO Daisuke with some porting from me. kbuild: -DMODULE for assembler source For consistency, it makes a lot of sense to add -DMODULE (plus the other special switches with CONFIG_MODVERSIONS) to .S files which get assembled. It surely fixes an ugly workaround for sparc64, guess we'll see if it breaks something for other archs (I wouldn't think so, though) kbuild: Additional config targets for testing This patch adds the following targets, which generate some configs useful for testing - which kind should be clear from the names: o allyesconfig o allmodconfig o allnoconfig o randconfig It also adds o defconfig which does the same as make oldconfig but uses the defaults for all new options without asking. The actual patch was done by Ghozlane Toumi, maintained in kbuild-2.5 by Keith Owens, and extracted by Sam Ravnborg and patch@luckynet.dynu.com. kbuild: Remove 2048 symbol limit from make xconfig This patch was extracted from Keith Owens' kbuild-2.5 by patch@luckynet.dynu.com kbuild: Remove absolute path in output of drivers/scsi/script_asm.pl Extracted from Keith Owens' kbuild-2.5 by patch@luckynet.dynu.com. kbuild: Fix tolower() usage in scripts/fixdep.c Not that it matters, since it's only used with 7bit chars, but anyway. kbuild: Use -nostdinc with kernel sources Kernel code should not include userspace headers. To avoid that happening by mistake, ask gcc not to look into the usual places when looking for headers. We still need to find gcc headers for vararg and similar, though - achieved by the "-iwithprefix include" which will look into /include, as proposed by Russell King. A similar patch went into 2.4, so not much breakage expected. Sparc32 code cleanups from 2.4.x. asm-generic.h: Add forward siginfo decl for the sake of HAVE_ARCH_SIGINFO_T platforms. init/main.c: Revert recent GCC requirement change. Fix generic device layer init sequence. Generic BUS objects have to be registered before devices assosciated with them are probed. Therefore subsys_initcall is inappropriate for such setups. It does not work to use core_initcall for this because the generic device layer bits need to be setup first too. So we rename unused_initcall to postcore_initcall and use this new initcall level for generic BUS object init. This fixes bootup on Alpha, and Sparc64. X86 was working by what looks to be luck in link order. SunRPC: Fix size_t vs. unsigned int arg descrepancy. register_netdevice: Fix return value handling on success. sparc64/kernel/Makefile: Remove bogus binfmt_elf32.o dependency target. Sparc64: Propagate exec MM handling changes to sparc32 emulation layer. Sparc64: Propagate forget_pte changes to sparc64 ioremap. Sparc: Adjust to new {clear,copy}_user_page calling convention. IDE: Print I/O ports more portably. IDE: Add missing printf format specifier when printing Sparc IRQ. IDE: Print 64-bit types more portably. kbuild: remove CLEAN_DIRS from Makfile kbuild: small toplevel makefile tidyup - add comment about the vmlinux link components - get rid of DRIVERS-y - always linking in sound/sound.o doesn't harm - sort SUBDIRS in link order - always use := to assign - remove supeflous $(TOPDIR) for lib/lib.a kbuild: Enforce UTS limit, use LANG=C for date/time Patch by Keith Owens, ported to 2.5. If the length of $(VERSION).$(PATCHLEVEL).$(SUBLEVEL)$(EXTRAVERSION) exceeds 64 characters it silently corrupts the utsname data, resulting in garbage for uname -r and problems running the kernel and modules. Abort if KERNELRELEASE is too long. Truncation is not good enough, it results in ambiguous /lib/modules/`uname -r` contents. Ensure that the date/time in uname are always in LANG=C. Users with other languages report that 8 bit values cause the boot messages to go haywire. Don't reference driver after you set pointer to NULL in device_detach [PATCH] Shared zlib include fix for 2.5 and 2.4-ac. This switches linux/zlib.h to include zconf.h as instead of just "zconf.h". [PATCH] Futex update I: Trivial comment removal This comment refers to the original implementation. Obsolete. [PATCH] Futex II: Copy-from-user can fail. This patch handles the case where copy_from_user fails (it could have been unmapped from this address space by another thread). [PATCH] Futex update III: don't use put_page... This uses page_cache_release() instead of put_page(), as it might be a pagecache page. [PATCH] fs/locks.c: add and use IS_{POSIX, FLOCK, LEASE} macros Another trivial part of a Matthew Wilcox patch. This just defines macros for distinguishing the differnet types of locks. [PATCH] Futex update IV: use a waitqueue. This turns the simple task pointer into a waitqueue, which is needed for the poll() patch. device detach locking, one more time: get driver and reset it in struct device before calling remove() kbuild: Add rules for compiling programs on the host At a couple of places, we need to build programs which run on the compile host during building the kernel. Add a new variable "host-progs" to declare such programs. Rules.make will automatically provide rules to compile these programs with appropriate command lines etc. Compared to having explicit rules spread around for doing so, this has the following advantages: o shorter o automatically figures out dependencies and handles changing command lines o Nicer output in quiet mode ;-) Compiling host programs from multiple sources is also possible, analogous to multi-part objects. E.g. scripts/Makefile has host-progs := tkparse tkparse-objs := tkparse.o tkcond.o tkgen.o kbuild: Fix extracting of CONFIG_ references by Sam Ravnborg: fixdep, when adding dependencies to config entires fails to take into account the last part of a config entry, if it contains more than one underscore. Example: CONFIG_PROC_FS generates $(wildcard include/config/proc.h) but should generate $(wildcard include/config/proc/fs.h) Attached patch fixes this. kbuild: dependency generation fixes o Fix a bug in fixdep.c, which could cause segfaults o Make sure that we build fixdep first of all, since we need it to build host programs. USB: split some pci specific pieces out of hcd.c into a separate file. ISDN: Add missing #include ISDN: Fix a typo in drivers/isdn/hisax/elsa.c One of the janitorial cleanups introduced pci_request_region() instead of request_region(). ISDN/CAPI: Remove some left-over from the capi_driver removal ISDN: Add PPP statistics in bytes (by Karsten Keil) ISDN: Export all hisax symbols from drivers/isdn/hisax/config.o As a courtesy to s390, remove drivers/isdn/hisax/fsm.o from $(export-objs), so that it doesn't clash with the S390 fsm.o, which is exporting symbols as well ;) ISDN: hisax/hfc_pci.c: sync with 2.4 ISDN: Add support for USR PCI TA (by Karsten Keil) ISDN: Cisco HDLC update Only handle reverse ARP special, leave the rest to the network layer. (by Bjoern A. Zeeb) ISDN: Add support for Eicon Diva 2.02 (by Joerg Petersohn/Karsten Keil) ISDN: Add DoV (Data over Voice) support (by Guy Ellis) ISDN: Add support for Formula-n enter:now, a.k.a. Gerdes Power ISDN (by Christoph Ersfeld/Karsten Keil) ISDN: Add in-kernel ISAPnP support to HiSax driven cards (by Karsten Keil) ISDN: MPPP crash fix prevent kernel oops with MPPP add some debugging code for critical MPPP situations (by Karsten Keil) ISDN: LED support for netjet driver (by Guy Ellis) ISDN/CAPI: Don't use special slab caches for CAPI objects In normal use, only handful of these structures are allocated, so there's no point in setting up private slab caches for them. [PATCH] 2.5.20 ov511.c compile fixes This patch allows ov511 to build again by removing references to urb->next. It now resubmits in the completion handler and properly sets urb->interval. [PATCH] uhci-hcd misc This has two minor tweaks to the uhci-hcd driver: - removes some duplicated code (HCD framework does that test) - corrects a FIXME comment (no issue) [PATCH] relocate error checks As was discussed a few weeks back, this moves most of the sanity checks and input conditioning for the HCD framework's usb_submit_urb() support directly into usb_submit_urb(), so that all HCDs (not just those using the sharable HCD framework support) can rely on them. USB: usb-midi driver: fixed memory flag, as pointed out by Oliver Neukum USB: hcd cleanups and documentation Implement many of the hcd cleanups that David Brownell had previously submitted. JFS: Fix structure alignment problem on 64-bit machines The code in xtLog assumes that the structures xdlistlock_t and maplock_t are the same size, as they are overlaid. On 64-bit machines, this was not the case. Ensure that the xdlistlock_t is 16 bytes in length regardless of the size of a pointer. JFS: Add hch's copyright. He has added significant content to these files. JFS: remove obsolete declaration IBM PCI Hotplug driver: polling thread locking cleanup removed a lot of bizzare polling locking logic, causing the driver to not sleep for 2 seconds with some locks held. This improves userspace interaction by a few orders of magnitude :) PCI Hotplug core: added #include to fix compile time warning IBM PCI Hotplug driver: added __init and __exit to functions that needed it. Thanks to Andrey Panin for pointing these out to me. [PATCH] fix urb->next removal in konicawc driver This patch removes the use of urb->next in konicawc.c and resubmits the URB in the completion handler. [PATCH] fix urb->next removal in usbvideo This patch removes the use of urb->next in usbvideo.c and resubmits the URB in the completion handler. [PATCH] USB pwc webcam patch thanks to suggestions from David Brownell I modified pwc driver to compile and work under 2.5.18. Also it fixes mas ISO transfer sizes which could potentionaly affect older kernels too. With these changes it works - I'll test stability and post results later. [PATCH] USB SA-1111 patch against usb-2.5 bitkeeper This adds SA-1111 support for ohci-hcd and fixes usb-ohci too. [PATCH] fs/locks.c: Only yield once for flocks This patch removes the annoying and confusing `wait' argument from many places. The only change in behaviour is that we now yield once when unblocking other BSD-style flocks instead of once for each lock. This slightly improves the semantics for userspace. Before, when we had two tasks waiting on a lock, the first one would receive the lock. Now, the one with the highest priority receives the lock. [PATCH] fs/locks.c: remove MSNFS define This patch removes the MSNFS defines. These locks have a wider purpose than emulating Microsoft NFS semantics. [PATCH] Move vmalloc wrappers out of include/linux/vmalloc.h This moves the vmalloc wrappers from into mm/vmalloc.c. Doing this will later allow us to remove from , along with some other #include fixups. [PATCH] Remove from This removes from . This then goes and fixes all of the files (x86 and PPC) which relied on implicit includes which don't happen anymore. This also takes out of fs/mpage.c and puts it into include/linux/bio.h where it belongs since references 'kdev_t' directly. A quick summary of the of the added includes: arch/i386/kernel/microcode.c: needs extern for num_physpages, in linux/mm.h include/linux/spinlock.h: local_irq* is defined in but this was never directly included. [PATCH] More work on removing from Fixes for include file cleanups [PATCH] fs/locks.c: more cleanup Define the for_each_lock macro and start replacing ugly special for loops with it. Rejig the interface between sys_flock and flock_lock_file to always pass a struct file_lock rather than a command. Eliminate some gotos by simplifying the logic. Remove some redundant initialisation. [PATCH] capability.c thinko Apparently my understanding of C brace rules was blurred and there is a thinko in kernel/capability.c - what we want is what the indention reflects but what we get is an unconditional return of -EINVAL. [PATCH] set_cpus_allowed fix Mike Kravetz pointed out that the set_cpus_allowed optimization he suggested has a small but possible race condition wherein the system could still be operating in the context of the task but it is not running. On top of this, the runqueue lock can be dropped in load_balance and thus we can race and set task->cpu at a very unpleasant time. My solution is to just remove the optimization. [PATCH] matroxfb dies when you try to use secondary head in 2.5.x James introduced bad bug which causes NULL pointer dereference as soon as you'll try to use secondary head because of screen_base is not initialized (it is initialized on wrong head). [PATCH] missing GET_CPU_IDX in i386 entry.S resume_kernel uses CPU_IDX but never uses GET_CPU_IDX to get the index. This is an issue when smp and preemption are both enabled. I also removed the unused GET_CURRENT_CPU_IDX. [PATCH] s/390 patches for 2.5.20 (1 of 4). This first patch contains the changes to arch/{s390,s390x} [PATCH] s/390 patches for 2.5.20 (2 of 4). Second patch of the s/390 update. Contains all the include file changes in include/asm-{s390,s390x}. [PATCH] s/390 patches for 2.5.20 (3 of 4). Third patch of the s/390 update. Everything from drivers/s390. This patch contains the main bunch of changes. Major changes are: 1) the restructured dasd driver, 2) the new tape driver, 3) the start of the common io layer rework, 4) the gpl'ed lcs driver and 5) the gpl'ed qdio layer [PATCH] s/390 patches for 2.5.20 (4 of 4). Fourth and last part of the s/390 update. Docu stuff. ACPI sleep depends on software suspend (at least for now) [PATCH] Reduce number of LOOKUP calls in nfs_lookup_revalidate() fs/nfs/inode.c: - Ensure that we initialize NFS_MTIME_UPDATE(inode). - If a directory's mtime changed due to some action on our part (file creation/deletion/...) don't force a dcache revalidation by changing NFS_MTIME_UPDATE(inode). [PATCH] Catch a few more cases where we need to renew inode->d_time fs/nfs/dir.c: - Call nfs_renew_times() when we deliberately create a negative dentry. or rename a file. Fall-out from header file cleanups "make clean": get rid of temporary directories too Kernel version 2.5.21