The state of the BKL-removal tree
The bkl-removal tree lives at:
git://git.lwn.net/linux-2.6.git bkl-removal
It has accumulated a fair amount of stuff at this point. Lots of pushdown work and a significant API change: the open() and fasync() file_operations are now called without the big kernel lock.
This tree has been in linux-next since mid-May. In the absence of reasons to the contrary, my intent is to ask for this to be merged for 2.6.27. Objections, anybody?
Andi Kleen (1):
Remove BKL from remote_llseek v2
Arnd Bergmann (68):
agp-frontend: BKL pushdown
ans-lcd: BKL pushdown
apm-emulation: BKL pushdown
apm_32: BKL pushdown
basler-excite: BKL pushdown
bf561-coreb: BKL pushdown
block-dasd_eer: BKL pushdown
bluetooth-vhci: BKL pushdown
briq_panel: BKL pushdown
bvme6000-rtc: BKL pushdown
cpwatchdog: BKL pushdown
crypto-prng: BKL pushdown
crypto-zcrypt_api: BKL pushdown
display7seg: BKL pushdown
dlm-user: BKL pushdown
ds1286: BKL pushdown
ds1620: BKL pushdown
efirtc: BKL pushdown
envctrl: BKL pushdown
flash: BKL pushdown
genrtc: BKL pushdown
hdpu_cpustate: BKL pushdown
hp_sdc_rtc: BKL pushdown
hpet: BKL pushdown
hw-random: BKL pushdown
infiniband-ucma: BKL pushdown
ip27-rtc: BKL pushdown
ipmi-watchdog: BKL pushdown
irnet_ppp: BKL pushdown
jsflash: BKL pushdown
lcd: BKL pushdown
macintosh-smu: BKL pushdown
media-radio-miropcm20-rds: BKL pushdown
megaraid: BKL pushdown
message-i2o-i2o_config: BKL pushdown
microcode: BKL pushdown
monreader: BKL pushdown
monwriter: BKL pushdown
mousedev: BKL pushdown
mvme16x-rtc: BKL pushdown
mwave-mwavedd: BKL pushdown
net-tun: BKL pushdown
nvram: BKL pushdown
ocfs2-stack_user: BKL pushdown
openprom: BKL pushdown
parisc-eisa_eeprom: BKL pushdown
parisc-kernel-perf: BKL pushdown
pm_qos_params: BKL pushdown
riowatchdog: BKL pushdown
rtc: BKL pushdown
rtc-rtc-m41t80: BKL pushdown
sbus-rtc: BKL pushdown
scsi-tgt: BKL pushdown
serio: BKL pushdown
sony-laptop: BKL pushdown
sonypi: BKL pushdown
sparc-apc: BKL pushdown
sparc64-rtc: BKL pushdown
tpm-tpm: BKL pushdown
uctrl: BKL pushdown
uinput: BKL pushdown
uml-harddog: BKL pushdown
uml-mmapper: BKL pushdown
uml-random: BKL pushdown
via-pmu: BKL pushdown
vmcp: BKL pushdown
vmwatchdog: BKL pushdown
x86-mce: BKL pushdown
Jonathan Corbet (73):
bsg: cdev lock_kernel() pushdown
cris: cdev lock_kernel() pushdown
mips: cdev lock_kernel() pushdown
sh: cdev lock_kernel() pushdown
x86: cdev lock_kernel() pushdown
i2c: cdev lock_kernel() pushdown
cosa: cdev lock_kernel() pushdown
pcmcia: cdev lock_kernel() pushdown
rtc: cdev lock_kernel() pushdown drivers/s390: cdev lock_kernel() pushdown
AoE: cdev lock_kernel() pushdown
paride: cdev lock_kernel() pushdown
mtdchar: cdev lock_kernel() pushdown
UBI: cdev lock_kernel() pushdown
HID: cdev lock_kernel() pushdown
Input: cdev lock_kernel() pushdown
UIO: cdev lock_kernel() pushdown
ipmi: cdev lock_kernel() pushdown
mem: cdev lock_kernel() pushdown
misc: cdev lock_kernel() pushdown
viotape: cdev lock_kernel pushdown ()
mbcs: cdev lock_kernel() pushdown
lp: cdev lock_kernel() pushdown
drm: cdev lock_kernel() pushdown
phonedev: cdev lock_kernel() pushdown
ide-tape: cdev lock_kernel() pushdown
sg: cdev lock_kernel() pushdown
osst: cdev lock_kernel() pushdown.
aacraid: cdev lock_kernel() pushdown
st: cdev lock_kernel() pushdown
gdth: cdev lock_kernel() pushdown
isdn: cdev lock_kernel() pushdown
usbcore: cdev lock_kernel() pushdown
dvb: cdev lock_kernel() pushdown
fbmem: cdev lock_kernel() pushdown
sound: cdev lock_kernel() pushdown
snsc: cdev lock_kernel() pushdown
tty: cdev lock_kernel() pushdown
Add "no BKL needed" comments to several drivers
spidev: BKL pushdown
vcs: BKL pushdown
xilinx icap: BKL pushdown
tlckl: BKL pushdown
raw: BKL pushdown
dsp56k: BKL pushdown
infiniband: more BKL pushdown
phantom: BKL pushdown
bpp: bkl pushdown
videopix: BKL pushdown
dpt_i20: BKL pushdown
changer: BKL pushdown
CAPI: BKL pushdown
divamnt: BKL pushdown
adb: BKL pushdown
printer gadget: BKL pushdown
USB Monitor: BKL pushdown
usbdev: BKL pushdown
videodev: BKL pushdown
Add cycle_kernel_lock()
Add a bunch of cycle_kernel_lock() calls Add a comment in chrdev_open()
Remove the lock_kernel() call from chrdev_open() Make FAT users happier by not deadlocking Merge commit 'v2.6.26-rc6' into bkl-removal
mpt: fasync BKL pushdown
i2o: fasync BKL pushdown
tun: fasync BKL pushdown
tty_io: fasync BKL pushdown
Bluetooth VHCI: fasync BKL pushdown
ecryptfs: fasync BKL pushdown
ipmi: fasync BKL pushdown
snd/PCM: fasync BKL pushdown
Call fasync() functions without the BKL
Linus Torvalds (1):
Replace BKL with superblock lock in fat/msdos/vfat
Stephen Rothwell (1):
bkl-removal viotape fixup
arch/blackfin/mach-bf561/coreb.c | 4 ++ arch/cris/arch-v10/drivers/eeprom.c | 4 +- arch/cris/arch-v10/drivers/gpio.c | 3 ++ arch/cris/arch-v10/drivers/i2c.c | 2 + arch/cris/arch-v10/drivers/sync_serial.c | 34 +++++++++++++-------- arch/cris/arch-v32/drivers/cryptocop.c | 3 +- arch/cris/arch-v32/drivers/i2c.c | 2 + arch/cris/arch-v32/drivers/mach-a3/gpio.c | 4 ++ arch/cris/arch-v32/drivers/mach-fs/gpio.c | 5 ++- arch/cris/arch-v32/drivers/sync_serial.c | 33 ++++++++++++-------- arch/m68k/bvme6000/rtc.c | 7 ++++- arch/m68k/mvme16x/rtc.c | 4 ++ arch/mips/basler/excite/excite_iodev.c | 9 +++++- arch/mips/kernel/rtlx.c | 7 ++++- arch/mips/kernel/vpe.c | 12 ++++++-- arch/mips/sibyte/common/sb_tbprof.c | 25 +++++++++++---- arch/parisc/kernel/perf.c | 4 ++ arch/s390/crypto/prng.c | 2 + arch/sh/boards/landisk/gio.c | 10 ++++-- arch/sparc/kernel/apc.c | 2 + arch/sparc64/kernel/time.c | 7 ++++- arch/um/drivers/harddog_kern.c | 3 ++ arch/um/drivers/mmapper_kern.c | 2 + arch/um/drivers/random.c | 3 ++ arch/x86/kernel/apm_32.c | 4 ++ arch/x86/kernel/cpu/mcheck/mce_64.c | 4 ++ arch/x86/kernel/cpuid.c | 25 +++++++++++----- arch/x86/kernel/microcode.c | 2 + arch/x86/kernel/msr.c | 16 +++++++-- block/bsg.c | 7 ++++- drivers/block/aoe/aoechr.c | 7 ++++- drivers/block/paride/pg.c | 22 ++++++++++---- drivers/block/paride/pt.c | 8 ++++- drivers/bluetooth/hci_vhci.c | 14 +++++++-- drivers/char/agp/frontend.c | 4 ++ drivers/char/apm-emulation.c | 3 ++ drivers/char/briq_panel.c | 9 ++++- drivers/char/cs5535_gpio.c | 2 + drivers/char/drm/drm_fops.c | 9 ++++- drivers/char/ds1286.c | 4 ++ drivers/char/ds1620.c | 9 +++++- drivers/char/dsp56k.c | 16 +++++++--- drivers/char/dtlk.c | 3 ++ drivers/char/efirtc.c | 2 + drivers/char/genrtc.c | 7 ++++- drivers/char/hpet.c | 4 ++ drivers/char/hw_random/core.c | 2 + drivers/char/ip2/ip2main.c | 34 +-------------------- drivers/char/ip27-rtc.c | 4 ++ drivers/char/ipmi/ipmi_devintf.c | 10 +++++- drivers/char/ipmi/ipmi_watchdog.c | 3 ++ drivers/char/lcd.c | 3 ++ drivers/char/lp.c | 38 ++++++++++++++++------- drivers/char/mbcs.c | 5 +++ drivers/char/mem.c | 10 +++++- drivers/char/misc.c | 3 ++ drivers/char/mwave/mwavedd.c | 2 + drivers/char/nvram.c | 4 ++ drivers/char/pc8736x_gpio.c | 2 + drivers/char/ppdev.c | 2 + drivers/char/raw.c | 3 ++ drivers/char/rtc.c | 4 ++ drivers/char/scx200_gpio.c | 2 + drivers/char/snsc.c | 5 ++- drivers/char/sonypi.c | 3 ++ drivers/char/tb0219.c | 2 + drivers/char/tlclk.c | 19 +++++++---- drivers/char/tpm/tpm.c | 5 +++ drivers/char/tty_io.c | 41 +++++++++++++++++++++---- drivers/char/vc_screen.c | 9 ++++- drivers/char/viotape.c | 3 ++ drivers/char/vr41xx_giu.c | 2 + drivers/char/xilinx_hwicap/xilinx_hwicap.c | 6 +++- drivers/hid/hidraw.c | 3 ++ drivers/i2c/i2c-dev.c | 22 ++++++++++---- drivers/ide/ide-tape.c | 7 ++++- drivers/infiniband/core/ucm.c | 2 + drivers/infiniband/core/ucma.c | 3 ++ drivers/infiniband/core/user_mad.c | 7 ++++- drivers/infiniband/core/uverbs_main.c | 9 ++++- drivers/infiniband/hw/ipath/ipath_file_ops.c | 2 + drivers/input/input.c | 16 +++++++-- drivers/input/misc/hp_sdc_rtc.c | 2 + drivers/input/misc/uinput.c | 3 ++ drivers/input/mousedev.c | 12 ++++++- drivers/input/serio/serio_raw.c | 6 +++- drivers/isdn/capi/capi.c | 17 +++++++---- drivers/isdn/hardware/eicon/divamnt.c | 16 +++++++--- drivers/isdn/hardware/eicon/divasi.c | 2 + drivers/isdn/hardware/eicon/divasmain.c | 2 + drivers/isdn/i4l/isdn_common.c | 3 +- drivers/macintosh/adb.c | 18 ++++++++--- drivers/macintosh/ans-lcd.c | 2 + drivers/macintosh/smu.c | 3 ++ drivers/macintosh/via-pmu.c | 3 ++ drivers/media/dvb/dvb-core/dvbdev.c | 4 ++ drivers/media/radio/miropcm20-rds.c | 4 ++ drivers/media/video/videodev.c | 4 ++ drivers/message/fusion/mptctl.c | 6 +++- drivers/message/i2o/i2o_config.c | 12 +++++-- drivers/misc/hdpuftrs/hdpu_cpustate.c | 9 +++++- drivers/misc/phantom.c | 9 ++++- drivers/misc/sony-laptop.c | 3 ++ drivers/mtd/mtdchar.c | 22 ++++++++++---- drivers/mtd/ubi/cdev.c | 7 ++++- drivers/net/ppp_generic.c | 2 + drivers/net/tun.c | 13 ++++++-- drivers/net/wan/cosa.c | 22 ++++++++++---- drivers/parisc/eisa_eeprom.c | 3 ++ drivers/pcmcia/pcmcia_ioctl.c | 25 +++++++++++---- drivers/rtc/rtc-dev.c | 12 ++++++-- drivers/rtc/rtc-m41t80.c | 7 ++++- drivers/s390/block/dasd_eer.c | 6 ++++ drivers/s390/char/fs3270.c | 23 ++++++++++---- drivers/s390/char/monreader.c | 4 ++ drivers/s390/char/monwriter.c | 3 ++ drivers/s390/char/tape_char.c | 12 ++++++-- drivers/s390/char/vmcp.c | 4 ++ drivers/s390/char/vmlogrdr.c | 8 ++++- drivers/s390/char/vmur.c | 12 ++++++-- drivers/s390/char/vmwatchdog.c | 7 ++++- drivers/s390/crypto/zcrypt_api.c | 3 ++ drivers/sbus/char/bpp.c | 3 ++ drivers/sbus/char/cpwatchdog.c | 4 ++ drivers/sbus/char/display7seg.c | 1 + drivers/sbus/char/envctrl.c | 2 + drivers/sbus/char/flash.c | 6 +++- drivers/sbus/char/jsflash.c | 13 ++++++-- drivers/sbus/char/openprom.c | 3 ++ drivers/sbus/char/riowatchdog.c | 2 + drivers/sbus/char/rtc.c | 3 ++ drivers/sbus/char/uctrl.c | 3 ++ drivers/sbus/char/vfc_dev.c | 5 +++ drivers/scsi/3w-9xxx.c | 3 ++ drivers/scsi/3w-xxxx.c | 3 ++ drivers/scsi/aacraid/linit.c | 3 ++ drivers/scsi/ch.c | 4 ++ drivers/scsi/dpt_i2o.c | 5 +++ drivers/scsi/gdth.c | 3 ++ drivers/scsi/megaraid.c | 5 ++- drivers/scsi/megaraid/megaraid_mm.c | 2 + drivers/scsi/megaraid/megaraid_sas.c | 2 + drivers/scsi/osst.c | 15 +++++++++- drivers/scsi/scsi_tgt_if.c | 2 + drivers/scsi/sg.c | 16 ++++++++-- drivers/scsi/st.c | 11 ++++++- drivers/spi/spidev.c | 3 ++ drivers/telephony/phonedev.c | 3 ++ drivers/uio/uio.c | 17 +++++++--- drivers/usb/core/devio.c | 2 + drivers/usb/core/file.c | 3 ++ drivers/usb/gadget/printer.c | 3 +- drivers/usb/mon/mon_bin.c | 6 ++++ drivers/video/fbmem.c | 15 +++++++-- fs/char_dev.c | 7 ++-- fs/cifs/cifsfs.c | 2 +- fs/dlm/user.c | 9 +++++- fs/ecryptfs/file.c | 3 ++ fs/fat/cache.c | 2 +- fs/fat/dir.c | 4 +- fs/fat/file.c | 6 ---- fs/fat/inode.c | 26 ++++++++++------ fs/fcntl.c | 3 -- fs/gfs2/ops_file.c | 4 +- fs/msdos/namei.c | 35 ++++++++++++---------- fs/ncpfs/file.c | 12 +++++++- fs/nfs/file.c | 6 +++- fs/ocfs2/stack_user.c | 3 ++ fs/read_write.c | 38 +++++++----------------- fs/smbfs/file.c | 11 ++++++- fs/vfat/namei.c | 35 ++++++++++++---------- include/linux/fs.h | 3 +- include/linux/smp_lock.h | 13 ++++++++ kernel/pm_qos_params.c | 7 ++++- net/irda/irnet/irnet.h | 1 + net/irda/irnet/irnet_ppp.c | 3 ++ sound/core/pcm_native.c | 8 ++++- sound/core/sound.c | 15 +++++++++- sound/sound_core.c | 5 +++ 179 files changed, 1084 insertions(+), 352 deletions(-)
jon
Quoted 3 lines · Jonathan Corbet
This tree has been in linux-next since mid-May. In the absence of reasons to the contrary, my intent is to ask for this to be merged for 2.6.27. Objections, anybody?
i see this still has this unnecessary hunk:
Quoted 2 lines
bf561-coreb: BKL pushdown
arch/blackfin/mach-bf561/coreb.c | 4 ++
-mike
PATCH 1/1 IB/uverbs: BKL is not needed for ib_uverbs_open()
In reply to Jonathan CorbetRemove explicit lock_kernel() calls and document why the code is safe.
---
Hi Jon,
Is this sort of patch the next step in the BKL-removal plan? Is now an appropriate time to send such things?
If this looks good to you I will do a careful audit of the rest of drivers/infiniband/core and send you a few more similar patches.
Thanks,
Roland
drivers/infiniband/core/uverbs_main.c | 20 +++++++++++++------- 1 files changed, 13 insertions(+), 7 deletions(-)
drivers/infiniband/core/uverbs_main.c +13 -7
@@ -45,7 +45,6 @@#include <linux/file.h> #include <linux/mount.h> #include <linux/cdev.h> -#include <linux/smp_lock.h> #include <asm/uaccess.h>
@@ -611,23 +610,32 @@ static int ib_uverbs_mmap(struct file *filp, struct vm_area_struct *vma)return file->device->ib_dev->mmap (file->ucontext, vma); } +/* +* ib_uverbs_open() does not need the BKL: +* +* - dev_table[] accesses are protected by map_lock, the +* ib_uverbs_device structures are properly reference counted, and +* everything else is purely local to the file being created, so +* races against other open calls are not a proble; +* - there is no ioctl method to race against; +* - the device is added to dev_table[] as the last part of module +* initialization, the open method will either immediately run +* -ENXIO, or all required initialization will be done. +*/ static int ib_uverbs_open (struct inode *inode,struct file *filp) {struct ib_uverbs_device *dev;struct ib_uverbs_file *file;int ret; -lock_kernel ();spin_lock (&map_lock); dev = dev_table[iminor (inode) -IB_UVERBS_BASE_MINOR ];if (dev)kref_get (&dev->ref);spin_unlock (&map_lock); -if (!dev) { -unlock_kernel (); +if (!dev)return -ENXIO ; - }if (!try_module_get (dev->ib_dev->owner)) { ret = -ENODEV ;
@@ -648,7 +656,6 @@ static int ib_uverbs_open(struct inode *inode, struct file *filp) filp->private_data = file; -unlock_kernel ();return 0 ; err_module:
@@ -656,7 +663,6 @@ err_module: err:kref_put (&dev->ref, ib_uverbs_release_dev); -unlock_kernel ();return ret; }
Signature · 1 line
1.5.6
Re: [PATCH] IB/uverbs: BKL is not needed for ib_uverbs_open()
Quoted 5 lines · On Fri, 27 Jun 2008 14:43:20 -0700 Roland Dreier <[email protected]> wrote:
Is this sort of patch the next step in the BKL-removal plan? Is now an appropriate time to send such things?
If this looks good to you I will do a careful audit of the rest of
drivers/infiniband/coreand send you a few more similar patches.
That's the general idea: audit the code where the BKL appears and remove it when it's safe to do so. It's always an appropriate time; the only question is whether you think this change is good to go into the tree for 2.6.27, or whether it should hang out in linux-next for one more cycle. I would assume that the extra time would not be necessary.
Thanks,
jon
Re: [ofa-general] Re: [PATCH] IB/uverbs: BKL is not needed for ib_uverbs_open()
Quoted 6 lines
That's the general idea: audit the code where the BKL appears and remove it when it's safe to do so. It's always an appropriate time; the only question is whether you think this change is good to go into the tree for 2.6.27, or whether it should hang out in linux-next for one more cycle. I would assume that the extra time would not be necessary.
Yes, this change should be good to go for 2.6.27.
Thanks,
Roland