index : litmus-rt.git	archive/unc-master-3.0 archived-2013.1 archived-private-master archived-semi-part demo ecrts-pgm-final ecrts14-pgm-final gpusync-rtss12 gpusync/staging linux-tip litmus2008-patch-series master pgm prop/litmus-signals prop/robust-tie-break staging test tracing-devel v2.6.34-with-arm-patches v2015.1 wip-2011.2-bbb wip-2011.2-bbb-trace wip-2012.3-gpu wip-2012.3-gpu-preport wip-2012.3-gpu-rtss13 wip-2012.3-gpu-sobliv-budget-w-kshark wip-aedzl-final wip-aedzl-revised wip-arbit-deadline wip-aux-tasks wip-bbb wip-bbb-prio-don wip-better-break wip-binary-heap wip-budget wip-color wip-color-jlh wip-d10-hz1000 wip-default-clustering wip-dissipation-jerickso wip-dissipation2-jerickso wip-ecrts14-pgm wip-edf-hsb wip-edf-os wip-edf-tie-break wip-edzl-critique wip-edzl-final wip-edzl-revised wip-events wip-extra-debug wip-fix-switch-jerickso wip-fix3 wip-fmlp-dequeue wip-ft-irq-flag wip-gpu-cleanup wip-gpu-interrupts wip-gpu-rtas12 wip-gpu-rtss12 wip-gpu-rtss12-srp wip-gpusync-merge wip-ikglp wip-k-fmlp wip-kernel-coloring wip-kernthreads wip-klmirqd-to-aux wip-kshark wip-litmus-3.2 wip-litmus2011.2 wip-litmus3.0-2011.2 wip-master-2.6.33-rt wip-mc wip-mc-bipasa wip-mc-jerickso wip-mc2-cache-slack wip-mcrit-mac wip-merge-3.0 wip-merge-v3.0 wip-migration-affinity wip-mmap-uncache wip-modechange wip-nested-locking wip-omlp-gedf wip-pai wip-percore-lib wip-performance wip-pgm wip-pgm-split wip-pm-ovd wip-prio-inh wip-prioq-dgl wip-refactored-gedf wip-release-master-fix wip-robust-tie-break wip-rt-kshark wip-rtas12-pgm wip-semi-part wip-semi-part-edfos-jerickso wip-shared-lib wip-shared-lib2 wip-shared-mem wip-splitting-jerickso wip-splitting-omlp-jerickso wip-stage-binheap wip-sun-port wip-timer-trace wip-tracepoints
	The LITMUS^RT kernel.	Bjoern Brandenburg

aboutsummaryrefslogtreecommitdiffstats

log msg author committer range

Branch	Commit message	Author	Age
archive/unc-master-3.0	P-FP: fix BUG_ON releated to priority inheritance	Bjoern Brandenburg	13 years
archived-2013.1	uncachedev: mmap memory that is not cached by CPUs	Glenn Elliott	12 years
archived-private-master	Merge branch 'wip-2.6.34' into old-private-master	Andrea Bastoni	15 years
archived-semi-part	Merge branch 'wip-semi-part' of ssh://cvs/cvs/proj/litmus/repo/litmus2010 int...	Andrea Bastoni	15 years
demo	Further refinements	Jonathan Herman	14 years
ecrts-pgm-final	Merge branch 'wip-ecrts14-pgm' of ssh://rtsrv.cs.unc.edu/home/litmus/litmus-r...	Glenn Elliott	12 years
ecrts14-pgm-final	Merge branch 'wip-ecrts14-pgm' of ssh://rtsrv.cs.unc.edu/home/litmus/litmus-r...	Glenn Elliott	12 years
gpusync-rtss12	Final GPUSync implementation.	Glenn Elliott	12 years
gpusync/staging	Rename IKGLP R2DGLP.	Glenn Elliott	12 years
linux-tip	Merge branch 'slab/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/p...	Linus Torvalds	15 years
litmus2008-patch-series	add i386 feather-trace implementation	Bjoern B. Brandenburg	16 years
master	PSN-EDF: use inferred_sporadic_job_release_at	Bjoern Brandenburg	9 years
pgm	make it compile	Glenn Elliott	12 years
prop/litmus-signals	Infrastructure for Litmus signals.	Glenn Elliott	13 years
prop/robust-tie-break	Fixed bug in edf_higher_prio().	Glenn Elliott	13 years
staging	Fix tracepoint compilation error	Felipe Cerqueira	13 years
test	9/23/2016	Namhoon Kim	9 years
tracing-devel	Test kernel tracing events capabilities	Andrea Bastoni	16 years
v2.6.34-with-arm-patches	smsc911x: Add spinlocks around registers access	Catalin Marinas	15 years
v2015.1	Add ARM syscall def for get_current_budget	Bjoern Brandenburg	10 years
wip-2011.2-bbb	Litmus core: simplify np-section protocol	Bjoern B. Brandenburg	14 years
wip-2011.2-bbb-trace	Refactor sched_trace_log_message() -> debug_trace_log_message()	Andrea Bastoni	14 years
wip-2012.3-gpu	SOBLIV draining support for C-EDF.	Glenn Elliott	12 years
wip-2012.3-gpu-preport	pick up last C-RM file	Glenn Elliott	12 years
wip-2012.3-gpu-rtss13	Fix critical bug in GPU tracker.	Glenn Elliott	12 years
wip-2012.3-gpu-sobliv-budget-w-kshark	Proper sobliv draining and many bug fixes.	Glenn Elliott	12 years
wip-aedzl-final	Make it easier to compile AEDZL interfaces in liblitmus.	Glenn Elliott	15 years
wip-aedzl-revised	Add sched_trace data for Apative EDZL	Glenn Elliott	15 years
wip-arbit-deadline	Fix compilation bug.	Glenn Elliott	13 years
wip-aux-tasks	Description of refined aux task inheritance.	Glenn Elliott	13 years
wip-bbb	GSN-EDF & Core: improve debug TRACE'ing for NP sections	Bjoern B. Brandenburg	14 years
wip-bbb-prio-don	use correct timestamp	Bjoern B. Brandenburg	14 years
wip-better-break	Implement hash-based EDF tie-breaking.	Glenn Elliott	13 years
wip-binary-heap	Make C-EDF work with simplified binheap_delete	Glenn Elliott	13 years
wip-budget	Added support for choices in budget policy enforcement.	Glenn Elliott	15 years
wip-color	Summarize schedulability with final record	Jonathan Herman	13 years
wip-color-jlh	sched_color: Fixed two bugs causing crashing on experiment restart and a rare...	Jonathan Herman	13 years
wip-d10-hz1000	Enable HZ=1000 on District 10	Bjoern B. Brandenburg	15 years
wip-default-clustering	Feature: Make default C-EDF clustering compile-time configurable.	Glenn Elliott	15 years
wip-dissipation-jerickso	Update from 2.6.36 to 2.6.36.4	Jeremy Erickson	11 years
wip-dissipation2-jerickso	Update 2.6.36 to 2.6.36.4	Jeremy Erickson	11 years
wip-ecrts14-pgm	Merge branch 'wip-ecrts14-pgm' of ssh://rtsrv.cs.unc.edu/home/litmus/litmus-r...	Glenn Elliott	12 years
wip-edf-hsb	last tested version	Jonathan Herman	14 years
wip-edf-os	Lookup table EDF-os	Jeremy Erickson	12 years
wip-edf-tie-break	Merge branch 'wip-edf-tie-break' of ssh://rtsrv.cs.unc.edu/home/litmus/litmus...	Glenn Elliott	13 years
wip-edzl-critique	Use hr_timer's active checks instead of having own flag.	Glenn Elliott	15 years
wip-edzl-final	Implementation of the EDZL scheduler.	Glenn Elliott	15 years
wip-edzl-revised	Clean up comments.	Glenn Elliott	15 years
wip-events	Added support for tracing arbitrary actions.	Jonathan Herman	15 years
wip-extra-debug	DBG: add additional tracing	Bjoern B. Brandenburg	15 years
wip-fix-switch-jerickso	Attempt to fix race condition with plugin switching	Jeremy Erickson	15 years
wip-fix3	sched: show length of runqueue clock deactivation in /proc/sched_debug	Bjoern B. Brandenburg	15 years
wip-fmlp-dequeue	Improve FMLP queue management.	Glenn Elliott	14 years
wip-ft-irq-flag	Feather-Trace: keep track of interrupt-related interference.	Bjoern B. Brandenburg	14 years
wip-gpu-cleanup	Enable sched_trace log injection from userspace	Glenn Elliott	13 years
wip-gpu-interrupts	Remove option for threading of all softirqs.	Glenn Elliott	14 years
wip-gpu-rtas12	Generalized GPU cost predictors + EWMA. (untested)	Glenn Elliott	13 years
wip-gpu-rtss12	Final GPUSync implementation.	Glenn Elliott	13 years
wip-gpu-rtss12-srp	experimental changes to support GPUs under SRP	Glenn Elliott	13 years
wip-gpusync-merge	Cleanup priority tracking for budget enforcement.	Glenn Elliott	11 years
wip-ikglp	Move RSM and IKGLP imp. to own .c files	Glenn Elliott	13 years
wip-k-fmlp	Merge branch 'mpi-master' into wip-k-fmlp	Glenn Elliott	14 years
wip-kernel-coloring	Added recolor syscall	Namhoon Kim	7 years
wip-kernthreads	Kludge work-queue processing into klitirqd.	Glenn Elliott	15 years
wip-klmirqd-to-aux	Allow klmirqd threads to be given names.	Glenn Elliott	13 years
wip-kshark	Merge branch 'mpi-staging' into wip-kshark	Jonathan Herman	13 years
wip-litmus-3.2	Merge commit 'v3.2' into litmus-staging	Andrea Bastoni	13 years
wip-litmus2011.2	Cleanup: Coding conformance for affinity stuff.	Glenn Elliott	14 years
wip-litmus3.0-2011.2	Feather-Trace: keep track of interrupt-related interference.	Bjoern B. Brandenburg	14 years
wip-master-2.6.33-rt	Avoid deadlock when switching task policy to BACKGROUND (ugly)	Andrea Bastoni	15 years
wip-mc	Removed ARM-specific hacks which disabled less common mixed-criticality featu...	Jonathan Herman	12 years
wip-mc-bipasa	MC-EDF added	bipasa chattopadhyay	13 years
wip-mc-jerickso	Split C/D queues	Jeremy Erickson	15 years
wip-mc2-cache-slack	Manually patched mc^2 related code	Ming Yang	10 years
wip-mcrit-mac	cosmetic	Mac Mollison	15 years
wip-merge-3.0	Prevent Linux to send IPI and queue tasks on remote CPUs.	Andrea Bastoni	14 years
wip-merge-v3.0	Prevent Linux to send IPI and queue tasks on remote CPUs.	Andrea Bastoni	14 years
wip-migration-affinity	NULL affinity dereference in C-EDF.	Glenn Elliott	14 years
wip-mmap-uncache	share branch with others	Glenn Elliott	13 years
wip-modechange	RTSS 2017 submission	Namhoon Kim	8 years
wip-nested-locking	Appears to be working.	Bryan Ward	12 years
wip-omlp-gedf	First implementation of G-OMLP.	Glenn Elliott	15 years
wip-pai	Some cleanup of PAI	Glenn Elliott	14 years
wip-percore-lib	9/21/2016	Namhoon Kim	9 years
wip-performance	CONFIG_DONT_PREEMPT_ON_TIE: Don't preeempt a scheduled task on priority tie.	Glenn Elliott	14 years
wip-pgm	Add PGM support to C-FL	Glenn Elliott	12 years
wip-pgm-split	First draft of C-FL-split	Namhoon Kim	12 years
wip-pm-ovd	Add preemption-and-migration overhead tracing support	Andrea Bastoni	15 years
wip-prio-inh	P-EDF updated to use the generic pi framework.	Glenn Elliott	15 years
wip-prioq-dgl	BUG FIX: Support DGLs with PRIOQ_MUTEX	Glenn Elliott	13 years
wip-refactored-gedf	Generalizd architecture for GEDF-style scheduelrs to reduce code redundancy.	Glenn Elliott	15 years
wip-release-master-fix	bugfix: release master CPU must signal task was picked	Bjoern B. Brandenburg	14 years
wip-robust-tie-break	EDF priority tie-breaks.	Glenn Elliott	13 years
wip-rt-kshark	Move task time accounting into the complete_job method.	Jonathan Herman	13 years
wip-rtas12-pgm	Scheduling of PGM jobs.	Glenn Elliott	13 years
wip-semi-part	Fix compile error with newer GCC	Jeremy Erickson	12 years
wip-semi-part-edfos-jerickso	Use initial CPU set by client	Jeremy Erickson	12 years
wip-shared-lib	TODO: Fix condition checks in replicate_page_move_mapping()	Namhoon Kim	9 years
wip-shared-lib2	RTAS 2017 Submission ver.	Namhoon Kim	9 years
wip-shared-mem	Initial commit for shared library	Namhoon Kim	9 years
wip-splitting-jerickso	Fix release behavior	Jeremy Erickson	13 years
wip-splitting-omlp-jerickso	Bjoern's Dissertation Code with Priority Donation	Jeremy Erickson	13 years
wip-stage-binheap	An efficient binary heap implementation.	Glenn Elliott	13 years
wip-sun-port	Dynamic memory allocation and clean exit for FeatherTrace	Christopher Kenna	15 years
wip-timer-trace	bugfix: C-EDF, clear scheduled field of the correct CPU upon task_exit	Andrea Bastoni	15 years
wip-tracepoints	Add kernel-style events for sched_trace_XXX() functions	Andrea Bastoni	14 years
Tag	Download	Author	Age
2015.1	commit 8e51b37822...	Bjoern Brandenburg	10 years
2013.1	commit bcaacec1ca...	Glenn Elliott	12 years
2012.3	commit c158b5fbe4...	Jonathan Herman	13 years
2012.2	commit b53c479a0f...	Glenn Elliott	13 years
2012.1	commit 83b11ea1c6...	Bjoern B. Brandenburg	14 years
rtas12-mc-beta-exp	commit 8e236ee20f...	Christopher Kenna	14 years
2011.1	commit d11808b5c6...	Christopher Kenna	15 years
v2.6.37-rc4	commit e8a7e48bb2...	Linus Torvalds	15 years
v2.6.37-rc3	commit 3561d43fd2...	Linus Torvalds	15 years
v2.6.37-rc2	commit e53beacd23...	Linus Torvalds	15 years
v2.6.37-rc1	commit c8ddb2713c...	Linus Torvalds	15 years
v2.6.36	commit f6f94e2ab1...	Linus Torvalds	15 years
2010.2	commit 5c5456402d...	Bjoern B. Brandenburg	15 years
v2.6.36-rc8	commit cd07202cc8...	Linus Torvalds	15 years
v2.6.36-rc7	commit cb655d0f3d...	Linus Torvalds	15 years
v2.6.36-rc6	commit 899611ee7d...	Linus Torvalds	15 years
v2.6.36-rc5	commit b30a3f6257...	Linus Torvalds	15 years
v2.6.36-rc4	commit 49553c2ef8...	Linus Torvalds	15 years
v2.6.36-rc3	commit 2bfc96a127...	Linus Torvalds	15 years
v2.6.36-rc2	commit 76be97c1fc...	Linus Torvalds	15 years
v2.6.36-rc1	commit da5cabf80e...	Linus Torvalds	15 years
v2.6.35	commit 9fe6206f40...	Linus Torvalds	15 years
v2.6.35-rc6	commit b37fa16e78...	Linus Torvalds	15 years
v2.6.35-rc5	commit 1c5474a65b...	Linus Torvalds	15 years
v2.6.35-rc4	commit 815c4163b6...	Linus Torvalds	15 years
v2.6.35-rc3	commit 7e27d6e778...	Linus Torvalds	15 years
v2.6.35-rc2	commit e44a21b726...	Linus Torvalds	15 years
v2.6.35-rc1	commit 67a3e12b05...	Linus Torvalds	15 years
2010.1	commit 7c1ff4c544...	Andrea Bastoni	15 years
v2.6.34	commit e40152ee1e...	Linus Torvalds	15 years
v2.6.33.4	commit 4640b4e7d9...	Greg Kroah-Hartman	15 years
v2.6.34-rc7	commit b57f95a382...	Linus Torvalds	15 years
v2.6.34-rc6	commit 66f41d4c5c...	Linus Torvalds	15 years
v2.6.33.3	commit 3e7ad8ed97...	Greg Kroah-Hartman	15 years
v2.6.34-rc5	commit 01bf0b6457...	Linus Torvalds	15 years
v2.6.34-rc4	commit 0d0fb0f9c5...	Linus Torvalds	15 years
v2.6.33.2	commit 19f00f070c...	Greg Kroah-Hartman	15 years
v2.6.34-rc3	commit 2eaa9cfdf3...	Linus Torvalds	15 years
v2.6.34-rc2	commit 220bf991b0...	Linus Torvalds	15 years
v2.6.33.1	commit dbdafe5ccf...	Greg Kroah-Hartman	16 years
v2.6.34-rc1	commit 57d54889cd...	Linus Torvalds	16 years
v2.6.33	commit 60b341b778...	Linus Torvalds	16 years
v2.6.33-rc8	commit 724e6d3fe8...	Linus Torvalds	16 years
v2.6.33-rc7	commit 29275254ca...	Linus Torvalds	16 years
v2.6.33-rc6	commit abe94c756c...	Linus Torvalds	16 years
v2.6.33-rc5	commit 92dcffb916...	Linus Torvalds	16 years
v2.6.33-rc4	commit 7284ce6c9f...	Linus Torvalds	16 years
v2.6.33-rc3	commit 74d2e4f8d7...	Linus Torvalds	16 years
v2.6.33-rc2	commit 6b7b284958...	Linus Torvalds	16 years
v2.6.33-rc1	commit 55639353a0...	Linus Torvalds	16 years
v2.6.32	commit 22763c5cf3...	Linus Torvalds	16 years
v2.6.32-rc8	commit 648f4e3e50...	Linus Torvalds	16 years
v2.6.32-rc7	commit 156171c71a...	Linus Torvalds	16 years
v2.6.32-rc6	commit b419148e56...	Linus Torvalds	16 years
v2.6.32-rc5	commit 012abeea66...	Linus Torvalds	16 years
v2.6.32-rc4	commit 161291396e...	Linus Torvalds	16 years
v2.6.32-rc3	commit 374576a8b6...	Linus Torvalds	16 years
v2.6.32-rc1	commit 17d857be64...	Linus Torvalds	16 years
v2.6.32-rc2	commit 17d857be64...	Linus Torvalds	16 years
v2.6.31	commit 74fca6a428...	Linus Torvalds	16 years
v2.6.31-rc9	commit e07cccf404...	Linus Torvalds	16 years
v2.6.31-rc8	commit 326ba5010a...	Linus Torvalds	16 years
v2.6.31-rc7	commit 422bef879e...	Linus Torvalds	16 years
v2.6.31-rc6	commit 64f1607ffb...	Linus Torvalds	16 years
v2.6.31-rc5	commit ed680c4ad4...	Linus Torvalds	16 years
v2.6.31-rc4	commit 4be3bd7849...	Linus Torvalds	16 years
v2.6.31-rc3	commit 6847e154e3...	Linus Torvalds	16 years
v2.6.31-rc2	commit 8e4a718ff3...	Linus Torvalds	16 years
v2.6.31-rc1	commit 28d0325ce6...	Linus Torvalds	16 years
v2.6.30	commit 07a2039b8e...	Linus Torvalds	16 years
v2.6.30-rc8	commit 9fa7eb283c...	Linus Torvalds	16 years
v2.6.30-rc7	commit 59a3759d0f...	Linus Torvalds	16 years
v2.6.30-rc6	commit 1406de8e11...	Linus Torvalds	16 years
v2.6.30-rc5	commit 091bf7624d...	Linus Torvalds	16 years
v2.6.30-rc4	commit 091438dd56...	Linus Torvalds	16 years
v2.6.30-rc3	commit 0910697403...	Linus Torvalds	16 years
v2.6.30-rc2	commit 0882e8dd3a...	Linus Torvalds	16 years
v2.6.30-rc1	commit 577c9c456f...	Linus Torvalds	16 years
v2.6.29	commit 8e0ee43bc2...	Linus Torvalds	16 years
v2.6.29-rc8	commit 041b62374c...	Linus Torvalds	17 years
v2.6.29-rc7	commit fec6c6fec3...	Linus Torvalds	17 years
v2.6.29-rc6	commit 20f4d6c3a2...	Linus Torvalds	17 years
v2.6.29-rc5	commit d2f8d7ee1a...	Linus Torvalds	17 years
v2.6.29-rc4	commit 8e4921515c...	Linus Torvalds	17 years
v2.6.29-rc3	commit 18e352e4a7...	Linus Torvalds	17 years
v2.6.29-rc2	commit 1de9e8e70f...	Linus Torvalds	17 years
v2.6.29-rc1	commit c59765042f...	Linus Torvalds	17 years
v2.6.28	commit 4a6908a3a0...	Linus Torvalds	17 years
v2.6.28-rc9	commit 929096fe9f...	Linus Torvalds	17 years
v2.6.28-rc8	commit 8b1fae4e42...	Linus Torvalds	17 years
v2.6.28-rc7	commit 061e41fdb5...	Linus Torvalds	17 years
v2.6.28-rc6	commit 13d428afc0...	Linus Torvalds	17 years
v2.6.28-rc5	commit 9bf1a2445f...	Linus Torvalds	17 years
v2.6.28-rc4	commit f7160c7573...	Linus Torvalds	17 years
v2.6.28-rc3	commit 45beca08dd...	Linus Torvalds	17 years
v2.6.28-rc2	commit 0173a3265b...	Linus Torvalds	17 years
v2.6.28-rc1	commit 57f8f7b60d...	Linus Torvalds	17 years
v2.6.27	commit 3fa8749e58...	Linus Torvalds	17 years
v2.6.27-rc9	commit 4330ed8ed4...	Linus Torvalds	17 years
v2.6.27-rc8	commit 94aca1dac6...	Linus Torvalds	17 years
v2.6.27-rc7	commit 72d31053f6...	Linus Torvalds	17 years
v2.6.27-rc6	commit adee14b2e1...	Linus Torvalds	17 years
v2.6.27-rc5	commit 24342c34a0...	Linus Torvalds	17 years
v2.6.27-rc4	commit 6a55617ed5...	Linus Torvalds	17 years
v2.6.27-rc3	commit 30a2f3c60a...	Linus Torvalds	17 years
v2.6.27-rc2	commit 0967d61ea0...	Linus Torvalds	17 years
v2.6.27-rc1	commit 6e86841d05...	Linus Torvalds	17 years
v2.6.26	commit bce7f793da...	Linus Torvalds	17 years
v2.6.26-rc9	commit b7279469d6...	Linus Torvalds	17 years
v2.6.26-rc8	commit 543cf4cb3f...	Linus Torvalds	17 years
v2.6.26-rc7	commit d70ac829b7...	Linus Torvalds	17 years
v2.6.26-rc6	commit 5dd34572ad...	Linus Torvalds	17 years
v2.6.26-rc5	commit 53c8ba9540...	Linus Torvalds	17 years
v2.6.26-rc4	commit e490517a03...	Linus Torvalds	17 years
v2.6.26-rc3	commit b8291ad07a...	Linus Torvalds	17 years
v2.6.26-rc2	commit 492c2e476e...	Linus Torvalds	17 years
v2.6.26-rc1	commit 2ddcca36c8...	Linus Torvalds	17 years
v2.6.25	commit 4b119e21d0...	Linus Torvalds	17 years
v2.6.25-rc9	commit 120dd64cac...	Linus Torvalds	17 years
v2.6.25-rc8	commit 0e81a8ae37...	Linus Torvalds	17 years
v2.6.25-rc7	commit 05dda977f2...	Linus Torvalds	17 years
v2.6.25-rc6	commit a978b30af3...	Linus Torvalds	17 years
v2.6.25-rc5	commit cdeeeae056...	Linus Torvalds	18 years
v2.6.25-rc4	commit 29e8c3c304...	Linus Torvalds	18 years
v2.6.25-rc3	commit bfa274e243...	Linus Torvalds	18 years
v2.6.25-rc2	commit 101142c37b...	Linus Torvalds	18 years
v2.6.25-rc1	commit 19af35546d...	Linus Torvalds	18 years
v2.6.24	commit 49914084e7...	Linus Torvalds	18 years
v2.6.24-rc8	commit cbd9c88369...	Linus Torvalds	18 years
v2.6.24-rc7	commit 3ce5445046...	Linus Torvalds	18 years
v2.6.24-rc6	commit ea67db4cdb...	Linus Torvalds	18 years
v2.6.24-rc5	commit 82d29bf6dc...	Linus Torvalds	18 years
v2.6.24-rc4	commit 09b56adc98...	Linus Torvalds	18 years
v2.6.24-rc3	commit d9f8bcbf67...	Linus Torvalds	18 years
v2.6.24-rc2	commit dbeeb816e8...	Linus Torvalds	18 years
v2.6.24-rc1	commit c9927c2bf4...	Linus Torvalds	18 years
v2.6.23	commit bbf25010f1...	Linus Torvalds	18 years
v2.6.23-rc9	commit 3146b39c18...	Linus Torvalds	18 years
v2.6.23-rc8	commit 4942de4a0e...	Linus Torvalds	18 years
v2.6.23-rc7	commit 81cfe79b9c...	Linus Torvalds	18 years
v2.6.23-rc6	commit 0d4cbb5e7f...	Linus Torvalds	18 years
v2.6.23-rc5	commit 40ffbfad6b...	Linus Torvalds	18 years
v2.6.23-rc4	commit b07d68b5ca...	Linus Torvalds	18 years
v2.6.23-rc3	commit 39d3520c92...	Linus Torvalds	18 years
v2.6.23-rc2	commit d4ac2477fa...	Linus Torvalds	18 years
v2.6.23-rc1	commit f695baf2df...	Linus Torvalds	18 years
v2.6.22	commit 7dcca30a32...	Linus Torvalds	18 years
v2.6.22-rc7	commit a38d6181ff...	Linus Torvalds	18 years
v2.6.22-rc6	commit 189548642c...	Linus Torvalds	18 years
v2.6.22-rc5	commit 188e1f81ba...	Linus Torvalds	18 years
v2.6.22-rc4	commit 5ecd3100e6...	Linus Torvalds	18 years
v2.6.22-rc3	commit c420bc9f09...	Linus Torvalds	18 years
v2.6.22-rc2	commit 55b637c6a0...	Linus Torvalds	18 years
v2.6.22-rc1	commit 39403865d2...	Linus Torvalds	18 years
v2.6.21	commit de46c33745...	Linus Torvalds	18 years
v2.6.21-rc7	commit 94a05509a9...	Linus Torvalds	18 years
v2.6.21-rc6	commit a21bd69e15...	Linus Torvalds	18 years
v2.6.21-rc5	commit e0f2e3a06b...	Linus Torvalds	18 years
v2.6.21-rc4	commit db98e0b434...	Linus Torvalds	19 years
v2.6.21-rc3	commit 08e15e81a4...	Linus Torvalds	19 years
v2.6.21-rc2	commit 606135a308...	Linus Torvalds	19 years
v2.6.21-rc1	commit c8f71b01a5...	Linus Torvalds	19 years
v2.6.20	commit 62d0cfcb27...	Linus Torvalds	19 years
v2.6.20-rc7	commit f56df2f4db...	Linus Torvalds	19 years
v2.6.20-rc6	commit 99abfeafb5...	Linus Torvalds	19 years
v2.6.20-rc5	commit a8b3485287...	Linus Torvalds	19 years
v2.6.20-rc4	commit bf81b46482...	Linus Torvalds	19 years
v2.6.20-rc3	commit 669df1b478...	Linus Torvalds	19 years
v2.6.20-rc2	commit 3bf8ba38f3...	Linus Torvalds	19 years
v2.6.20-rc1	commit cc016448b0...	Linus Torvalds	19 years
v2.6.19	commit 0215ffb08c...	Linus Torvalds	19 years
v2.6.19-rc6	commit 44597f65f6...	Linus Torvalds	19 years
v2.6.19-rc5	commit 80c2188127...	Linus Torvalds	19 years
v2.6.19-rc4	commit ae99a78af3...	Linus Torvalds	19 years
v2.6.19-rc3	commit 7059abedd2...	Linus Torvalds	19 years
v2.6.19-rc2	commit b4bd8c6643...	Linus Torvalds	19 years
v2.6.19-rc1	commit d223a60106...	Linus Torvalds	19 years
v2.6.18	commit e478bec0ba...	Linus Torvalds	19 years
v2.6.18-rc7	commit 95064a75eb...	Linus Torvalds	19 years
v2.6.18-rc6	commit c336923b66...	Linus Torvalds	19 years
v2.6.18-rc5	commit 60d4684068...	Linus Torvalds	19 years
v2.6.18-rc4	commit 9f737633e6...	Linus Torvalds	19 years
v2.6.18-rc3	commit b6ff50833a...	Linus Torvalds	19 years
v2.6.18-rc2	commit 82d6897fef...	Linus Torvalds	19 years
v2.6.18-rc1	commit 120bda20c6...	Linus Torvalds	19 years
v2.6.17	commit 427abfa28a...	Linus Torvalds	19 years
v2.6.17-rc6	commit 1def630a6a...	Linus Torvalds	19 years
v2.6.17-rc5	commit a8bd60705a...	Linus Torvalds	19 years
v2.6.17-rc4	commit d8c3291c73...	Linus Torvalds	19 years
v2.6.17-rc3	commit 2be4d50295...	Linus Torvalds	19 years
v2.6.17-rc2	commit 8bbde0e6d5...	Linus Torvalds	19 years
v2.6.17-rc1	commit 6246b6128b...	Linus Torvalds	19 years
v2.6.16	commit 7705a8792b...	Linus Torvalds	19 years
v2.6.16-rc6	commit 535744878e...	Linus Torvalds	20 years
v2.6.16-rc5	commit b9a33cebac...	Linus Torvalds	20 years
v2.6.16-rc4	commit bd71c2b174...	Linus Torvalds	20 years
v2.6.16-rc3	commit e9bb4c9929...	Linus Torvalds	20 years
v2.6.16-rc2	commit 826eeb53a6...	Linus Torvalds	20 years
v2.6.16-rc1	commit 2664b25051...	Linus Torvalds	20 years
v2.6.15	commit 88026842b0...	Linus Torvalds	20 years
v2.6.15-rc7	commit f89f5948fc...	Linus Torvalds	20 years
v2.6.15-rc6	commit df7addbb45...	Linus Torvalds	20 years
v2.6.15-rc5	commit 436b0f76f2...	Linus Torvalds	20 years
v2.6.15-rc4	commit 5666c0947e...	Linus Torvalds	20 years
v2.6.15-rc3	commit 624f54be20...	Linus Torvalds	20 years
v2.6.15-rc2	commit 3bedff1d73...	Linus Torvalds	20 years
v2.6.15-rc1	commit cd52d1ee9a...	Linus Torvalds	20 years
v2.6.14	commit 741b2252a5...	Linus Torvalds	20 years
v2.6.14-rc5	commit 93918e9afc...	Linus Torvalds	20 years
v2.6.14-rc4	commit 907a426179...	Linus Torvalds	20 years
v2.6.14-rc3	commit 1c9426e8a5...	Linus Torvalds	20 years
v2.6.14-rc2	commit 676d55ae30...	Linus Torvalds	20 years
v2.6.14-rc1	commit 2f4ba45a75...	Linus Torvalds	20 years
v2.6.13	commit 02b3e4e2d7...	Linus Torvalds	20 years
v2.6.13-rc7	commit 0572e3da3f...	Linus Torvalds	20 years
v2.6.13-rc6	commit 6fc32179de...	Linus Torvalds	20 years
v2.6.13-rc5	commit 9a351e30d7...	Linus Torvalds	20 years
v2.6.13-rc4	commit 6395352334...	Linus Torvalds	20 years
v2.6.11	tree c39ae07f39...
v2.6.11-tree	tree c39ae07f39...
v2.6.12	commit 9ee1c939d1...
v2.6.12-rc2	commit 1da177e4c3...
v2.6.12-rc3	commit a2755a80f4...
v2.6.12-rc4	commit 88d7bd8cb9...
v2.6.12-rc5	commit 2a24ab628a...
v2.6.12-rc6	commit 7cef5677ef...
v2.6.13-rc1	commit 4c91aedb75...
v2.6.13-rc2	commit a18bcb7450...
v2.6.13-rc3	commit c32511e271...

generated by cgit v1.2.2 (git 2.25.0) at 2025-09-10 06:47:17 -0400

/*
** Write ahead logging implementation copyright Chris Mason 2000
**
** The background commits make this code very interrelated, and
** overly complex.  I need to rethink things a bit....The major players:
**
** journal_begin -- call with the number of blocks you expect to log.
**                  If the current transaction is too
** 		    old, it will block until the current transaction is
** 		    finished, and then start a new one.
**		    Usually, your transaction will get joined in with
**                  previous ones for speed.
**
** journal_join  -- same as journal_begin, but won't block on the current
**                  transaction regardless of age.  Don't ever call
**                  this.  Ever.  There are only two places it should be
**                  called from, and they are both inside this file.
**
** journal_mark_dirty -- adds blocks into this transaction.  clears any flags
**                       that might make them get sent to disk
**                       and then marks them BH_JDirty.  Puts the buffer head
**                       into the current transaction hash.
**
** journal_end -- if the current transaction is batchable, it does nothing
**                   otherwise, it could do an async/synchronous commit, or
**                   a full flush of all log and real blocks in the
**                   transaction.
**
** flush_old_commits -- if the current transaction is too old, it is ended and
**                      commit blocks are sent to disk.  Forces commit blocks
**                      to disk for all backgrounded commits that have been
**                      around too long.
**		     -- Note, if you call this as an immediate flush from
**		        from within kupdate, it will ignore the immediate flag
*/

#include <linux/time.h>
#include <linux/semaphore.h>
#include <linux/vmalloc.h>
#include <linux/reiserfs_fs.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/fcntl.h>
#include <linux/stat.h>
#include <linux/string.h>
#include <linux/buffer_head.h>
#include <linux/workqueue.h>
#include <linux/writeback.h>
#include <linux/blkdev.h>
#include <linux/backing-dev.h>
#include <linux/uaccess.h>
#include <linux/slab.h>

#include <asm/system.h>

/* gets a struct reiserfs_journal_list * from a list head */
#define JOURNAL_LIST_ENTRY(h) (list_entry((h), struct reiserfs_journal_list, \
                               j_list))
#define JOURNAL_WORK_ENTRY(h) (list_entry((h), struct reiserfs_journal_list, \
                               j_working_list))

/* the number of mounted filesystems.  This is used to decide when to
** start and kill the commit workqueue
*/
static int reiserfs_mounted_fs_count;

static struct workqueue_struct *commit_wq;

#define JOURNAL_TRANS_HALF 1018	/* must be correct to keep the desc and commit
				   structs at 4k */
#define BUFNR 64		/*read ahead */

/* cnode stat bits.  Move these into reiserfs_fs.h */

#define BLOCK_FREED 2		/* this block was freed, and can't be written.  */
#define BLOCK_FREED_HOLDER 3	/* this block was freed during this transaction, and can't be written */

#define BLOCK_NEEDS_FLUSH 4	/* used in flush_journal_list */
#define BLOCK_DIRTIED 5

/* journal list state bits */
#define LIST_TOUCHED 1
#define LIST_DIRTY   2
#define LIST_COMMIT_PENDING  4	/* someone will commit this list */

/* flags for do_journal_end */
#define FLUSH_ALL   1		/* flush commit and real blocks */
#define COMMIT_NOW  2		/* end and commit this transaction */
#define WAIT        4		/* wait for the log blocks to hit the disk */

static int do_journal_end(struct reiserfs_transaction_handle *,
			  struct super_block *, unsigned long nblocks,
			  int flags);
static int flush_journal_list(struct super_block *s,
			      struct reiserfs_journal_list *jl, int flushall);
static int flush_commit_list(struct super_block *s,
			     struct reiserfs_journal_list *jl, int flushall);
static int can_dirty(struct reiserfs_journal_cnode *cn);
static int journal_join(struct reiserfs_transaction_handle *th,
			struct super_block *sb, unsigned long nblocks);
static int release_journal_dev(struct super_block *super,
			       struct reiserfs_journal *journal);
static int dirty_one_transaction(struct super_block *s,
				 struct reiserfs_journal_list *jl);
static void flush_async_commits(struct work_struct *work);
static void queue_log_writer(struct super_block *s);

/* values for join in do_journal_begin_r */
enum {
	JBEGIN_REG = 0,		/* regular journal begin */
	JBEGIN_JOIN = 1,	/* join the running transaction if at all possible */
	JBEGIN_ABORT = 2,	/* called from cleanup code, ignores aborted flag */
};

static int do_journal_begin_r(struct reiserfs_transaction_handle *th,
			      struct super_block *sb,
			      unsigned long nblocks, int join);

static void init_journal_hash(struct super_block *sb)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	memset(journal->j_hash_table, 0,
	       JOURNAL_HASH_SIZE * sizeof(struct reiserfs_journal_cnode *));
}

/*
** clears BH_Dirty and sticks the buffer on the clean list.  Called because I can't allow refile_buffer to
** make schedule happen after I've freed a block.  Look at remove_from_transaction and journal_mark_freed for
** more details.
*/
static int reiserfs_clean_and_file_buffer(struct buffer_head *bh)
{
	if (bh) {
		clear_buffer_dirty(bh);
		clear_buffer_journal_test(bh);
	}
	return 0;
}

static struct reiserfs_bitmap_node *allocate_bitmap_node(struct super_block
							 *sb)
{
	struct reiserfs_bitmap_node *bn;
	static int id;

	bn = kmalloc(sizeof(struct reiserfs_bitmap_node), GFP_NOFS);
	if (!bn) {
		return NULL;
	}
	bn->data = kzalloc(sb->s_blocksize, GFP_NOFS);
	if (!bn->data) {
		kfree(bn);
		return NULL;
	}
	bn->id = id++;
	INIT_LIST_HEAD(&bn->list);
	return bn;
}

static struct reiserfs_bitmap_node *get_bitmap_node(struct super_block *sb)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct reiserfs_bitmap_node *bn = NULL;
	struct list_head *entry = journal->j_bitmap_nodes.next;

	journal->j_used_bitmap_nodes++;
      repeat:

	if (entry != &journal->j_bitmap_nodes) {
		bn = list_entry(entry, struct reiserfs_bitmap_node, list);
		list_del(entry);
		memset(bn->data, 0, sb->s_blocksize);
		journal->j_free_bitmap_nodes--;
		return bn;
	}
	bn = allocate_bitmap_node(sb);
	if (!bn) {
		yield();
		goto repeat;
	}
	return bn;
}
static inline void free_bitmap_node(struct super_block *sb,
				    struct reiserfs_bitmap_node *bn)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	journal->j_used_bitmap_nodes--;
	if (journal->j_free_bitmap_nodes > REISERFS_MAX_BITMAP_NODES) {
		kfree(bn->data);
		kfree(bn);
	} else {
		list_add(&bn->list, &journal->j_bitmap_nodes);
		journal->j_free_bitmap_nodes++;
	}
}

static void allocate_bitmap_nodes(struct super_block *sb)
{
	int i;
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct reiserfs_bitmap_node *bn = NULL;
	for (i = 0; i < REISERFS_MIN_BITMAP_NODES; i++) {
		bn = allocate_bitmap_node(sb);
		if (bn) {
			list_add(&bn->list, &journal->j_bitmap_nodes);
			journal->j_free_bitmap_nodes++;
		} else {
			break;	/* this is ok, we'll try again when more are needed */
		}
	}
}

static int set_bit_in_list_bitmap(struct super_block *sb,
				  b_blocknr_t block,
				  struct reiserfs_list_bitmap *jb)
{
	unsigned int bmap_nr = block / (sb->s_blocksize << 3);
	unsigned int bit_nr = block % (sb->s_blocksize << 3);

	if (!jb->bitmaps[bmap_nr]) {
		jb->bitmaps[bmap_nr] = get_bitmap_node(sb);
	}
	set_bit(bit_nr, (unsigned long *)jb->bitmaps[bmap_nr]->data);
	return 0;
}

static void cleanup_bitmap_list(struct super_block *sb,
				struct reiserfs_list_bitmap *jb)
{
	int i;
	if (jb->bitmaps == NULL)
		return;

	for (i = 0; i < reiserfs_bmap_count(sb); i++) {
		if (jb->bitmaps[i]) {
			free_bitmap_node(sb, jb->bitmaps[i]);
			jb->bitmaps[i] = NULL;
		}
	}
}

/*
** only call this on FS unmount.
*/
static int free_list_bitmaps(struct super_block *sb,
			     struct reiserfs_list_bitmap *jb_array)
{
	int i;
	struct reiserfs_list_bitmap *jb;
	for (i = 0; i < JOURNAL_NUM_BITMAPS; i++) {
		jb = jb_array + i;
		jb->journal_list = NULL;
		cleanup_bitmap_list(sb, jb);
		vfree(jb->bitmaps);
		jb->bitmaps = NULL;
	}
	return 0;
}

static int free_bitmap_nodes(struct super_block *sb)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct list_head *next = journal->j_bitmap_nodes.next;
	struct reiserfs_bitmap_node *bn;

	while (next != &journal->j_bitmap_nodes) {
		bn = list_entry(next, struct reiserfs_bitmap_node, list);
		list_del(next);
		kfree(bn->data);
		kfree(bn);
		next = journal->j_bitmap_nodes.next;
		journal->j_free_bitmap_nodes--;
	}

	return 0;
}

/*
** get memory for JOURNAL_NUM_BITMAPS worth of bitmaps.
** jb_array is the array to be filled in.
*/
int reiserfs_allocate_list_bitmaps(struct super_block *sb,
				   struct reiserfs_list_bitmap *jb_array,
				   unsigned int bmap_nr)
{
	int i;
	int failed = 0;
	struct reiserfs_list_bitmap *jb;
	int mem = bmap_nr * sizeof(struct reiserfs_bitmap_node *);

	for (i = 0; i < JOURNAL_NUM_BITMAPS; i++) {
		jb = jb_array + i;
		jb->journal_list = NULL;
		jb->bitmaps = vzalloc(mem);
		if (!jb->bitmaps) {
			reiserfs_warning(sb, "clm-2000", "unable to "
					 "allocate bitmaps for journal lists");
			failed = 1;
			break;
		}
	}
	if (failed) {
		free_list_bitmaps(sb, jb_array);
		return -1;
	}
	return 0;
}

/*
** find an available list bitmap.  If you can't find one, flush a commit list
** and try again
*/
static struct reiserfs_list_bitmap *get_list_bitmap(struct super_block *sb,
						    struct reiserfs_journal_list
						    *jl)
{
	int i, j;
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct reiserfs_list_bitmap *jb = NULL;

	for (j = 0; j < (JOURNAL_NUM_BITMAPS * 3); j++) {
		i = journal->j_list_bitmap_index;
		journal->j_list_bitmap_index = (i + 1) % JOURNAL_NUM_BITMAPS;
		jb = journal->j_list_bitmap + i;
		if (journal->j_list_bitmap[i].journal_list) {
			flush_commit_list(sb,
					  journal->j_list_bitmap[i].
					  journal_list, 1);
			if (!journal->j_list_bitmap[i].journal_list) {
				break;
			}
		} else {
			break;
		}
	}
	if (jb->journal_list) {	/* double check to make sure if flushed correctly */
		return NULL;
	}
	jb->journal_list = jl;
	return jb;
}

/*
** allocates a new chunk of X nodes, and links them all together as a list.
** Uses the cnode->next and cnode->prev pointers
** returns NULL on failure
*/
static struct reiserfs_journal_cnode *allocate_cnodes(int num_cnodes)
{
	struct reiserfs_journal_cnode *head;
	int i;
	if (num_cnodes <= 0) {
		return NULL;
	}
	head = vzalloc(num_cnodes * sizeof(struct reiserfs_journal_cnode));
	if (!head) {
		return NULL;
	}
	head[0].prev = NULL;
	head[0].next = head + 1;
	for (i = 1; i < num_cnodes; i++) {
		head[i].prev = head + (i - 1);
		head[i].next = head + (i + 1);	/* if last one, overwrite it after the if */
	}
	head[num_cnodes - 1].next = NULL;
	return head;
}

/*
** pulls a cnode off the free list, or returns NULL on failure
*/
static struct reiserfs_journal_cnode *get_cnode(struct super_block *sb)
{
	struct reiserfs_journal_cnode *cn;
	struct reiserfs_journal *journal = SB_JOURNAL(sb);

	reiserfs_check_lock_depth(sb, "get_cnode");

	if (journal->j_cnode_free <= 0) {
		return NULL;
	}
	journal->j_cnode_used++;
	journal->j_cnode_free--;
	cn = journal->j_cnode_free_list;
	if (!cn) {
		return cn;
	}
	if (cn->next) {
		cn->next->prev = NULL;
	}
	journal->j_cnode_free_list = cn->next;
	memset(cn, 0, sizeof(struct reiserfs_journal_cnode));
	return cn;
}

/*
** returns a cnode to the free list
*/
static void free_cnode(struct super_block *sb,
		       struct reiserfs_journal_cnode *cn)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);

	reiserfs_check_lock_depth(sb, "free_cnode");

	journal->j_cnode_used--;
	journal->j_cnode_free++;
	/* memset(cn, 0, sizeof(struct reiserfs_journal_cnode)) ; */
	cn->next = journal->j_cnode_free_list;
	if (journal->j_cnode_free_list) {
		journal->j_cnode_free_list->prev = cn;
	}
	cn->prev = NULL;	/* not needed with the memset, but I might kill the memset, and forget to do this */
	journal->j_cnode_free_list = cn;
}

static void clear_prepared_bits(struct buffer_head *bh)
{
	clear_buffer_journal_prepared(bh);
	clear_buffer_journal_restore_dirty(bh);
}

/* return a cnode with same dev, block number and size in table, or null if not found */
static inline struct reiserfs_journal_cnode *get_journal_hash_dev(struct
								  super_block
								  *sb,
								  struct
								  reiserfs_journal_cnode
								  **table,
								  long bl)
{
	struct reiserfs_journal_cnode *cn;
	cn = journal_hash(table, sb, bl);
	while (cn) {
		if (cn->blocknr == bl && cn->sb == sb)
			return cn;
		cn = cn->hnext;
	}
	return (struct reiserfs_journal_cnode *)0;
}

/*
** this actually means 'can this block be reallocated yet?'.  If you set search_all, a block can only be allocated
** if it is not in the current transaction, was not freed by the current transaction, and has no chance of ever
** being overwritten by a replay after crashing.
**
** If you don't set search_all, a block can only be allocated if it is not in the current transaction.  Since deleting
** a block removes it from the current transaction, this case should never happen.  If you don't set search_all, make
** sure you never write the block without logging it.
**
** next_zero_bit is a suggestion about the next block to try for find_forward.
** when bl is rejected because it is set in a journal list bitmap, we search
** for the next zero bit in the bitmap that rejected bl.  Then, we return that
** through next_zero_bit for find_forward to try.
**
** Just because we return something in next_zero_bit does not mean we won't
** reject it on the next call to reiserfs_in_journal
**
*/
int reiserfs_in_journal(struct super_block *sb,
			unsigned int bmap_nr, int bit_nr, int search_all,
			b_blocknr_t * next_zero_bit)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct reiserfs_journal_cnode *cn;
	struct reiserfs_list_bitmap *jb;
	int i;
	unsigned long bl;

	*next_zero_bit = 0;	/* always start this at zero. */

	PROC_INFO_INC(sb, journal.in_journal);
	/* If we aren't doing a search_all, this is a metablock, and it will be logged before use.
	 ** if we crash before the transaction that freed it commits,  this transaction won't
	 ** have committed either, and the block will never be written
	 */
	if (search_all) {
		for (i = 0; i < JOURNAL_NUM_BITMAPS; i++) {
			PROC_INFO_INC(sb, journal.in_journal_bitmap);
			jb = journal->j_list_bitmap + i;
			if (jb->journal_list && jb->bitmaps[bmap_nr] &&
			    test_bit(bit_nr,
				     (unsigned long *)jb->bitmaps[bmap_nr]->
				     data)) {
				*next_zero_bit =
				    find_next_zero_bit((unsigned long *)
						       (jb->bitmaps[bmap_nr]->
							data),
						       sb->s_blocksize << 3,
						       bit_nr + 1);
				return 1;
			}
		}
	}

	bl = bmap_nr * (sb->s_blocksize << 3) + bit_nr;
	/* is it in any old transactions? */
	if (search_all
	    && (cn =
		get_journal_hash_dev(sb, journal->j_list_hash_table, bl))) {
		return 1;
	}

	/* is it in the current transaction.  This should never happen */
	if ((cn = get_journal_hash_dev(sb, journal->j_hash_table, bl))) {
		BUG();
		return 1;
	}

	PROC_INFO_INC(sb, journal.in_journal_reusable);
	/* safe for reuse */
	return 0;
}

/* insert cn into table
*/
static inline void insert_journal_hash(struct reiserfs_journal_cnode **table,
				       struct reiserfs_journal_cnode *cn)
{
	struct reiserfs_journal_cnode *cn_orig;

	cn_orig = journal_hash(table, cn->sb, cn->blocknr);
	cn->hnext = cn_orig;
	cn->hprev = NULL;
	if (cn_orig) {
		cn_orig->hprev = cn;
	}
	journal_hash(table, cn->sb, cn->blocknr) = cn;
}

/* lock the current transaction */
static inline void lock_journal(struct super_block *sb)
{
	PROC_INFO_INC(sb, journal.lock_journal);

	reiserfs_mutex_lock_safe(&SB_JOURNAL(sb)->j_mutex, sb);
}

/* unlock the current transaction */
static inline void unlock_journal(struct super_block *sb)
{
	mutex_unlock(&SB_JOURNAL(sb)->j_mutex);
}

static inline void get_journal_list(struct reiserfs_journal_list *jl)
{
	jl->j_refcount++;
}

static inline void put_journal_list(struct super_block *s,
				    struct reiserfs_journal_list *jl)
{
	if (jl->j_refcount < 1) {
		reiserfs_panic(s, "journal-2", "trans id %u, refcount at %d",
			       jl->j_trans_id, jl->j_refcount);
	}
	if (--jl->j_refcount == 0)
		kfree(jl);
}

/*
** this used to be much more involved, and I'm keeping it just in case things get ugly again.
** it gets called by flush_commit_list, and cleans up any data stored about blocks freed during a
** transaction.
*/
static void cleanup_freed_for_journal_list(struct super_block *sb,
					   struct reiserfs_journal_list *jl)
{

	struct reiserfs_list_bitmap *jb = jl->j_list_bitmap;
	if (jb) {
		cleanup_bitmap_list(sb, jb);
	}
	jl->j_list_bitmap->journal_list = NULL;
	jl->j_list_bitmap = NULL;
}

static int journal_list_still_alive(struct super_block *s,
				    unsigned int trans_id)
{
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	struct list_head *entry = &journal->j_journal_list;
	struct reiserfs_journal_list *jl;

	if (!list_empty(entry)) {
		jl = JOURNAL_LIST_ENTRY(entry->next);
		if (jl->j_trans_id <= trans_id) {
			return 1;
		}
	}
	return 0;
}

/*
 * If page->mapping was null, we failed to truncate this page for
 * some reason.  Most likely because it was truncated after being
 * logged via data=journal.
 *
 * This does a check to see if the buffer belongs to one of these
 * lost pages before doing the final put_bh.  If page->mapping was
 * null, it tries to free buffers on the page, which should make the
 * final page_cache_release drop the page from the lru.
 */
static void release_buffer_page(struct buffer_head *bh)
{
	struct page *page = bh->b_page;
	if (!page->mapping && trylock_page(page)) {
		page_cache_get(page);
		put_bh(bh);
		if (!page->mapping)
			try_to_free_buffers(page);
		unlock_page(page);
		page_cache_release(page);
	} else {
		put_bh(bh);
	}
}

static void reiserfs_end_buffer_io_sync(struct buffer_head *bh, int uptodate)
{
	char b[BDEVNAME_SIZE];

	if (buffer_journaled(bh)) {
		reiserfs_warning(NULL, "clm-2084",
				 "pinned buffer %lu:%s sent to disk",
				 bh->b_blocknr, bdevname(bh->b_bdev, b));
	}
	if (uptodate)
		set_buffer_uptodate(bh);
	else
		clear_buffer_uptodate(bh);

	unlock_buffer(bh);
	release_buffer_page(bh);
}

static void reiserfs_end_ordered_io(struct buffer_head *bh, int uptodate)
{
	if (uptodate)
		set_buffer_uptodate(bh);
	else
		clear_buffer_uptodate(bh);
	unlock_buffer(bh);
	put_bh(bh);
}

static void submit_logged_buffer(struct buffer_head *bh)
{
	get_bh(bh);
	bh->b_end_io = reiserfs_end_buffer_io_sync;
	clear_buffer_journal_new(bh);
	clear_buffer_dirty(bh);
	if (!test_clear_buffer_journal_test(bh))
		BUG();
	if (!buffer_uptodate(bh))
		BUG();
	submit_bh(WRITE, bh);
}

static void submit_ordered_buffer(struct buffer_head *bh)
{
	get_bh(bh);
	bh->b_end_io = reiserfs_end_ordered_io;
	clear_buffer_dirty(bh);
	if (!buffer_uptodate(bh))
		BUG();
	submit_bh(WRITE, bh);
}

#define CHUNK_SIZE 32
struct buffer_chunk {
	struct buffer_head *bh[CHUNK_SIZE];
	int nr;
};

static void write_chunk(struct buffer_chunk *chunk)
{
	int i;
	for (i = 0; i < chunk->nr; i++) {
		submit_logged_buffer(chunk->bh[i]);
	}
	chunk->nr = 0;
}

static void write_ordered_chunk(struct buffer_chunk *chunk)
{
	int i;
	for (i = 0; i < chunk->nr; i++) {
		submit_ordered_buffer(chunk->bh[i]);
	}
	chunk->nr = 0;
}

static int add_to_chunk(struct buffer_chunk *chunk, struct buffer_head *bh,
			spinlock_t * lock, void (fn) (struct buffer_chunk *))
{
	int ret = 0;
	BUG_ON(chunk->nr >= CHUNK_SIZE);
	chunk->bh[chunk->nr++] = bh;
	if (chunk->nr >= CHUNK_SIZE) {
		ret = 1;
		if (lock)
			spin_unlock(lock);
		fn(chunk);
		if (lock)
			spin_lock(lock);
	}
	return ret;
}

static atomic_t nr_reiserfs_jh = ATOMIC_INIT(0);
static struct reiserfs_jh *alloc_jh(void)
{
	struct reiserfs_jh *jh;
	while (1) {
		jh = kmalloc(sizeof(*jh), GFP_NOFS);
		if (jh) {
			atomic_inc(&nr_reiserfs_jh);
			return jh;
		}
		yield();
	}
}

/*
 * we want to free the jh when the buffer has been written
 * and waited on
 */
void reiserfs_free_jh(struct buffer_head *bh)
{
	struct reiserfs_jh *jh;

	jh = bh->b_private;
	if (jh) {
		bh->b_private = NULL;
		jh->bh = NULL;
		list_del_init(&jh->list);
		kfree(jh);
		if (atomic_read(&nr_reiserfs_jh) <= 0)
			BUG();
		atomic_dec(&nr_reiserfs_jh);
		put_bh(bh);
	}
}

static inline int __add_jh(struct reiserfs_journal *j, struct buffer_head *bh,
			   int tail)
{
	struct reiserfs_jh *jh;

	if (bh->b_private) {
		spin_lock(&j->j_dirty_buffers_lock);
		if (!bh->b_private) {
			spin_unlock(&j->j_dirty_buffers_lock);
			goto no_jh;
		}
		jh = bh->b_private;
		list_del_init(&jh->list);
	} else {
	      no_jh:
		get_bh(bh);
		jh = alloc_jh();
		spin_lock(&j->j_dirty_buffers_lock);
		/* buffer must be locked for __add_jh, should be able to have
		 * two adds at the same time
		 */
		BUG_ON(bh->b_private);
		jh->bh = bh;
		bh->b_private = jh;
	}
	jh->jl = j->j_current_jl;
	if (tail)
		list_add_tail(&jh->list, &jh->jl->j_tail_bh_list);
	else {
		list_add_tail(&jh->list, &jh->jl->j_bh_list);
	}
	spin_unlock(&j->j_dirty_buffers_lock);
	return 0;
}

int reiserfs_add_tail_list(struct inode *inode, struct buffer_head *bh)
{
	return __add_jh(SB_JOURNAL(inode->i_sb), bh, 1);
}
int reiserfs_add_ordered_list(struct inode *inode, struct buffer_head *bh)
{
	return __add_jh(SB_JOURNAL(inode->i_sb), bh, 0);
}

#define JH_ENTRY(l) list_entry((l), struct reiserfs_jh, list)
static int write_ordered_buffers(spinlock_t * lock,
				 struct reiserfs_journal *j,
				 struct reiserfs_journal_list *jl,
				 struct list_head *list)
{
	struct buffer_head *bh;
	struct reiserfs_jh *jh;
	int ret = j->j_errno;
	struct buffer_chunk chunk;
	struct list_head tmp;
	INIT_LIST_HEAD(&tmp);

	chunk.nr = 0;
	spin_lock(lock);
	while (!list_empty(list)) {
		jh = JH_ENTRY(list->next);
		bh = jh->bh;
		get_bh(bh);
		if (!trylock_buffer(bh)) {
			if (!buffer_dirty(bh)) {
				list_move(&jh->list, &tmp);
				goto loop_next;
			}
			spin_unlock(lock);
			if (chunk.nr)
				write_ordered_chunk(&chunk);
			wait_on_buffer(bh);
			cond_resched();
			spin_lock(lock);
			goto loop_next;
		}
		/* in theory, dirty non-uptodate buffers should never get here,
		 * but the upper layer io error paths still have a few quirks.
		 * Handle them here as gracefully as we can
		 */
		if (!buffer_uptodate(bh) && buffer_dirty(bh)) {
			clear_buffer_dirty(bh);
			ret = -EIO;
		}
		if (buffer_dirty(bh)) {
			list_move(&jh->list, &tmp);
			add_to_chunk(&chunk, bh, lock, write_ordered_chunk);
		} else {
			reiserfs_free_jh(bh);
			unlock_buffer(bh);
		}
	      loop_next:
		put_bh(bh);
		cond_resched_lock(lock);
	}
	if (chunk.nr) {
		spin_unlock(lock);
		write_ordered_chunk(&chunk);
		spin_lock(lock);
	}
	while (!list_empty(&tmp)) {
		jh = JH_ENTRY(tmp.prev);
		bh = jh->bh;
		get_bh(bh);
		reiserfs_free_jh(bh);

		if (buffer_locked(bh)) {
			spin_unlock(lock);
			wait_on_buffer(bh);
			spin_lock(lock);
		}
		if (!buffer_uptodate(bh)) {
			ret = -EIO;
		}
		/* ugly interaction with invalidatepage here.
		 * reiserfs_invalidate_page will pin any buffer that has a valid
		 * journal head from an older transaction.  If someone else sets
		 * our buffer dirty after we write it in the first loop, and
		 * then someone truncates the page away, nobody will ever write
		 * the buffer. We're safe if we write the page one last time
		 * after freeing the journal header.
		 */
		if (buffer_dirty(bh) && unlikely(bh->b_page->mapping == NULL)) {
			spin_unlock(lock);
			ll_rw_block(WRITE, 1, &bh);
			spin_lock(lock);
		}
		put_bh(bh);
		cond_resched_lock(lock);
	}
	spin_unlock(lock);
	return ret;
}

static int flush_older_commits(struct super_block *s,
			       struct reiserfs_journal_list *jl)
{
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	struct reiserfs_journal_list *other_jl;
	struct reiserfs_journal_list *first_jl;
	struct list_head *entry;
	unsigned int trans_id = jl->j_trans_id;
	unsigned int other_trans_id;
	unsigned int first_trans_id;

      find_first:
	/*
	 * first we walk backwards to find the oldest uncommitted transation
	 */
	first_jl = jl;
	entry = jl->j_list.prev;
	while (1) {
		other_jl = JOURNAL_LIST_ENTRY(entry);
		if (entry == &journal->j_journal_list ||
		    atomic_read(&other_jl->j_older_commits_done))
			break;

		first_jl = other_jl;
		entry = other_jl->j_list.prev;
	}

	/* if we didn't find any older uncommitted transactions, return now */
	if (first_jl == jl) {
		return 0;
	}

	first_trans_id = first_jl->j_trans_id;

	entry = &first_jl->j_list;
	while (1) {
		other_jl = JOURNAL_LIST_ENTRY(entry);
		other_trans_id = other_jl->j_trans_id;

		if (other_trans_id < trans_id) {
			if (atomic_read(&other_jl->j_commit_left) != 0) {
				flush_commit_list(s, other_jl, 0);

				/* list we were called with is gone, return */
				if (!journal_list_still_alive(s, trans_id))
					return 1;

				/* the one we just flushed is gone, this means all
				 * older lists are also gone, so first_jl is no longer
				 * valid either.  Go back to the beginning.
				 */
				if (!journal_list_still_alive
				    (s, other_trans_id)) {
					goto find_first;
				}
			}
			entry = entry->next;
			if (entry == &journal->j_journal_list)
				return 0;
		} else {
			return 0;
		}
	}
	return 0;
}

static int reiserfs_async_progress_wait(struct super_block *s)
{
	struct reiserfs_journal *j = SB_JOURNAL(s);

	if (atomic_read(&j->j_async_throttle)) {
		reiserfs_write_unlock(s);
		congestion_wait(BLK_RW_ASYNC, HZ / 10);
		reiserfs_write_lock(s);
	}

	return 0;
}

/*
** if this journal list still has commit blocks unflushed, send them to disk.
**
** log areas must be flushed in order (transaction 2 can't commit before transaction 1)
** Before the commit block can by written, every other log block must be safely on disk
**
*/
static int flush_commit_list(struct super_block *s,
			     struct reiserfs_journal_list *jl, int flushall)
{
	int i;
	b_blocknr_t bn;
	struct buffer_head *tbh = NULL;
	unsigned int trans_id = jl->j_trans_id;
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	int retval = 0;
	int write_len;

	reiserfs_check_lock_depth(s, "flush_commit_list");

	if (atomic_read(&jl->j_older_commits_done)) {
		return 0;
	}

	/* before we can put our commit blocks on disk, we have to make sure everyone older than
	 ** us is on disk too
	 */
	BUG_ON(jl->j_len <= 0);
	BUG_ON(trans_id == journal->j_trans_id);

	get_journal_list(jl);
	if (flushall) {
		if (flush_older_commits(s, jl) == 1) {
			/* list disappeared during flush_older_commits.  return */
			goto put_jl;
		}
	}

	/* make sure nobody is trying to flush this one at the same time */
	reiserfs_mutex_lock_safe(&jl->j_commit_mutex, s);

	if (!journal_list_still_alive(s, trans_id)) {
		mutex_unlock(&jl->j_commit_mutex);
		goto put_jl;
	}
	BUG_ON(jl->j_trans_id == 0);

	/* this commit is done, exit */
	if (atomic_read(&(jl->j_commit_left)) <= 0) {
		if (flushall) {
			atomic_set(&(jl->j_older_commits_done), 1);
		}
		mutex_unlock(&jl->j_commit_mutex);
		goto put_jl;
	}

	if (!list_empty(&jl->j_bh_list)) {
		int ret;

		/*
		 * We might sleep in numerous places inside
		 * write_ordered_buffers. Relax the write lock.
		 */
		reiserfs_write_unlock(s);
		ret = write_ordered_buffers(&journal->j_dirty_buffers_lock,
					    journal, jl, &jl->j_bh_list);
		if (ret < 0 && retval == 0)
			retval = ret;
		reiserfs_write_lock(s);
	}
	BUG_ON(!list_empty(&jl->j_bh_list));
	/*
	 * for the description block and all the log blocks, submit any buffers
	 * that haven't already reached the disk.  Try to write at least 256
	 * log blocks. later on, we will only wait on blocks that correspond
	 * to this transaction, but while we're unplugging we might as well
	 * get a chunk of data on there.
	 */
	atomic_inc(&journal->j_async_throttle);
	write_len = jl->j_len + 1;
	if (write_len < 256)
		write_len = 256;
	for (i = 0 ; i < write_len ; i++) {
		bn = SB_ONDISK_JOURNAL_1st_BLOCK(s) + (jl->j_start + i) %
		    SB_ONDISK_JOURNAL_SIZE(s);
		tbh = journal_find_get_block(s, bn);
		if (tbh) {
			if (buffer_dirty(tbh)) {
		            reiserfs_write_unlock(s);
			    ll_rw_block(WRITE, 1, &tbh);
			    reiserfs_write_lock(s);
			}
			put_bh(tbh) ;
		}
	}
	atomic_dec(&journal->j_async_throttle);

	for (i = 0; i < (jl->j_len + 1); i++) {
		bn = SB_ONDISK_JOURNAL_1st_BLOCK(s) +
		    (jl->j_start + i) % SB_ONDISK_JOURNAL_SIZE(s);
		tbh = journal_find_get_block(s, bn);

		reiserfs_write_unlock(s);
		wait_on_buffer(tbh);
		reiserfs_write_lock(s);
		// since we're using ll_rw_blk above, it might have skipped over
		// a locked buffer.  Double check here
		//
		/* redundant, sync_dirty_buffer() checks */
		if (buffer_dirty(tbh)) {
			reiserfs_write_unlock(s);
			sync_dirty_buffer(tbh);
			reiserfs_write_lock(s);
		}
		if (unlikely(!buffer_uptodate(tbh))) {
#ifdef CONFIG_REISERFS_CHECK
			reiserfs_warning(s, "journal-601",
					 "buffer write failed");
#endif
			retval = -EIO;
		}
		put_bh(tbh);	/* once for journal_find_get_block */
		put_bh(tbh);	/* once due to original getblk in do_journal_end */
		atomic_dec(&(jl->j_commit_left));
	}

	BUG_ON(atomic_read(&(jl->j_commit_left)) != 1);

	/* If there was a write error in the journal - we can't commit
	 * this transaction - it will be invalid and, if successful,
	 * will just end up propagating the write error out to
	 * the file system. */
	if (likely(!retval && !reiserfs_is_journal_aborted (journal))) {
		if (buffer_dirty(jl->j_commit_bh))
			BUG();
		mark_buffer_dirty(jl->j_commit_bh) ;
		reiserfs_write_unlock(s);
		if (reiserfs_barrier_flush(s))
			__sync_dirty_buffer(jl->j_commit_bh, WRITE_FLUSH_FUA);
		else
			sync_dirty_buffer(jl->j_commit_bh);
		reiserfs_write_lock(s);
	}

	/* If there was a write error in the journal - we can't commit this
	 * transaction - it will be invalid and, if successful, will just end
	 * up propagating the write error out to the filesystem. */
	if (unlikely(!buffer_uptodate(jl->j_commit_bh))) {
#ifdef CONFIG_REISERFS_CHECK
		reiserfs_warning(s, "journal-615", "buffer write failed");
#endif
		retval = -EIO;
	}
	bforget(jl->j_commit_bh);
	if (journal->j_last_commit_id != 0 &&
	    (jl->j_trans_id - journal->j_last_commit_id) != 1) {
		reiserfs_warning(s, "clm-2200", "last commit %lu, current %lu",
				 journal->j_last_commit_id, jl->j_trans_id);
	}
	journal->j_last_commit_id = jl->j_trans_id;

	/* now, every commit block is on the disk.  It is safe to allow blocks freed during this transaction to be reallocated */
	cleanup_freed_for_journal_list(s, jl);

	retval = retval ? retval : journal->j_errno;

	/* mark the metadata dirty */
	if (!retval)
		dirty_one_transaction(s, jl);
	atomic_dec(&(jl->j_commit_left));

	if (flushall) {
		atomic_set(&(jl->j_older_commits_done), 1);
	}
	mutex_unlock(&jl->j_commit_mutex);
      put_jl:
	put_journal_list(s, jl);

	if (retval)
		reiserfs_abort(s, retval, "Journal write error in %s",
			       __func__);
	return retval;
}

/*
** flush_journal_list frequently needs to find a newer transaction for a given block.  This does that, or
** returns NULL if it can't find anything
*/
static struct reiserfs_journal_list *find_newer_jl_for_cn(struct
							  reiserfs_journal_cnode
							  *cn)
{
	struct super_block *sb = cn->sb;
	b_blocknr_t blocknr = cn->blocknr;

	cn = cn->hprev;
	while (cn) {
		if (cn->sb == sb && cn->blocknr == blocknr && cn->jlist) {
			return cn->jlist;
		}
		cn = cn->hprev;
	}
	return NULL;
}

static int newer_jl_done(struct reiserfs_journal_cnode *cn)
{
	struct super_block *sb = cn->sb;
	b_blocknr_t blocknr = cn->blocknr;

	cn = cn->hprev;
	while (cn) {
		if (cn->sb == sb && cn->blocknr == blocknr && cn->jlist &&
		    atomic_read(&cn->jlist->j_commit_left) != 0)
				    return 0;
		cn = cn->hprev;
	}
	return 1;
}

static void remove_journal_hash(struct super_block *,
				struct reiserfs_journal_cnode **,
				struct reiserfs_journal_list *, unsigned long,
				int);

/*
** once all the real blocks have been flushed, it is safe to remove them from the
** journal list for this transaction.  Aside from freeing the cnode, this also allows the
** block to be reallocated for data blocks if it had been deleted.
*/
static void remove_all_from_journal_list(struct super_block *sb,
					 struct reiserfs_journal_list *jl,
					 int debug)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct reiserfs_journal_cnode *cn, *last;
	cn = jl->j_realblock;

	/* which is better, to lock once around the whole loop, or
	 ** to lock for each call to remove_journal_hash?
	 */
	while (cn) {
		if (cn->blocknr != 0) {
			if (debug) {
				reiserfs_warning(sb, "reiserfs-2201",
						 "block %u, bh is %d, state %ld",
						 cn->blocknr, cn->bh ? 1 : 0,
						 cn->state);
			}
			cn->state = 0;
			remove_journal_hash(sb, journal->j_list_hash_table,
					    jl, cn->blocknr, 1);
		}
		last = cn;
		cn = cn->next;
		free_cnode(sb, last);
	}
	jl->j_realblock = NULL;
}

/*
** if this timestamp is greater than the timestamp we wrote last to the header block, write it to the header block.
** once this is done, I can safely say the log area for this transaction won't ever be replayed, and I can start
** releasing blocks in this transaction for reuse as data blocks.
** called by flush_journal_list, before it calls remove_all_from_journal_list
**
*/
static int _update_journal_header_block(struct super_block *sb,
					unsigned long offset,
					unsigned int trans_id)
{
	struct reiserfs_journal_header *jh;
	struct reiserfs_journal *journal = SB_JOURNAL(sb);

	if (reiserfs_is_journal_aborted(journal))
		return -EIO;

	if (trans_id >= journal->j_last_flush_trans_id) {
		if (buffer_locked((journal->j_header_bh))) {
			reiserfs_write_unlock(sb);
			wait_on_buffer((journal->j_header_bh));
			reiserfs_write_lock(sb);
			if (unlikely(!buffer_uptodate(journal->j_header_bh))) {
#ifdef CONFIG_REISERFS_CHECK
				reiserfs_warning(sb, "journal-699",
						 "buffer write failed");
#endif
				return -EIO;
			}
		}
		journal->j_last_flush_trans_id = trans_id;
		journal->j_first_unflushed_offset = offset;
		jh = (struct reiserfs_journal_header *)(journal->j_header_bh->
							b_data);
		jh->j_last_flush_trans_id = cpu_to_le32(trans_id);
		jh->j_first_unflushed_offset = cpu_to_le32(offset);
		jh->j_mount_id = cpu_to_le32(journal->j_mount_id);

		set_buffer_dirty(journal->j_header_bh);
		reiserfs_write_unlock(sb);

		if (reiserfs_barrier_flush(sb))
			__sync_dirty_buffer(journal->j_header_bh, WRITE_FLUSH_FUA);
		else
			sync_dirty_buffer(journal->j_header_bh);

		reiserfs_write_lock(sb);
		if (!buffer_uptodate(journal->j_header_bh)) {
			reiserfs_warning(sb, "journal-837",
					 "IO error during journal replay");
			return -EIO;
		}
	}
	return 0;
}

static int update_journal_header_block(struct super_block *sb,
				       unsigned long offset,
				       unsigned int trans_id)
{
	return _update_journal_header_block(sb, offset, trans_id);
}

/*
** flush any and all journal lists older than you are
** can only be called from flush_journal_list
*/
static int flush_older_journal_lists(struct super_block *sb,
				     struct reiserfs_journal_list *jl)
{
	struct list_head *entry;
	struct reiserfs_journal_list *other_jl;
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	unsigned int trans_id = jl->j_trans_id;

	/* we know we are the only ones flushing things, no extra race
	 * protection is required.
	 */
      restart:
	entry = journal->j_journal_list.next;
	/* Did we wrap? */
	if (entry == &journal->j_journal_list)
		return 0;
	other_jl = JOURNAL_LIST_ENTRY(entry);
	if (other_jl->j_trans_id < trans_id) {
		BUG_ON(other_jl->j_refcount <= 0);
		/* do not flush all */
		flush_journal_list(sb, other_jl, 0);

		/* other_jl is now deleted from the list */
		goto restart;
	}
	return 0;
}

static void del_from_work_list(struct super_block *s,
			       struct reiserfs_journal_list *jl)
{
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	if (!list_empty(&jl->j_working_list)) {
		list_del_init(&jl->j_working_list);
		journal->j_num_work_lists--;
	}
}

/* flush a journal list, both commit and real blocks
**
** always set flushall to 1, unless you are calling from inside
** flush_journal_list
**
** IMPORTANT.  This can only be called while there are no journal writers,
** and the journal is locked.  That means it can only be called from
** do_journal_end, or by journal_release
*/
static int flush_journal_list(struct super_block *s,
			      struct reiserfs_journal_list *jl, int flushall)
{
	struct reiserfs_journal_list *pjl;
	struct reiserfs_journal_cnode *cn, *last;
	int count;
	int was_jwait = 0;
	int was_dirty = 0;
	struct buffer_head *saved_bh;
	unsigned long j_len_saved = jl->j_len;
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	int err = 0;

	BUG_ON(j_len_saved <= 0);

	if (atomic_read(&journal->j_wcount) != 0) {
		reiserfs_warning(s, "clm-2048", "called with wcount %d",
				 atomic_read(&journal->j_wcount));
	}
	BUG_ON(jl->j_trans_id == 0);

	/* if flushall == 0, the lock is already held */
	if (flushall) {
		reiserfs_mutex_lock_safe(&journal->j_flush_mutex, s);
	} else if (mutex_trylock(&journal->j_flush_mutex)) {
		BUG();
	}

	count = 0;
	if (j_len_saved > journal->j_trans_max) {
		reiserfs_panic(s, "journal-715", "length is %lu, trans id %lu",
			       j_len_saved, jl->j_trans_id);
		return 0;
	}

	/* if all the work is already done, get out of here */
	if (atomic_read(&(jl->j_nonzerolen)) <= 0 &&
	    atomic_read(&(jl->j_commit_left)) <= 0) {
		goto flush_older_and_return;
	}

	/* start by putting the commit list on disk.  This will also flush
	 ** the commit lists of any olders transactions
	 */
	flush_commit_list(s, jl, 1);

	if (!(jl->j_state & LIST_DIRTY)
	    && !reiserfs_is_journal_aborted(journal))
		BUG();

	/* are we done now? */
	if (atomic_read(&(jl->j_nonzerolen)) <= 0 &&
	    atomic_read(&(jl->j_commit_left)) <= 0) {
		goto flush_older_and_return;
	}

	/* loop through each cnode, see if we need to write it,
	 ** or wait on a more recent transaction, or just ignore it
	 */
	if (atomic_read(&(journal->j_wcount)) != 0) {
		reiserfs_panic(s, "journal-844", "journal list is flushing, "
			       "wcount is not 0");
	}
	cn = jl->j_realblock;
	while (cn) {
		was_jwait = 0;
		was_dirty = 0;
		saved_bh = NULL;
		/* blocknr of 0 is no longer in the hash, ignore it */
		if (cn->blocknr == 0) {
			goto free_cnode;
		}

		/* This transaction failed commit. Don't write out to the disk */
		if (!(jl->j_state & LIST_DIRTY))
			goto free_cnode;

		pjl = find_newer_jl_for_cn(cn);
		/* the order is important here.  We check pjl to make sure we
		 ** don't clear BH_JDirty_wait if we aren't the one writing this
		 ** block to disk
		 */
		if (!pjl && cn->bh) {
			saved_bh = cn->bh;

			/* we do this to make sure nobody releases the buffer while
			 ** we are working with it
			 */
			get_bh(saved_bh);

			if (buffer_journal_dirty(saved_bh)) {
				BUG_ON(!can_dirty(cn));
				was_jwait = 1;
				was_dirty = 1;
			} else if (can_dirty(cn)) {
				/* everything with !pjl && jwait should be writable */
				BUG();
			}
		}

		/* if someone has this block in a newer transaction, just make
		 ** sure they are committed, and don't try writing it to disk
		 */
		if (pjl) {
			if (atomic_read(&pjl->j_commit_left))
				flush_commit_list(s, pjl, 1);
			goto free_cnode;
		}

		/* bh == NULL when the block got to disk on its own, OR,
		 ** the block got freed in a future transaction
		 */
		if (saved_bh == NULL) {
			goto free_cnode;
		}

		/* this should never happen.  kupdate_one_transaction has this list
		 ** locked while it works, so we should never see a buffer here that
		 ** is not marked JDirty_wait
		 */
		if ((!was_jwait) && !buffer_locked(saved_bh)) {
			reiserfs_warning(s, "journal-813",
					 "BAD! buffer %llu %cdirty %cjwait, "
					 "not in a newer tranasction",
					 (unsigned long long)saved_bh->
					 b_blocknr, was_dirty ? ' ' : '!',
					 was_jwait ? ' ' : '!');
		}
		if (was_dirty) {
			/* we inc again because saved_bh gets decremented at free_cnode */
			get_bh(saved_bh);
			set_bit(BLOCK_NEEDS_FLUSH, &cn->state);
			lock_buffer(saved_bh);
			BUG_ON(cn->blocknr != saved_bh->b_blocknr);
			if (buffer_dirty(saved_bh))
				submit_logged_buffer(saved_bh);
			else
				unlock_buffer(saved_bh);
			count++;
		} else {
			reiserfs_warning(s, "clm-2082",
					 "Unable to flush buffer %llu in %s",
					 (unsigned long long)saved_bh->
					 b_blocknr, __func__);
		}
	      free_cnode:
		last = cn;
		cn = cn->next;
		if (saved_bh) {
			/* we incremented this to keep others from taking the buffer head away */
			put_bh(saved_bh);
			if (atomic_read(&(saved_bh->b_count)) < 0) {
				reiserfs_warning(s, "journal-945",
						 "saved_bh->b_count < 0");
			}
		}
	}
	if (count > 0) {
		cn = jl->j_realblock;
		while (cn) {
			if (test_bit(BLOCK_NEEDS_FLUSH, &cn->state)) {
				if (!cn->bh) {
					reiserfs_panic(s, "journal-1011",
						       "cn->bh is NULL");
				}

				reiserfs_write_unlock(s);
				wait_on_buffer(cn->bh);
				reiserfs_write_lock(s);

				if (!cn->bh) {
					reiserfs_panic(s, "journal-1012",
						       "cn->bh is NULL");
				}
				if (unlikely(!buffer_uptodate(cn->bh))) {
#ifdef CONFIG_REISERFS_CHECK
					reiserfs_warning(s, "journal-949",
							 "buffer write failed");
#endif
					err = -EIO;
				}
				/* note, we must clear the JDirty_wait bit after the up to date
				 ** check, otherwise we race against our flushpage routine
				 */
				BUG_ON(!test_clear_buffer_journal_dirty
				       (cn->bh));

				/* drop one ref for us */
				put_bh(cn->bh);
				/* drop one ref for journal_mark_dirty */
				release_buffer_page(cn->bh);
			}
			cn = cn->next;
		}
	}

	if (err)
		reiserfs_abort(s, -EIO,
			       "Write error while pushing transaction to disk in %s",
			       __func__);
      flush_older_and_return:

	/* before we can update the journal header block, we _must_ flush all
	 ** real blocks from all older transactions to disk.  This is because
	 ** once the header block is updated, this transaction will not be
	 ** replayed after a crash
	 */
	if (flushall) {
		flush_older_journal_lists(s, jl);
	}

	err = journal->j_errno;
	/* before we can remove everything from the hash tables for this
	 ** transaction, we must make sure it can never be replayed
	 **
	 ** since we are only called from do_journal_end, we know for sure there
	 ** are no allocations going on while we are flushing journal lists.  So,
	 ** we only need to update the journal header block for the last list
	 ** being flushed
	 */
	if (!err && flushall) {
		err =
		    update_journal_header_block(s,
						(jl->j_start + jl->j_len +
						 2) % SB_ONDISK_JOURNAL_SIZE(s),
						jl->j_trans_id);
		if (err)
			reiserfs_abort(s, -EIO,
				       "Write error while updating journal header in %s",
				       __func__);
	}
	remove_all_from_journal_list(s, jl, 0);
	list_del_init(&jl->j_list);
	journal->j_num_lists--;
	del_from_work_list(s, jl);

	if (journal->j_last_flush_id != 0 &&
	    (jl->j_trans_id - journal->j_last_flush_id) != 1) {
		reiserfs_warning(s, "clm-2201", "last flush %lu, current %lu",
				 journal->j_last_flush_id, jl->j_trans_id);
	}
	journal->j_last_flush_id = jl->j_trans_id;

	/* not strictly required since we are freeing the list, but it should
	 * help find code using dead lists later on
	 */
	jl->j_len = 0;
	atomic_set(&(jl->j_nonzerolen), 0);
	jl->j_start = 0;
	jl->j_realblock = NULL;
	jl->j_commit_bh = NULL;
	jl->j_trans_id = 0;
	jl->j_state = 0;
	put_journal_list(s, jl);
	if (flushall)
		mutex_unlock(&journal->j_flush_mutex);
	return err;
}

static int test_transaction(struct super_block *s,
                            struct reiserfs_journal_list *jl)
{
	struct reiserfs_journal_cnode *cn;

	if (jl->j_len == 0 || atomic_read(&jl->j_nonzerolen) == 0)
		return 1;

	cn = jl->j_realblock;
	while (cn) {
		/* if the blocknr == 0, this has been cleared from the hash,
		 ** skip it
		 */
		if (cn->blocknr == 0) {
			goto next;
		}
		if (cn->bh && !newer_jl_done(cn))
			return 0;
	      next:
		cn = cn->next;
		cond_resched();
	}
	return 0;
}

static int write_one_transaction(struct super_block *s,
				 struct reiserfs_journal_list *jl,
				 struct buffer_chunk *chunk)
{
	struct reiserfs_journal_cnode *cn;
	int ret = 0;

	jl->j_state |= LIST_TOUCHED;
	del_from_work_list(s, jl);
	if (jl->j_len == 0 || atomic_read(&jl->j_nonzerolen) == 0) {
		return 0;
	}

	cn = jl->j_realblock;
	while (cn) {
		/* if the blocknr == 0, this has been cleared from the hash,
		 ** skip it
		 */
		if (cn->blocknr == 0) {
			goto next;
		}
		if (cn->bh && can_dirty(cn) && buffer_dirty(cn->bh)) {
			struct buffer_head *tmp_bh;
			/* we can race against journal_mark_freed when we try
			 * to lock_buffer(cn->bh), so we have to inc the buffer
			 * count, and recheck things after locking
			 */
			tmp_bh = cn->bh;
			get_bh(tmp_bh);
			lock_buffer(tmp_bh);
			if (cn->bh && can_dirty(cn) && buffer_dirty(tmp_bh)) {
				if (!buffer_journal_dirty(tmp_bh) ||
				    buffer_journal_prepared(tmp_bh))
					BUG();
				add_to_chunk(chunk, tmp_bh, NULL, write_chunk);
				ret++;
			} else {
				/* note, cn->bh might be null now */
				unlock_buffer(tmp_bh);
			}
			put_bh(tmp_bh);
		}
	      next:
		cn = cn->next;
		cond_resched();
	}
	return ret;
}

/* used by flush_commit_list */
static int dirty_one_transaction(struct super_block *s,
				 struct reiserfs_journal_list *jl)
{
	struct reiserfs_journal_cnode *cn;
	struct reiserfs_journal_list *pjl;
	int ret = 0;

	jl->j_state |= LIST_DIRTY;
	cn = jl->j_realblock;
	while (cn) {
		/* look for a more recent transaction that logged this
		 ** buffer.  Only the most recent transaction with a buffer in
		 ** it is allowed to send that buffer to disk
		 */
		pjl = find_newer_jl_for_cn(cn);
		if (!pjl && cn->blocknr && cn->bh
		    && buffer_journal_dirty(cn->bh)) {
			BUG_ON(!can_dirty(cn));
			/* if the buffer is prepared, it will either be logged
			 * or restored.  If restored, we need to make sure
			 * it actually gets marked dirty
			 */
			clear_buffer_journal_new(cn->bh);
			if (buffer_journal_prepared(cn->bh)) {
				set_buffer_journal_restore_dirty(cn->bh);
			} else {
				set_buffer_journal_test(cn->bh);
				mark_buffer_dirty(cn->bh);
			}
		}
		cn = cn->next;
	}
	return ret;
}

static int kupdate_transactions(struct super_block *s,
				struct reiserfs_journal_list *jl,
				struct reiserfs_journal_list **next_jl,
				unsigned int *next_trans_id,
				int num_blocks, int num_trans)
{
	int ret = 0;
	int written = 0;
	int transactions_flushed = 0;
	unsigned int orig_trans_id = jl->j_trans_id;
	struct buffer_chunk chunk;
	struct list_head *entry;
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	chunk.nr = 0;

	reiserfs_mutex_lock_safe(&journal->j_flush_mutex, s);
	if (!journal_list_still_alive(s, orig_trans_id)) {
		goto done;
	}

	/* we've got j_flush_mutex held, nobody is going to delete any
	 * of these lists out from underneath us
	 */
	while ((num_trans && transactions_flushed < num_trans) ||
	       (!num_trans && written < num_blocks)) {

		if (jl->j_len == 0 || (jl->j_state & LIST_TOUCHED) ||
		    atomic_read(&jl->j_commit_left)
		    || !(jl->j_state & LIST_DIRTY)) {
			del_from_work_list(s, jl);
			break;
		}
		ret = write_one_transaction(s, jl, &chunk);

		if (ret < 0)
			goto done;
		transactions_flushed++;
		written += ret;
		entry = jl->j_list.next;

		/* did we wrap? */
		if (entry == &journal->j_journal_list) {
			break;
		}
		jl = JOURNAL_LIST_ENTRY(entry);

		/* don't bother with older transactions */
		if (jl->j_trans_id <= orig_trans_id)
			break;
	}
	if (chunk.nr) {
		write_chunk(&chunk);
	}

      done:
	mutex_unlock(&journal->j_flush_mutex);
	return ret;
}

/* for o_sync and fsync heavy applications, they tend to use
** all the journa list slots with tiny transactions.  These
** trigger lots and lots of calls to update the header block, which
** adds seeks and slows things down.
**
** This function tries to clear out a large chunk of the journal lists
** at once, which makes everything faster since only the newest journal
** list updates the header block
*/
static int flush_used_journal_lists(struct super_block *s,
				    struct reiserfs_journal_list *jl)
{
	unsigned long len = 0;
	unsigned long cur_len;
	int ret;
	int i;
	int limit = 256;
	struct reiserfs_journal_list *tjl;
	struct reiserfs_journal_list *flush_jl;
	unsigned int trans_id;
	struct reiserfs_journal *journal = SB_JOURNAL(s);

	flush_jl = tjl = jl;

	/* in data logging mode, try harder to flush a lot of blocks */
	if (reiserfs_data_log(s))
		limit = 1024;
	/* flush for 256 transactions or limit blocks, whichever comes first */
	for (i = 0; i < 256 && len < limit; i++) {
		if (atomic_read(&tjl->j_commit_left) ||
		    tjl->j_trans_id < jl->j_trans_id) {
			break;
		}
		cur_len = atomic_read(&tjl->j_nonzerolen);
		if (cur_len > 0) {
			tjl->j_state &= ~LIST_TOUCHED;
		}
		len += cur_len;
		flush_jl = tjl;
		if (tjl->j_list.next == &journal->j_journal_list)
			break;
		tjl = JOURNAL_LIST_ENTRY(tjl->j_list.next);
	}
	/* try to find a group of blocks we can flush across all the
	 ** transactions, but only bother if we've actually spanned
	 ** across multiple lists
	 */
	if (flush_jl != jl) {
		ret = kupdate_transactions(s, jl, &tjl, &trans_id, len, i);
	}
	flush_journal_list(s, flush_jl, 1);
	return 0;
}

/*
** removes any nodes in table with name block and dev as bh.
** only touchs the hnext and hprev pointers.
*/
void remove_journal_hash(struct super_block *sb,
			 struct reiserfs_journal_cnode **table,
			 struct reiserfs_journal_list *jl,
			 unsigned long block, int remove_freed)
{
	struct reiserfs_journal_cnode *cur;
	struct reiserfs_journal_cnode **head;

	head = &(journal_hash(table, sb, block));
	if (!head) {
		return;
	}
	cur = *head;
	while (cur) {
		if (cur->blocknr == block && cur->sb == sb
		    && (jl == NULL || jl == cur->jlist)
		    && (!test_bit(BLOCK_FREED, &cur->state) || remove_freed)) {
			if (cur->hnext) {
				cur->hnext->hprev = cur->hprev;
			}
			if (cur->hprev) {
				cur->hprev->hnext = cur->hnext;
			} else {
				*head = cur->hnext;
			}
			cur->blocknr = 0;
			cur->sb = NULL;
			cur->state = 0;
			if (cur->bh && cur->jlist)	/* anybody who clears the cur->bh will also dec the nonzerolen */
				atomic_dec(&(cur->jlist->j_nonzerolen));
			cur->bh = NULL;
			cur->jlist = NULL;
		}
		cur = cur->hnext;
	}
}

static void free_journal_ram(struct super_block *sb)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	kfree(journal->j_current_jl);
	journal->j_num_lists--;

	vfree(journal->j_cnode_free_orig);
	free_list_bitmaps(sb, journal->j_list_bitmap);
	free_bitmap_nodes(sb);	/* must be after free_list_bitmaps */
	if (journal->j_header_bh) {
		brelse(journal->j_header_bh);
	}
	/* j_header_bh is on the journal dev, make sure not to release the journal
	 * dev until we brelse j_header_bh
	 */
	release_journal_dev(sb, journal);
	vfree(journal);
}

/*
** call on unmount.  Only set error to 1 if you haven't made your way out
** of read_super() yet.  Any other caller must keep error at 0.
*/
static int do_journal_release(struct reiserfs_transaction_handle *th,
			      struct super_block *sb, int error)
{
	struct reiserfs_transaction_handle myth;
	int flushed = 0;
	struct reiserfs_journal *journal = SB_JOURNAL(sb);

	/* we only want to flush out transactions if we were called with error == 0
	 */
	if (!error && !(sb->s_flags & MS_RDONLY)) {
		/* end the current trans */
		BUG_ON(!th->t_trans_id);
		do_journal_end(th, sb, 10, FLUSH_ALL);

		/* make sure something gets logged to force our way into the flush code */
		if (!journal_join(&myth, sb, 1)) {
			reiserfs_prepare_for_journal(sb,
						     SB_BUFFER_WITH_SB(sb),
						     1);
			journal_mark_dirty(&myth, sb,
					   SB_BUFFER_WITH_SB(sb));
			do_journal_end(&myth, sb, 1, FLUSH_ALL);
			flushed = 1;
		}
	}

	/* this also catches errors during the do_journal_end above */
	if (!error && reiserfs_is_journal_aborted(journal)) {
		memset(&myth, 0, sizeof(myth));
		if (!journal_join_abort(&myth, sb, 1)) {
			reiserfs_prepare_for_journal(sb,
						     SB_BUFFER_WITH_SB(sb),
						     1);
			journal_mark_dirty(&myth, sb,
					   SB_BUFFER_WITH_SB(sb));
			do_journal_end(&myth, sb, 1, FLUSH_ALL);
		}
	}

	reiserfs_mounted_fs_count--;
	/* wait for all commits to finish */
	cancel_delayed_work(&SB_JOURNAL(sb)->j_work);

	/*
	 * We must release the write lock here because
	 * the workqueue job (flush_async_commit) needs this lock
	 */
	reiserfs_write_unlock(sb);
	flush_workqueue(commit_wq);

	if (!reiserfs_mounted_fs_count) {
		destroy_workqueue(commit_wq);
		commit_wq = NULL;
	}

	free_journal_ram(sb);

	reiserfs_write_lock(sb);

	return 0;
}

/*
** call on unmount.  flush all journal trans, release all alloc'd ram
*/
int journal_release(struct reiserfs_transaction_handle *th,
		    struct super_block *sb)
{
	return do_journal_release(th, sb, 0);
}

/*
** only call from an error condition inside reiserfs_read_super!
*/
int journal_release_error(struct reiserfs_transaction_handle *th,
			  struct super_block *sb)
{
	return do_journal_release(th, sb, 1);
}

/* compares description block with commit block.  returns 1 if they differ, 0 if they are the same */
static int journal_compare_desc_commit(struct super_block *sb,
				       struct reiserfs_journal_desc *desc,
				       struct reiserfs_journal_commit *commit)
{
	if (get_commit_trans_id(commit) != get_desc_trans_id(desc) ||
	    get_commit_trans_len(commit) != get_desc_trans_len(desc) ||
	    get_commit_trans_len(commit) > SB_JOURNAL(sb)->j_trans_max ||
	    get_commit_trans_len(commit) <= 0) {
		return 1;
	}
	return 0;
}

/* returns 0 if it did not find a description block
** returns -1 if it found a corrupt commit block
** returns 1 if both desc and commit were valid
*/
static int journal_transaction_is_valid(struct super_block *sb,
					struct buffer_head *d_bh,
					unsigned int *oldest_invalid_trans_id,
					unsigned long *newest_mount_id)
{
	struct reiserfs_journal_desc *desc;
	struct reiserfs_journal_commit *commit;
	struct buffer_head *c_bh;
	unsigned long offset;

	if (!d_bh)
		return 0;

	desc = (struct reiserfs_journal_desc *)d_bh->b_data;
	if (get_desc_trans_len(desc) > 0
	    && !memcmp(get_journal_desc_magic(d_bh), JOURNAL_DESC_MAGIC, 8)) {
		if (oldest_invalid_trans_id && *oldest_invalid_trans_id
		    && get_desc_trans_id(desc) > *oldest_invalid_trans_id) {
			reiserfs_debug(sb, REISERFS_DEBUG_CODE,
				       "journal-986: transaction "
				       "is valid returning because trans_id %d is greater than "
				       "oldest_invalid %lu",
				       get_desc_trans_id(desc),
				       *oldest_invalid_trans_id);
			return 0;
		}
		if (newest_mount_id
		    && *newest_mount_id > get_desc_mount_id(desc)) {
			reiserfs_debug(sb, REISERFS_DEBUG_CODE,
				       "journal-1087: transaction "
				       "is valid returning because mount_id %d is less than "
				       "newest_mount_id %lu",
				       get_desc_mount_id(desc),
				       *newest_mount_id);
			return -1;
		}
		if (get_desc_trans_len(desc) > SB_JOURNAL(sb)->j_trans_max) {
			reiserfs_warning(sb, "journal-2018",
					 "Bad transaction length %d "
					 "encountered, ignoring transaction",
					 get_desc_trans_len(desc));
			return -1;
		}
		offset = d_bh->b_blocknr - SB_ONDISK_JOURNAL_1st_BLOCK(sb);

		/* ok, we have a journal description block, lets see if the transaction was valid */
		c_bh =
		    journal_bread(sb,
				  SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
				  ((offset + get_desc_trans_len(desc) +
				    1) % SB_ONDISK_JOURNAL_SIZE(sb)));
		if (!c_bh)
			return 0;
		commit = (struct reiserfs_journal_commit *)c_bh->b_data;
		if (journal_compare_desc_commit(sb, desc, commit)) {
			reiserfs_debug(sb, REISERFS_DEBUG_CODE,
				       "journal_transaction_is_valid, commit offset %ld had bad "
				       "time %d or length %d",
				       c_bh->b_blocknr -
				       SB_ONDISK_JOURNAL_1st_BLOCK(sb),
				       get_commit_trans_id(commit),
				       get_commit_trans_len(commit));
			brelse(c_bh);
			if (oldest_invalid_trans_id) {
				*oldest_invalid_trans_id =
				    get_desc_trans_id(desc);
				reiserfs_debug(sb, REISERFS_DEBUG_CODE,
					       "journal-1004: "
					       "transaction_is_valid setting oldest invalid trans_id "
					       "to %d",
					       get_desc_trans_id(desc));
			}
			return -1;
		}
		brelse(c_bh);
		reiserfs_debug(sb, REISERFS_DEBUG_CODE,
			       "journal-1006: found valid "
			       "transaction start offset %llu, len %d id %d",
			       d_bh->b_blocknr -
			       SB_ONDISK_JOURNAL_1st_BLOCK(sb),
			       get_desc_trans_len(desc),
			       get_desc_trans_id(desc));
		return 1;
	} else {
		return 0;
	}
}

static void brelse_array(struct buffer_head **heads, int num)
{
	int i;
	for (i = 0; i < num; i++) {
		brelse(heads[i]);
	}
}

/*
** given the start, and values for the oldest acceptable transactions,
** this either reads in a replays a transaction, or returns because the transaction
** is invalid, or too old.
*/
static int journal_read_transaction(struct super_block *sb,
				    unsigned long cur_dblock,
				    unsigned long oldest_start,
				    unsigned int oldest_trans_id,
				    unsigned long newest_mount_id)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct reiserfs_journal_desc *desc;
	struct reiserfs_journal_commit *commit;
	unsigned int trans_id = 0;
	struct buffer_head *c_bh;
	struct buffer_head *d_bh;
	struct buffer_head **log_blocks = NULL;
	struct buffer_head **real_blocks = NULL;
	unsigned int trans_offset;
	int i;
	int trans_half;

	d_bh = journal_bread(sb, cur_dblock);
	if (!d_bh)
		return 1;
	desc = (struct reiserfs_journal_desc *)d_bh->b_data;
	trans_offset = d_bh->b_blocknr - SB_ONDISK_JOURNAL_1st_BLOCK(sb);
	reiserfs_debug(sb, REISERFS_DEBUG_CODE, "journal-1037: "
		       "journal_read_transaction, offset %llu, len %d mount_id %d",
		       d_bh->b_blocknr - SB_ONDISK_JOURNAL_1st_BLOCK(sb),
		       get_desc_trans_len(desc), get_desc_mount_id(desc));
	if (get_desc_trans_id(desc) < oldest_trans_id) {
		reiserfs_debug(sb, REISERFS_DEBUG_CODE, "journal-1039: "
			       "journal_read_trans skipping because %lu is too old",
			       cur_dblock -
			       SB_ONDISK_JOURNAL_1st_BLOCK(sb));
		brelse(d_bh);
		return 1;
	}
	if (get_desc_mount_id(desc) != newest_mount_id) {
		reiserfs_debug(sb, REISERFS_DEBUG_CODE, "journal-1146: "
			       "journal_read_trans skipping because %d is != "
			       "newest_mount_id %lu", get_desc_mount_id(desc),
			       newest_mount_id);
		brelse(d_bh);
		return 1;
	}
	c_bh = journal_bread(sb, SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
			     ((trans_offset + get_desc_trans_len(desc) + 1) %
			      SB_ONDISK_JOURNAL_SIZE(sb)));
	if (!c_bh) {
		brelse(d_bh);
		return 1;
	}
	commit = (struct reiserfs_journal_commit *)c_bh->b_data;
	if (journal_compare_desc_commit(sb, desc, commit)) {
		reiserfs_debug(sb, REISERFS_DEBUG_CODE,
			       "journal_read_transaction, "
			       "commit offset %llu had bad time %d or length %d",
			       c_bh->b_blocknr -
			       SB_ONDISK_JOURNAL_1st_BLOCK(sb),
			       get_commit_trans_id(commit),
			       get_commit_trans_len(commit));
		brelse(c_bh);
		brelse(d_bh);
		return 1;
	}

	if (bdev_read_only(sb->s_bdev)) {
		reiserfs_warning(sb, "clm-2076",
				 "device is readonly, unable to replay log");
		brelse(c_bh);
		brelse(d_bh);
		return -EROFS;
	}

	trans_id = get_desc_trans_id(desc);
	/* now we know we've got a good transaction, and it was inside the valid time ranges */
	log_blocks = kmalloc(get_desc_trans_len(desc) *
			     sizeof(struct buffer_head *), GFP_NOFS);
	real_blocks = kmalloc(get_desc_trans_len(desc) *
			      sizeof(struct buffer_head *), GFP_NOFS);
	if (!log_blocks || !real_blocks) {
		brelse(c_bh);
		brelse(d_bh);
		kfree(log_blocks);
		kfree(real_blocks);
		reiserfs_warning(sb, "journal-1169",
				 "kmalloc failed, unable to mount FS");
		return -1;
	}
	/* get all the buffer heads */
	trans_half = journal_trans_half(sb->s_blocksize);
	for (i = 0; i < get_desc_trans_len(desc); i++) {
		log_blocks[i] =
		    journal_getblk(sb,
				   SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
				   (trans_offset + 1 +
				    i) % SB_ONDISK_JOURNAL_SIZE(sb));
		if (i < trans_half) {
			real_blocks[i] =
			    sb_getblk(sb,
				      le32_to_cpu(desc->j_realblock[i]));
		} else {
			real_blocks[i] =
			    sb_getblk(sb,
				      le32_to_cpu(commit->
						  j_realblock[i - trans_half]));
		}
		if (real_blocks[i]->b_blocknr > SB_BLOCK_COUNT(sb)) {
			reiserfs_warning(sb, "journal-1207",
					 "REPLAY FAILURE fsck required! "
					 "Block to replay is outside of "
					 "filesystem");
			goto abort_replay;
		}
		/* make sure we don't try to replay onto log or reserved area */
		if (is_block_in_log_or_reserved_area
		    (sb, real_blocks[i]->b_blocknr)) {
			reiserfs_warning(sb, "journal-1204",
					 "REPLAY FAILURE fsck required! "
					 "Trying to replay onto a log block");
		      abort_replay:
			brelse_array(log_blocks, i);
			brelse_array(real_blocks, i);
			brelse(c_bh);
			brelse(d_bh);
			kfree(log_blocks);
			kfree(real_blocks);
			return -1;
		}
	}
	/* read in the log blocks, memcpy to the corresponding real block */
	ll_rw_block(READ, get_desc_trans_len(desc), log_blocks);
	for (i = 0; i < get_desc_trans_len(desc); i++) {

		reiserfs_write_unlock(sb);
		wait_on_buffer(log_blocks[i]);
		reiserfs_write_lock(sb);

		if (!buffer_uptodate(log_blocks[i])) {
			reiserfs_warning(sb, "journal-1212",
					 "REPLAY FAILURE fsck required! "
					 "buffer write failed");
			brelse_array(log_blocks + i,
				     get_desc_trans_len(desc) - i);
			brelse_array(real_blocks, get_desc_trans_len(desc));
			brelse(c_bh);
			brelse(d_bh);
			kfree(log_blocks);
			kfree(real_blocks);
			return -1;
		}
		memcpy(real_blocks[i]->b_data, log_blocks[i]->b_data,
		       real_blocks[i]->b_size);
		set_buffer_uptodate(real_blocks[i]);
		brelse(log_blocks[i]);
	}
	/* flush out the real blocks */
	for (i = 0; i < get_desc_trans_len(desc); i++) {
		set_buffer_dirty(real_blocks[i]);
		write_dirty_buffer(real_blocks[i], WRITE);
	}
	for (i = 0; i < get_desc_trans_len(desc); i++) {
		wait_on_buffer(real_blocks[i]);
		if (!buffer_uptodate(real_blocks[i])) {
			reiserfs_warning(sb, "journal-1226",
					 "REPLAY FAILURE, fsck required! "
					 "buffer write failed");
			brelse_array(real_blocks + i,
				     get_desc_trans_len(desc) - i);
			brelse(c_bh);
			brelse(d_bh);
			kfree(log_blocks);
			kfree(real_blocks);
			return -1;
		}
		brelse(real_blocks[i]);
	}
	cur_dblock =
	    SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
	    ((trans_offset + get_desc_trans_len(desc) +
	      2) % SB_ONDISK_JOURNAL_SIZE(sb));
	reiserfs_debug(sb, REISERFS_DEBUG_CODE,
		       "journal-1095: setting journal " "start to offset %ld",
		       cur_dblock - SB_ONDISK_JOURNAL_1st_BLOCK(sb));

	/* init starting values for the first transaction, in case this is the last transaction to be replayed. */
	journal->j_start = cur_dblock - SB_ONDISK_JOURNAL_1st_BLOCK(sb);
	journal->j_last_flush_trans_id = trans_id;
	journal->j_trans_id = trans_id + 1;
	/* check for trans_id overflow */
	if (journal->j_trans_id == 0)
		journal->j_trans_id = 10;
	brelse(c_bh);
	brelse(d_bh);
	kfree(log_blocks);
	kfree(real_blocks);
	return 0;
}

/* This function reads blocks starting from block and to max_block of bufsize
   size (but no more than BUFNR blocks at a time). This proved to improve
   mounting speed on self-rebuilding raid5 arrays at least.
   Right now it is only used from journal code. But later we might use it
   from other places.
   Note: Do not use journal_getblk/sb_getblk functions here! */
static struct buffer_head *reiserfs_breada(struct block_device *dev,
					   b_blocknr_t block, int bufsize,
					   b_blocknr_t max_block)
{
	struct buffer_head *bhlist[BUFNR];
	unsigned int blocks = BUFNR;
	struct buffer_head *bh;
	int i, j;

	bh = __getblk(dev, block, bufsize);
	if (buffer_uptodate(bh))
		return (bh);

	if (block + BUFNR > max_block) {
		blocks = max_block - block;
	}
	bhlist[0] = bh;
	j = 1;
	for (i = 1; i < blocks; i++) {
		bh = __getblk(dev, block + i, bufsize);
		if (buffer_uptodate(bh)) {
			brelse(bh);
			break;
		} else
			bhlist[j++] = bh;
	}
	ll_rw_block(READ, j, bhlist);
	for (i = 1; i < j; i++)
		brelse(bhlist[i]);
	bh = bhlist[0];
	wait_on_buffer(bh);
	if (buffer_uptodate(bh))
		return bh;
	brelse(bh);
	return NULL;
}

/*
** read and replay the log
** on a clean unmount, the journal header's next unflushed pointer will be to an invalid
** transaction.  This tests that before finding all the transactions in the log, which makes normal mount times fast.
**
** After a crash, this starts with the next unflushed transaction, and replays until it finds one too old, or invalid.
**
** On exit, it sets things up so the first transaction will work correctly.
*/
static int journal_read(struct super_block *sb)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct reiserfs_journal_desc *desc;
	unsigned int oldest_trans_id = 0;
	unsigned int oldest_invalid_trans_id = 0;
	time_t start;
	unsigned long oldest_start = 0;
	unsigned long cur_dblock = 0;
	unsigned long newest_mount_id = 9;
	struct buffer_head *d_bh;
	struct reiserfs_journal_header *jh;
	int valid_journal_header = 0;
	int replay_count = 0;
	int continue_replay = 1;
	int ret;
	char b[BDEVNAME_SIZE];

	cur_dblock = SB_ONDISK_JOURNAL_1st_BLOCK(sb);
	reiserfs_info(sb, "checking transaction log (%s)\n",
		      bdevname(journal->j_dev_bd, b));
	start = get_seconds();

	/* step 1, read in the journal header block.  Check the transaction it says
	 ** is the first unflushed, and if that transaction is not valid,
	 ** replay is done
	 */
	journal->j_header_bh = journal_bread(sb,
					     SB_ONDISK_JOURNAL_1st_BLOCK(sb)
					     + SB_ONDISK_JOURNAL_SIZE(sb));
	if (!journal->j_header_bh) {
		return 1;
	}
	jh = (struct reiserfs_journal_header *)(journal->j_header_bh->b_data);
	if (le32_to_cpu(jh->j_first_unflushed_offset) <
	    SB_ONDISK_JOURNAL_SIZE(sb)
	    && le32_to_cpu(jh->j_last_flush_trans_id) > 0) {
		oldest_start =
		    SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
		    le32_to_cpu(jh->j_first_unflushed_offset);
		oldest_trans_id = le32_to_cpu(jh->j_last_flush_trans_id) + 1;
		newest_mount_id = le32_to_cpu(jh->j_mount_id);
		reiserfs_debug(sb, REISERFS_DEBUG_CODE,
			       "journal-1153: found in "
			       "header: first_unflushed_offset %d, last_flushed_trans_id "
			       "%lu", le32_to_cpu(jh->j_first_unflushed_offset),
			       le32_to_cpu(jh->j_last_flush_trans_id));
		valid_journal_header = 1;

		/* now, we try to read the first unflushed offset.  If it is not valid,
		 ** there is nothing more we can do, and it makes no sense to read
		 ** through the whole log.
		 */
		d_bh =
		    journal_bread(sb,
				  SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
				  le32_to_cpu(jh->j_first_unflushed_offset));
		ret = journal_transaction_is_valid(sb, d_bh, NULL, NULL);
		if (!ret) {
			continue_replay = 0;
		}
		brelse(d_bh);
		goto start_log_replay;
	}

	/* ok, there are transactions that need to be replayed.  start with the first log block, find
	 ** all the valid transactions, and pick out the oldest.
	 */
	while (continue_replay
	       && cur_dblock <
	       (SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
		SB_ONDISK_JOURNAL_SIZE(sb))) {
		/* Note that it is required for blocksize of primary fs device and journal
		   device to be the same */
		d_bh =
		    reiserfs_breada(journal->j_dev_bd, cur_dblock,
				    sb->s_blocksize,
				    SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
				    SB_ONDISK_JOURNAL_SIZE(sb));
		ret =
		    journal_transaction_is_valid(sb, d_bh,
						 &oldest_invalid_trans_id,
						 &newest_mount_id);
		if (ret == 1) {
			desc = (struct reiserfs_journal_desc *)d_bh->b_data;
			if (oldest_start == 0) {	/* init all oldest_ values */
				oldest_trans_id = get_desc_trans_id(desc);
				oldest_start = d_bh->b_blocknr;
				newest_mount_id = get_desc_mount_id(desc);
				reiserfs_debug(sb, REISERFS_DEBUG_CODE,
					       "journal-1179: Setting "
					       "oldest_start to offset %llu, trans_id %lu",
					       oldest_start -
					       SB_ONDISK_JOURNAL_1st_BLOCK
					       (sb), oldest_trans_id);
			} else if (oldest_trans_id > get_desc_trans_id(desc)) {
				/* one we just read was older */
				oldest_trans_id = get_desc_trans_id(desc);
				oldest_start = d_bh->b_blocknr;
				reiserfs_debug(sb, REISERFS_DEBUG_CODE,
					       "journal-1180: Resetting "
					       "oldest_start to offset %lu, trans_id %lu",
					       oldest_start -
					       SB_ONDISK_JOURNAL_1st_BLOCK
					       (sb), oldest_trans_id);
			}
			if (newest_mount_id < get_desc_mount_id(desc)) {
				newest_mount_id = get_desc_mount_id(desc);
				reiserfs_debug(sb, REISERFS_DEBUG_CODE,
					       "journal-1299: Setting "
					       "newest_mount_id to %d",
					       get_desc_mount_id(desc));
			}
			cur_dblock += get_desc_trans_len(desc) + 2;
		} else {
			cur_dblock++;
		}
		brelse(d_bh);
	}

      start_log_replay:
	cur_dblock = oldest_start;
	if (oldest_trans_id) {
		reiserfs_debug(sb, REISERFS_DEBUG_CODE,
			       "journal-1206: Starting replay "
			       "from offset %llu, trans_id %lu",
			       cur_dblock - SB_ONDISK_JOURNAL_1st_BLOCK(sb),
			       oldest_trans_id);

	}
	replay_count = 0;
	while (continue_replay && oldest_trans_id > 0) {
		ret =
		    journal_read_transaction(sb, cur_dblock, oldest_start,
					     oldest_trans_id, newest_mount_id);
		if (ret < 0) {
			return ret;
		} else if (ret != 0) {
			break;
		}
		cur_dblock =
		    SB_ONDISK_JOURNAL_1st_BLOCK(sb) + journal->j_start;
		replay_count++;
		if (cur_dblock == oldest_start)
			break;
	}

	if (oldest_trans_id == 0) {
		reiserfs_debug(sb, REISERFS_DEBUG_CODE,
			       "journal-1225: No valid " "transactions found");
	}
	/* j_start does not get set correctly if we don't replay any transactions.
	 ** if we had a valid journal_header, set j_start to the first unflushed transaction value,
	 ** copy the trans_id from the header
	 */
	if (valid_journal_header && replay_count == 0) {
		journal->j_start = le32_to_cpu(jh->j_first_unflushed_offset);
		journal->j_trans_id =
		    le32_to_cpu(jh->j_last_flush_trans_id) + 1;
		/* check for trans_id overflow */
		if (journal->j_trans_id == 0)
			journal->j_trans_id = 10;
		journal->j_last_flush_trans_id =
		    le32_to_cpu(jh->j_last_flush_trans_id);
		journal->j_mount_id = le32_to_cpu(jh->j_mount_id) + 1;
	} else {
		journal->j_mount_id = newest_mount_id + 1;
	}
	reiserfs_debug(sb, REISERFS_DEBUG_CODE, "journal-1299: Setting "
		       "newest_mount_id to %lu", journal->j_mount_id);
	journal->j_first_unflushed_offset = journal->j_start;
	if (replay_count > 0) {
		reiserfs_info(sb,
			      "replayed %d transactions in %lu seconds\n",
			      replay_count, get_seconds() - start);
	}
	if (!bdev_read_only(sb->s_bdev) &&
	    _update_journal_header_block(sb, journal->j_start,
					 journal->j_last_flush_trans_id)) {
		/* replay failed, caller must call free_journal_ram and abort
		 ** the mount
		 */
		return -1;
	}
	return 0;
}

static struct reiserfs_journal_list *alloc_journal_list(struct super_block *s)
{
	struct reiserfs_journal_list *jl;
	jl = kzalloc(sizeof(struct reiserfs_journal_list),
		     GFP_NOFS | __GFP_NOFAIL);
	INIT_LIST_HEAD(&jl->j_list);
	INIT_LIST_HEAD(&jl->j_working_list);
	INIT_LIST_HEAD(&jl->j_tail_bh_list);
	INIT_LIST_HEAD(&jl->j_bh_list);
	mutex_init(&jl->j_commit_mutex);
	SB_JOURNAL(s)->j_num_lists++;
	get_journal_list(jl);
	return jl;
}

static void journal_list_init(struct super_block *sb)
{
	SB_JOURNAL(sb)->j_current_jl = alloc_journal_list(sb);
}

static int release_journal_dev(struct super_block *super,
			       struct reiserfs_journal *journal)
{
	int result;

	result = 0;

	if (journal->j_dev_bd != NULL) {
		result = blkdev_put(journal->j_dev_bd, journal->j_dev_mode);
		journal->j_dev_bd = NULL;
	}

	if (result != 0) {
		reiserfs_warning(super, "sh-457",
				 "Cannot release journal device: %i", result);
	}
	return result;
}

static int journal_init_dev(struct super_block *super,
			    struct reiserfs_journal *journal,
			    const char *jdev_name)
{
	int result;
	dev_t jdev;
	fmode_t blkdev_mode = FMODE_READ | FMODE_WRITE | FMODE_EXCL;
	char b[BDEVNAME_SIZE];

	result = 0;

	journal->j_dev_bd = NULL;
	jdev = SB_ONDISK_JOURNAL_DEVICE(super) ?
	    new_decode_dev(SB_ONDISK_JOURNAL_DEVICE(super)) : super->s_dev;

	if (bdev_read_only(super->s_bdev))
		blkdev_mode = FMODE_READ;

	/* there is no "jdev" option and journal is on separate device */
	if ((!jdev_name || !jdev_name[0])) {
		if (jdev == super->s_dev)
			blkdev_mode &= ~FMODE_EXCL;
		journal->j_dev_bd = blkdev_get_by_dev(jdev, blkdev_mode,
						      journal);
		journal->j_dev_mode = blkdev_mode;
		if (IS_ERR(journal->j_dev_bd)) {
			result = PTR_ERR(journal->j_dev_bd);
			journal->j_dev_bd = NULL;
			reiserfs_warning(super, "sh-458",
					 "cannot init journal device '%s': %i",
					 __bdevname(jdev, b), result);
			return result;
		} else if (jdev != super->s_dev)
			set_blocksize(journal->j_dev_bd, super->s_blocksize);

		return 0;
	}

	journal->j_dev_mode = blkdev_mode;
	journal->j_dev_bd = blkdev_get_by_path(jdev_name, blkdev_mode, journal);
	if (IS_ERR(journal->j_dev_bd)) {
		result = PTR_ERR(journal->j_dev_bd);
		journal->j_dev_bd = NULL;
		reiserfs_warning(super,
				 "journal_init_dev: Cannot open '%s': %i",
				 jdev_name, result);
		return result;
	}

	set_blocksize(journal->j_dev_bd, super->s_blocksize);
	reiserfs_info(super,
		      "journal_init_dev: journal device: %s\n",
		      bdevname(journal->j_dev_bd, b));
	return 0;
}

/**
 * When creating/tuning a file system user can assign some
 * journal params within boundaries which depend on the ratio
 * blocksize/standard_blocksize.
 *
 * For blocks >= standard_blocksize transaction size should
 * be not less then JOURNAL_TRANS_MIN_DEFAULT, and not more
 * then JOURNAL_TRANS_MAX_DEFAULT.
 *
 * For blocks < standard_blocksize these boundaries should be
 * decreased proportionally.
 */
#define REISERFS_STANDARD_BLKSIZE (4096)

static int check_advise_trans_params(struct super_block *sb,
				     struct reiserfs_journal *journal)
{
        if (journal->j_trans_max) {
	        /* Non-default journal params.
		   Do sanity check for them. */
	        int ratio = 1;
		if (sb->s_blocksize < REISERFS_STANDARD_BLKSIZE)
		        ratio = REISERFS_STANDARD_BLKSIZE / sb->s_blocksize;

		if (journal->j_trans_max > JOURNAL_TRANS_MAX_DEFAULT / ratio ||
		    journal->j_trans_max < JOURNAL_TRANS_MIN_DEFAULT / ratio ||
		    SB_ONDISK_JOURNAL_SIZE(sb) / journal->j_trans_max <
		    JOURNAL_MIN_RATIO) {
			reiserfs_warning(sb, "sh-462",
					 "bad transaction max size (%u). "
					 "FSCK?", journal->j_trans_max);
			return 1;
		}
		if (journal->j_max_batch != (journal->j_trans_max) *
		        JOURNAL_MAX_BATCH_DEFAULT/JOURNAL_TRANS_MAX_DEFAULT) {
			reiserfs_warning(sb, "sh-463",
					 "bad transaction max batch (%u). "
					 "FSCK?", journal->j_max_batch);
			return 1;
		}
	} else {
		/* Default journal params.
                   The file system was created by old version
		   of mkreiserfs, so some fields contain zeros,
		   and we need to advise proper values for them */
		if (sb->s_blocksize != REISERFS_STANDARD_BLKSIZE) {
			reiserfs_warning(sb, "sh-464", "bad blocksize (%u)",
					 sb->s_blocksize);
			return 1;
		}
		journal->j_trans_max = JOURNAL_TRANS_MAX_DEFAULT;
		journal->j_max_batch = JOURNAL_MAX_BATCH_DEFAULT;
		journal->j_max_commit_age = JOURNAL_MAX_COMMIT_AGE;
	}
	return 0;
}

/*
** must be called once on fs mount.  calls journal_read for you
*/
int journal_init(struct super_block *sb, const char *j_dev_name,
		 int old_format, unsigned int commit_max_age)
{
	int num_cnodes = SB_ONDISK_JOURNAL_SIZE(sb) * 2;
	struct buffer_head *bhjh;
	struct reiserfs_super_block *rs;
	struct reiserfs_journal_header *jh;
	struct reiserfs_journal *journal;
	struct reiserfs_journal_list *jl;
	char b[BDEVNAME_SIZE];
	int ret;

	/*
	 * Unlock here to avoid various RECLAIM-FS-ON <-> IN-RECLAIM-FS
	 * dependency inversion warnings.
	 */
	reiserfs_write_unlock(sb);
	journal = SB_JOURNAL(sb) = vzalloc(sizeof(struct reiserfs_journal));
	if (!journal) {
		reiserfs_warning(sb, "journal-1256",
				 "unable to get memory for journal structure");
		reiserfs_write_lock(sb);
		return 1;
	}
	INIT_LIST_HEAD(&journal->j_bitmap_nodes);
	INIT_LIST_HEAD(&journal->j_prealloc_list);
	INIT_LIST_HEAD(&journal->j_working_list);
	INIT_LIST_HEAD(&journal->j_journal_list);
	journal->j_persistent_trans = 0;
	ret = reiserfs_allocate_list_bitmaps(sb, journal->j_list_bitmap,
					   reiserfs_bmap_count(sb));
	reiserfs_write_lock(sb);
	if (ret)
		goto free_and_return;

	allocate_bitmap_nodes(sb);

	/* reserved for journal area support */
	SB_JOURNAL_1st_RESERVED_BLOCK(sb) = (old_format ?
						 REISERFS_OLD_DISK_OFFSET_IN_BYTES
						 / sb->s_blocksize +
						 reiserfs_bmap_count(sb) +
						 1 :
						 REISERFS_DISK_OFFSET_IN_BYTES /
						 sb->s_blocksize + 2);

	/* Sanity check to see is the standard journal fitting within first bitmap
	   (actual for small blocksizes) */
	if (!SB_ONDISK_JOURNAL_DEVICE(sb) &&
	    (SB_JOURNAL_1st_RESERVED_BLOCK(sb) +
	     SB_ONDISK_JOURNAL_SIZE(sb) > sb->s_blocksize * 8)) {
		reiserfs_warning(sb, "journal-1393",
				 "journal does not fit for area addressed "
				 "by first of bitmap blocks. It starts at "
				 "%u and its size is %u. Block size %ld",
				 SB_JOURNAL_1st_RESERVED_BLOCK(sb),
				 SB_ONDISK_JOURNAL_SIZE(sb),
				 sb->s_blocksize);
		goto free_and_return;
	}

	/*
	 * We need to unlock here to avoid creating the following
	 * dependency:
	 * reiserfs_lock -> sysfs_mutex
	 * Because the reiserfs mmap path creates the following dependency:
	 * mm->mmap -> reiserfs_lock, hence we have
	 * mm->mmap -> reiserfs_lock ->sysfs_mutex
	 * This would ends up in a circular dependency with sysfs readdir path
	 * which does sysfs_mutex -> mm->mmap_sem
	 * This is fine because the reiserfs lock is useless in mount path,
	 * at least until we call journal_begin. We keep it for paranoid
	 * reasons.
	 */
	reiserfs_write_unlock(sb);
	if (journal_init_dev(sb, journal, j_dev_name) != 0) {
		reiserfs_write_lock(sb);
		reiserfs_warning(sb, "sh-462",
				 "unable to initialize jornal device");
		goto free_and_return;
	}
	reiserfs_write_lock(sb);

	rs = SB_DISK_SUPER_BLOCK(sb);

	/* read journal header */
	bhjh = journal_bread(sb,
			     SB_ONDISK_JOURNAL_1st_BLOCK(sb) +
			     SB_ONDISK_JOURNAL_SIZE(sb));
	if (!bhjh) {
		reiserfs_warning(sb, "sh-459",
				 "unable to read journal header");
		goto free_and_return;
	}
	jh = (struct reiserfs_journal_header *)(bhjh->b_data);

	/* make sure that journal matches to the super block */
	if (is_reiserfs_jr(rs)
	    && (le32_to_cpu(jh->jh_journal.jp_journal_magic) !=
		sb_jp_journal_magic(rs))) {
		reiserfs_warning(sb, "sh-460",
				 "journal header magic %x (device %s) does "
				 "not match to magic found in super block %x",
				 jh->jh_journal.jp_journal_magic,
				 bdevname(journal->j_dev_bd, b),
				 sb_jp_journal_magic(rs));
		brelse(bhjh);
		goto free_and_return;
	}

	journal->j_trans_max = le32_to_cpu(jh->jh_journal.jp_journal_trans_max);
	journal->j_max_batch = le32_to_cpu(jh->jh_journal.jp_journal_max_batch);
	journal->j_max_commit_age =
	    le32_to_cpu(jh->jh_journal.jp_journal_max_commit_age);
	journal->j_max_trans_age = JOURNAL_MAX_TRANS_AGE;

	if (check_advise_trans_params(sb, journal) != 0)
	        goto free_and_return;
	journal->j_default_max_commit_age = journal->j_max_commit_age;

	if (commit_max_age != 0) {
		journal->j_max_commit_age = commit_max_age;
		journal->j_max_trans_age = commit_max_age;
	}

	reiserfs_info(sb, "journal params: device %s, size %u, "
		      "journal first block %u, max trans len %u, max batch %u, "
		      "max commit age %u, max trans age %u\n",
		      bdevname(journal->j_dev_bd, b),
		      SB_ONDISK_JOURNAL_SIZE(sb),
		      SB_ONDISK_JOURNAL_1st_BLOCK(sb),
		      journal->j_trans_max,
		      journal->j_max_batch,
		      journal->j_max_commit_age, journal->j_max_trans_age);

	brelse(bhjh);

	journal->j_list_bitmap_index = 0;
	journal_list_init(sb);

	memset(journal->j_list_hash_table, 0,
	       JOURNAL_HASH_SIZE * sizeof(struct reiserfs_journal_cnode *));

	INIT_LIST_HEAD(&journal->j_dirty_buffers);
	spin_lock_init(&journal->j_dirty_buffers_lock);

	journal->j_start = 0;
	journal->j_len = 0;
	journal->j_len_alloc = 0;
	atomic_set(&(journal->j_wcount), 0);
	atomic_set(&(journal->j_async_throttle), 0);
	journal->j_bcount = 0;
	journal->j_trans_start_time = 0;
	journal->j_last = NULL;
	journal->j_first = NULL;
	init_waitqueue_head(&(journal->j_join_wait));
	mutex_init(&journal->j_mutex);
	mutex_init(&journal->j_flush_mutex);

	journal->j_trans_id = 10;
	journal->j_mount_id = 10;
	journal->j_state = 0;
	atomic_set(&(journal->j_jlock), 0);
	reiserfs_write_unlock(sb);
	journal->j_cnode_free_list = allocate_cnodes(num_cnodes);
	reiserfs_write_lock(sb);
	journal->j_cnode_free_orig = journal->j_cnode_free_list;
	journal->j_cnode_free = journal->j_cnode_free_list ? num_cnodes : 0;
	journal->j_cnode_used = 0;
	journal->j_must_wait = 0;

	if (journal->j_cnode_free == 0) {
		reiserfs_warning(sb, "journal-2004", "Journal cnode memory "
		                 "allocation failed (%ld bytes). Journal is "
		                 "too large for available memory. Usually "
		                 "this is due to a journal that is too large.",
		                 sizeof (struct reiserfs_journal_cnode) * num_cnodes);
        	goto free_and_return;
	}

	init_journal_hash(sb);
	jl = journal->j_current_jl;
	jl->j_list_bitmap = get_list_bitmap(sb, jl);
	if (!jl->j_list_bitmap) {
		reiserfs_warning(sb, "journal-2005",
				 "get_list_bitmap failed for journal list 0");
		goto free_and_return;
	}
	if (journal_read(sb) < 0) {
		reiserfs_warning(sb, "reiserfs-2006",
				 "Replay Failure, unable to mount");
		goto free_and_return;
	}

	reiserfs_mounted_fs_count++;
	if (reiserfs_mounted_fs_count <= 1) {
		reiserfs_write_unlock(sb);
		commit_wq = alloc_workqueue("reiserfs", WQ_MEM_RECLAIM, 0);
		reiserfs_write_lock(sb);
	}

	INIT_DELAYED_WORK(&journal->j_work, flush_async_commits);
	journal->j_work_sb = sb;
	return 0;
      free_and_return:
	free_journal_ram(sb);
	return 1;
}

/*
** test for a polite end of the current transaction.  Used by file_write, and should
** be used by delete to make sure they don't write more than can fit inside a single
** transaction
*/
int journal_transaction_should_end(struct reiserfs_transaction_handle *th,
				   int new_alloc)
{
	struct reiserfs_journal *journal = SB_JOURNAL(th->t_super);
	time_t now = get_seconds();
	/* cannot restart while nested */
	BUG_ON(!th->t_trans_id);
	if (th->t_refcount > 1)
		return 0;
	if (journal->j_must_wait > 0 ||
	    (journal->j_len_alloc + new_alloc) >= journal->j_max_batch ||
	    atomic_read(&(journal->j_jlock)) ||
	    (now - journal->j_trans_start_time) > journal->j_max_trans_age ||
	    journal->j_cnode_free < (journal->j_trans_max * 3)) {
		return 1;
	}
	/* protected by the BKL here */
	journal->j_len_alloc += new_alloc;
	th->t_blocks_allocated += new_alloc ;
	return 0;
}

/* this must be called inside a transaction, and requires the
** kernel_lock to be held
*/
void reiserfs_block_writes(struct reiserfs_transaction_handle *th)
{
	struct reiserfs_journal *journal = SB_JOURNAL(th->t_super);
	BUG_ON(!th->t_trans_id);
	journal->j_must_wait = 1;
	set_bit(J_WRITERS_BLOCKED, &journal->j_state);
	return;
}

/* this must be called without a transaction started, and does not
** require BKL
*/
void reiserfs_allow_writes(struct super_block *s)
{
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	clear_bit(J_WRITERS_BLOCKED, &journal->j_state);
	wake_up(&journal->j_join_wait);
}

/* this must be called without a transaction started, and does not
** require BKL
*/
void reiserfs_wait_on_write_block(struct super_block *s)
{
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	wait_event(journal->j_join_wait,
		   !test_bit(J_WRITERS_BLOCKED, &journal->j_state));
}

static void queue_log_writer(struct super_block *s)
{
	wait_queue_t wait;
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	set_bit(J_WRITERS_QUEUED, &journal->j_state);

	/*
	 * we don't want to use wait_event here because
	 * we only want to wait once.
	 */
	init_waitqueue_entry(&wait, current);
	add_wait_queue(&journal->j_join_wait, &wait);
	set_current_state(TASK_UNINTERRUPTIBLE);
	if (test_bit(J_WRITERS_QUEUED, &journal->j_state)) {
		reiserfs_write_unlock(s);
		schedule();
		reiserfs_write_lock(s);
	}
	__set_current_state(TASK_RUNNING);
	remove_wait_queue(&journal->j_join_wait, &wait);
}

static void wake_queued_writers(struct super_block *s)
{
	struct reiserfs_journal *journal = SB_JOURNAL(s);
	if (test_and_clear_bit(J_WRITERS_QUEUED, &journal->j_state))
		wake_up(&journal->j_join_wait);
}

static void let_transaction_grow(struct super_block *sb, unsigned int trans_id)
{
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	unsigned long bcount = journal->j_bcount;
	while (1) {
		reiserfs_write_unlock(sb);
		schedule_timeout_uninterruptible(1);
		reiserfs_write_lock(sb);
		journal->j_current_jl->j_state |= LIST_COMMIT_PENDING;
		while ((atomic_read(&journal->j_wcount) > 0 ||
			atomic_read(&journal->j_jlock)) &&
		       journal->j_trans_id == trans_id) {
			queue_log_writer(sb);
		}
		if (journal->j_trans_id != trans_id)
			break;
		if (bcount == journal->j_bcount)
			break;
		bcount = journal->j_bcount;
	}
}

/* join == true if you must join an existing transaction.
** join == false if you can deal with waiting for others to finish
**
** this will block until the transaction is joinable.  send the number of blocks you
** expect to use in nblocks.
*/
static int do_journal_begin_r(struct reiserfs_transaction_handle *th,
			      struct super_block *sb, unsigned long nblocks,
			      int join)
{
	time_t now = get_seconds();
	unsigned int old_trans_id;
	struct reiserfs_journal *journal = SB_JOURNAL(sb);
	struct reiserfs_transaction_handle myth;
	int sched_count = 0;
	int retval;

	reiserfs_check_lock_depth(sb, "journal_begin");
	BUG_ON(nblocks > journal->j_trans_max);

	PROC_INFO_INC(sb, journal.journal_being);
	/* set here for journal_join */
	th->t_refcount = 1;
	th->t_super = sb;

      relock:
	lock_journal(sb);
	if (join != JBEGIN_ABORT && reiserfs_is_journal_aborted(journal)) {
		unlock_journal(sb);
		retval = journal->j_errno;
		goto out_fail;
	}
	journal->j_bcount++;

	if (test_bit(J_WRITERS_BLOCKED, &journal->j_state)) {
		unlock_journal(sb);
		reiserfs_write_unlock(sb);
		reiserfs_wait_on_write_block(sb);