Commit 6bfc09e232 ("rcu: Provide RCU CPU stall warnings for tiny RCU")
moved the rcu_cpu_stall_timeout module parameter from rcutree.c to
rcupdate.c, but failed to update Documentation/RCU/stallwarn.txt. This
commit therefore repairs this omission.
commit 96224daa16 ("documentation: Update sysfs path for rcu_cpu_stall_suppress")
updated the path for rcu_cpu_stall_suppress, but failed to update for
rcu_cpu_stall_timeout.
Signed-off-by: Xie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
PREEMPT_RCU and TREE_PREEMPT_RCU serve the same function after
TINY_PREEMPT_RCU has been removed. This patch removes TREE_PREEMPT_RCU
and uses PREEMPT_RCU config option in its place.
Signed-off-by: Pranith Kumar <bobby.prani@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The CONFIG_RCU_CPU_STALL_VERBOSE Kconfig parameter causes preemptible
RCU's CPU stall warnings to dump out any preempted tasks that are blocking
the current RCU grace period. This information is useful, and the default
has been CONFIG_RCU_CPU_STALL_VERBOSE=y for some years. It is therefore
time for this commit to remove this Kconfig parameter, so that future
kernel builds will always act as if CONFIG_RCU_CPU_STALL_VERBOSE=y.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
This commit documents RCU-tasks stall warning messages and also describes
when to use the new cond_resched_rcu_qs() API.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
All other add functions for lists have the new item as first argument
and the position where it is added as second argument. This was changed
for no good reason in this function and makes using it unnecessary
confusing.
The name was changed to hlist_add_behind() to cause unconverted code to
generate a compile error instead of using the wrong parameter order.
[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Ken Helias <kenhelias@firemail.de>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Acked-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> [intel driver bits]
Cc: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Reviewed-by: Lai Jiangshan <laijs@cn.fujitsu.com>
The kerneltrap.org site no longer works, so this commit updates it to
a working reference, namely gmane.
Signed-off-by: Pranith Kumar <bobby.prani@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
This commit updates the list of RCU API members in whatisRCU.txt.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Recent LKML discussings (see http://lwn.net/Articles/586838/ and
http://lwn.net/Articles/588300/ for the LWN writeups) brought out
some ways of misusing the return value from rcu_dereference() that
are not necessarily completely intuitive. This commit therefore
documents what can and cannot safely be done with these values.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Commit 6bfc09e232 (rcu: Provide RCU CPU stall warnings for tiny RCU)
moved the rcu_cpu_stall_suppress module parameter from rcutree.c to
rcupdate.c, but failed to update Documentation/RCU/stallwarn.txt. This
commit therefore repairs this omission.
Reported-by: Kirill Tkhai <tkhai@yandex.ru>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Some of the early history leaves out some citations and vice versa. This
commit fixes these up.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The call_rcu() family of primitives will take action to accelerate
grace periods when the number of callbacks pending on a given CPU
becomes excessive. Although this safety mechanism can be useful,
it is no substitute for users of call_rcu() having rate-limit controls
in place. This commit adds this nuance to the documentation.
Reported-by: "Michael S. Tsirkin" <mst@redhat.com>
Reported-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Some of the 00-INDEX files are somewhat outdated and some folders does
not contain 00-INDEX at all. Only outdated (with the notably exception
of spi) indexes are touched here, the 169 folders without 00-INDEX has
not been touched.
New 00-INDEX
- spi/* was added in a series of commits dating back to 2006
Added files (missing in (*/)00-INDEX)
- dmatest.txt was added by commit 851b7e16a0 ("dmatest: run test via
debugfs")
- this_cpu_ops.txt was added by commit a1b2a555d6 ("percpu: add
documentation on this_cpu operations")
- ww-mutex-design.txt was added by commit 040a0a3710 ("mutex: Add
support for wound/wait style locks")
- bcache.txt was added by commit cafe563591 ("bcache: A block layer
cache")
- kernel-per-CPU-kthreads.txt was added by commit 49717cb404
("kthread: Document ways of reducing OS jitter due to per-CPU
kthreads")
- phy.txt was added by commit ff76496347 ("drivers: phy: add generic
PHY framework")
- block/null_blk was added by commit 12f8f4fc03 ("null_blk:
documentation")
- module-signing.txt was added by commit 3cafea3076 ("Add
Documentation/module-signing.txt file")
- assoc_array.txt was added by commit 3cb989501c ("Add a generic
associative array implementation.")
- arm/IXP4xx was part of the initial repo
- arm/cluster-pm-race-avoidance.txt was added by commit 7fe31d28e8
("ARM: mcpm: introduce helpers for platform coherency exit/setup")
- arm/firmware.txt was added by commit 7366b92a77 ("ARM: Add
interface for registering and calling firmware-specific operations")
- arm/kernel_mode_neon.txt was added by commit 2afd0a0524 ("ARM:
7825/1: document the use of NEON in kernel mode")
- arm/tcm.txt was added by commit bc581770cf ("ARM: 5580/2: ARM TCM
(Tightly-Coupled Memory) support v3")
- arm/vlocks.txt was added by commit 9762f12d3e ("ARM: mcpm: Add
baremetal voting mutexes")
- blackfin/gptimers-example.c, Makefile was added by commit
4b60779d5e ("Blackfin: add an example showing how to use the
gptimers API")
- devicetree/usage-model.txt was added by commit 31134efc68 ("dt:
Linux DT usage model documentation")
- fb/api.txt was added by commit fb21c2f428 ("fbdev: Add FOURCC-based
format configuration API")
- fb/sm501.txt was added by commit e6a0498071 ("video, sm501: add
edid and commandline support")
- fb/udlfb.txt was added by commit 96f8d864af ("fbdev: move udlfb out
of staging.")
- filesystems/Makefile was added by commit 1e0051ae48
("Documentation/fs/: split txt and source files")
- filesystems/nfs/nfsd-admin-interfaces.txt was added by commit
8a4c6e19cf ("nfsd: document kernel interfaces for nfsd
configuration")
- ide/warm-plug-howto.txt was added by commit f74c91413e ("ide: add
warm-plug support for IDE devices (take 2)")
- laptops/Makefile was added by commit d49129accc
("Documentation/laptop/: split txt and source files")
- leds/leds-blinkm.txt was added by commit b54cf35a7f ("LEDS: add
BlinkM RGB LED driver, documentation and update MAINTAINERS")
- leds/ledtrig-oneshot.txt was added by commit 5e417281cd ("leds: add
oneshot trigger")
- leds/ledtrig-transient.txt was added by commit 44e1e9f8e7 ("leds:
add new transient trigger for one shot timer activation")
- m68k/README.buddha was part of the initial repo
- networking/LICENSE.(qla3xxx|qlcnic|qlge) was added by commits
40839129f7, c4e84bde1d, 5a4faa8737
- networking/Makefile was added by commit 3794f3e812 ("docsrc: build
Documentation/ sources")
- networking/i40evf.txt was added by commit 105bf2fe6b ("i40evf: add
driver to kernel build system")
- networking/ipsec.txt was added by commit b3c6efbc36 ("xfrm: Add
file to document IPsec corner case")
- networking/mac80211-auth-assoc-deauth.txt was added by commit
3cd7920a2b ("mac80211: add auth/assoc/deauth flow diagram")
- networking/netlink_mmap.txt was added by commit 5683264c39
("netlink: add documentation for memory mapped I/O")
- networking/nf_conntrack-sysctl.txt was added by commit c9f9e0e159
("netfilter: doc: add nf_conntrack sysctl api documentation") lan)
- networking/team.txt was added by commit 3d249d4ca7 ("net: introduce
ethernet teaming device")
- networking/vxlan.txt was added by commit d342894c5d ("vxlan:
virtual extensible lan")
- power/runtime_pm.txt was added by commit 5e928f77a0 ("PM: Introduce
core framework for run-time PM of I/O devices (rev. 17)")
- power/charger-manager.txt was added by commit 3bb3dbbd56
("power_supply: Add initial Charger-Manager driver")
- RCU/lockdep-splat.txt was added by commit d7bd2d68aa ("rcu:
Document interpretation of RCU-lockdep splats")
- s390/kvm.txt was added by 5ecee4b (KVM: s390: API documentation)
- s390/qeth.txt was added by commit b4d72c08b3 ("qeth: bridgeport
support - basic control")
- scheduler/sched-bwc.txt was added by commit 88ebc08ea9 ("sched: Add
documentation for bandwidth control")
- scsi/advansys.txt was added by commit 4bd6d7f356 ("[SCSI] advansys:
Move documentation to Documentation/scsi")
- scsi/bfa.txt was added by commit 1ec90174bd ("[SCSI] bfa: add
readme file")
- scsi/bnx2fc.txt was added by commit 12b8fc10ea ("[SCSI] bnx2fc: Add
driver documentation")
- scsi/cxgb3i.txt was added by commit c3673464eb ("[SCSI] cxgb3i: Add
cxgb3i iSCSI driver.")
- scsi/hpsa.txt was added by commit 992ebcf14f ("[SCSI] hpsa: Add
hpsa.txt to Documentation/scsi")
- scsi/link_power_management_policy.txt was added by commit
ca77329fb7 ("[libata] Link power management infrastructure")
- scsi/osd.txt was added by commit 78e0c621de ("[SCSI] osd:
Documentation for OSD library")
- scsi/scsi-parameter.txt was created/moved by commit 163475fb11
("Documentation: move SCSI parameters to their own text file")
- serial/driver was part of the initial repo
- serial/n_gsm.txt was added by commit 323e84122e ("n_gsm: add a
documentation")
- timers/Makefile was added by commit 3794f3e812 ("docsrc: build
Documentation/ sources")
- virt/kvm/s390.txt was added by commit d9101fca3d ("KVM: s390:
diagnose call documentation")
- vm/split_page_table_lock was added by commit 49076ec2cc ("mm:
dynamically allocate page->ptl if it cannot be embedded to struct
page")
- w1/slaves/w1_ds28e04 was added by commit fbf7f7b4e2 ("w1: Add
1-wire slave device driver for DS28E04-100")
- w1/masters/omap-hdq was added by commit e0a29382c6 ("hdq:
documentation for OMAP HDQ")
- x86/early-microcode.txt was added by commit 0d91ea86a8 ("x86, doc:
Documentation for early microcode loading")
- x86/earlyprintk.txt was added by commit a1aade4788 ("x86/doc:
mini-howto for using earlyprintk=dbgp")
- x86/entry_64.txt was added by commit 8b4777a4b5 ("x86-64: Document
some of entry_64.S")
- x86/pat.txt was added by commit d27554d874 ("x86: PAT
documentation")
Moved files
- arm/kernel_user_helpers.txt was moved out of arch/arm/kernel by
commit 37b8304642 ("ARM: kuser: move interface documentation out of
the source code")
- efi-stub.txt was moved out of x86/ and down into Documentation/ in
commit 4172fe2f8a ("EFI stub documentation updates")
- laptops/hpfall.c was moved out of hwmon/ and into laptops/ in commit
efcfed9bad ("Move hp_accel to drivers/platform/x86")
- commit 5616c23ad9 ("x86: doc: move x86-generic documentation from
Doc/x86/i386"):
* x86/usb-legacy-support.txt
* x86/boot.txt
* x86/zero_page.txt
- power/video_extension.txt was moved to acpi in commit 70e66e4df1
("ACPI / video: move video_extension.txt to Documentation/acpi")
Removed files (left in 00-INDEX)
- memory.txt was removed by commit 00ea8990aa ("memory.txt: remove
stray information")
- gpio.txt was moved to gpio/ in commit fd8e198cfc ("Documentation:
gpiolib: document new interface")
- networking/DLINK.txt was removed by commit 168e06ae26
("drivers/net: delete old parallel port de600/de620 drivers")
- serial/hayes-esp.txt was removed by commit f53a2ade0b ("tty: esp:
remove broken driver")
- s390/TAPE was removed by commit 9e280f6693 ("[S390] remove tape
block docu")
- vm/locking was removed by commit 57ea8171d2 ("mm: documentation:
remove hopelessly out-of-date locking doc")
- laptops/acer-wmi.txt was remvoed by commit 020036678e ("acer-wmi:
Delete out-of-date documentation")
Typos/misc issues
- rpc-server-gss.txt was added as knfsd-rpcgss.txt in commit
030d794bf4 ("SUNRPC: Use gssproxy upcall for server RPCGSS
authentication.")
- commit b88cf73d92 ("net: add missing entries to
Documentation/networking/00-INDEX")
* generic-hdlc.txt was added as generic_hdlc.txt
* spider_net.txt was added as spider-net.txt
- w1/master/mxc-w1 was added as mxc_w1 by commit a5fd9139f7 ("w1: add
1-wire master driver for i.MX27 / i.MX31")
- s390/zfcpdump.txt was added as zfcpdump by commit 6920c12a40
("[S390] Add Documentation/s390/00-INDEX.")
Signed-off-by: Henrik Austad <henrik@austad.us>
Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> [rcu bits]
Acked-by: Rob Landley <rob@landley.net>
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Mark Brown <broonie@kernel.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Gleb Natapov <gleb@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Len Brown <len.brown@intel.com>
Cc: James Bottomley <JBottomley@parallels.com>
Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
doc.2013.12.03a: Topic branch for documentation changes.
fixes.2013.12.12a: Topic branch for miscellaneous fixes.
rcutorture.2013.12.03a: Topic branch for new rcutorture/KVM scripting.
sparse.2013.12.12a: Topic branch for sparse-RCU changes.
Dave Jones got the following lockdep splat:
> ======================================================
> [ INFO: possible circular locking dependency detected ]
> 3.12.0-rc3+ #92 Not tainted
> -------------------------------------------------------
> trinity-child2/15191 is trying to acquire lock:
> (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50
>
> but task is already holding lock:
> (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230
>
> which lock already depends on the new lock.
>
>
> the existing dependency chain (in reverse order) is:
>
> -> #3 (&ctx->lock){-.-...}:
> [<ffffffff810cc243>] lock_acquire+0x93/0x200
> [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80
> [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0
> [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0
> [<ffffffff81732052>] __schedule+0x1d2/0xa20
> [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0
> [<ffffffff817352b6>] retint_kernel+0x26/0x30
> [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50
> [<ffffffff813f0504>] pty_write+0x54/0x60
> [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0
> [<ffffffff813e5838>] tty_write+0x158/0x2d0
> [<ffffffff811c4850>] vfs_write+0xc0/0x1f0
> [<ffffffff811c52cc>] SyS_write+0x4c/0xa0
> [<ffffffff8173d4e4>] tracesys+0xdd/0xe2
>
> -> #2 (&rq->lock){-.-.-.}:
> [<ffffffff810cc243>] lock_acquire+0x93/0x200
> [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80
> [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0
> [<ffffffff81054336>] do_fork+0x126/0x460
> [<ffffffff81054696>] kernel_thread+0x26/0x30
> [<ffffffff8171ff93>] rest_init+0x23/0x140
> [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403
> [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c
> [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4
>
> -> #1 (&p->pi_lock){-.-.-.}:
> [<ffffffff810cc243>] lock_acquire+0x93/0x200
> [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90
> [<ffffffff810979d1>] try_to_wake_up+0x31/0x350
> [<ffffffff81097d62>] default_wake_function+0x12/0x20
> [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40
> [<ffffffff8108ea38>] __wake_up_common+0x58/0x90
> [<ffffffff8108ff59>] __wake_up+0x39/0x50
> [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0
> [<ffffffff81111450>] __call_rcu+0x140/0x820
> [<ffffffff81111b8d>] call_rcu+0x1d/0x20
> [<ffffffff81093697>] cpu_attach_domain+0x287/0x360
> [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0
> [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a
> [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202
> [<ffffffff817200be>] kernel_init+0xe/0x190
> [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0
>
> -> #0 (&rdp->nocb_wq){......}:
> [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0
> [<ffffffff810cc243>] lock_acquire+0x93/0x200
> [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90
> [<ffffffff8108ff43>] __wake_up+0x23/0x50
> [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0
> [<ffffffff81111450>] __call_rcu+0x140/0x820
> [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30
> [<ffffffff81149abf>] put_ctx+0x4f/0x70
> [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230
> [<ffffffff81056b8d>] do_exit+0x30d/0xcc0
> [<ffffffff8105893c>] do_group_exit+0x4c/0xc0
> [<ffffffff810589c4>] SyS_exit_group+0x14/0x20
> [<ffffffff8173d4e4>] tracesys+0xdd/0xe2
>
> other info that might help us debug this:
>
> Chain exists of:
> &rdp->nocb_wq --> &rq->lock --> &ctx->lock
>
> Possible unsafe locking scenario:
>
> CPU0 CPU1
> ---- ----
> lock(&ctx->lock);
> lock(&rq->lock);
> lock(&ctx->lock);
> lock(&rdp->nocb_wq);
>
> *** DEADLOCK ***
>
> 1 lock held by trinity-child2/15191:
> #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230
>
> stack backtrace:
> CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92
> ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40
> ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0
> ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0
> Call Trace:
> [<ffffffff8172a363>] dump_stack+0x4e/0x82
> [<ffffffff81726741>] print_circular_bug+0x200/0x20f
> [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0
> [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60
> [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80
> [<ffffffff810cc243>] lock_acquire+0x93/0x200
> [<ffffffff8108ff43>] ? __wake_up+0x23/0x50
> [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90
> [<ffffffff8108ff43>] ? __wake_up+0x23/0x50
> [<ffffffff8108ff43>] __wake_up+0x23/0x50
> [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0
> [<ffffffff81111450>] __call_rcu+0x140/0x820
> [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50
> [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30
> [<ffffffff81149abf>] put_ctx+0x4f/0x70
> [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230
> [<ffffffff81056b8d>] do_exit+0x30d/0xcc0
> [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0
> [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10
> [<ffffffff8105893c>] do_group_exit+0x4c/0xc0
> [<ffffffff810589c4>] SyS_exit_group+0x14/0x20
> [<ffffffff8173d4e4>] tracesys+0xdd/0xe2
The underlying problem is that perf is invoking call_rcu() with the
scheduler locks held, but in NOCB mode, call_rcu() will with high
probability invoke the scheduler -- which just might want to use its
locks. The reason that call_rcu() needs to invoke the scheduler is
to wake up the corresponding rcuo callback-offload kthread, which
does the job of starting up a grace period and invoking the callbacks
afterwards.
One solution (championed on a related problem by Lai Jiangshan) is to
simply defer the wakeup to some point where scheduler locks are no longer
held. Since we don't want to unnecessarily incur the cost of such
deferral, the task before us is threefold:
1. Determine when it is likely that a relevant scheduler lock is held.
2. Defer the wakeup in such cases.
3. Ensure that all deferred wakeups eventually happen, preferably
sooner rather than later.
We use irqs_disabled_flags() as a proxy for relevant scheduler locks
being held. This works because the relevant locks are always acquired
with interrupts disabled. We may defer more often than needed, but that
is at least safe.
The wakeup deferral is tracked via a new field in the per-CPU and
per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup.
This flag is checked by the RCU core processing. The __rcu_pending()
function now checks this flag, which causes rcu_check_callbacks()
to initiate RCU core processing at each scheduling-clock interrupt
where this flag is set. Of course this is not sufficient because
scheduling-clock interrupts are often turned off (the things we used to
be able to count on!). So the flags are also checked on entry to any
state that RCU considers to be idle, which includes both NO_HZ_IDLE idle
state and NO_HZ_FULL user-mode-execution state.
This approach should allow call_rcu() to be invoked regardless of what
locks you might be holding, the key word being "should".
Reported-by: Dave Jones <davej@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Michael Opdenacker <michael.opdenacker@free-electrons.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
[ paulmck: Add "then" as suggested by Josh Triplett. ]
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Currently, rcutorture has separate torture_types to test synchronous,
asynchronous, and expedited grace-period primitives. This has
two disadvantages: (1) Three times the number of runs to cover the
combinations and (2) Little testing of concurrent combinations of the
three options. This commit therefore adds a pair of module parameters
that control normal and expedited state, with the default being both
types, randomly selected, by the fakewriter processes, thus reducing
source-code size and increasing test coverage. In addtion, the writer
task switches between asynchronous-normal and expedited grace-period
primitives driven by the same pair of module parameters.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Note that this commit also updates the formatting of serveral of the
bibtex entries to conform to that of my .bib files. I started
accumulating entries back in the 1980s, back when bibtex insisted
that comma (",") was a separator, not a terminator. This rule forced
commas to the fronts of lines. 25 years later, bibtex allows commas
to be terminators, but I am too lazy to rework all my .bib files.
Keeping the same format as my .bib files allows my to simply
incorporate my RCU.bib file into Documentation/RCU/RTFP.txt, which
is much easier than my earlier practice of keeping track of what
had changed and adding individual entries. (I sometimes find relevant
papers that were published some years back, for example.)
In addition, this change adds entries for papers published in the
last year or so.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
There was a time when rcu_barrier() was guaranteed to wait for at least
a grace period, but that time ended due to energy-efficiency concerns.
So now rcu_barrier() is a no-op if there are no RCU callbacks queued in
the system. This commit updates the documentation to reflect this change.
Now, rcu_barrier() often does wait for a grace period, so, one could
imagine some modification to rcu_barrier() to more efficiently handle
cases where both rcu_barrier() and a grace period are needed. But this
must wait until someone shows a real-world need for a change.
Reported-by: Bob Copeland <bob@cozybit.com>
Reported-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Because TINY_PREEMPT_RCU is no more, this commit removes its tracing
formats from the documentation.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
These interfaces never did get used, so this commit removes them,
their rcutorture tests, and documentation referencing them.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The full dynticks tree needs the latest RCU and sched
upstream updates in order to fix some dependencies.
Merge a common upstream merge point that has these
updates.
Conflicts:
include/linux/perf_event.h
kernel/rcutree.h
kernel/rcutree_plugin.h
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
We are planning to convert the dynticks Kconfig options layout
into a choice menu. The user must be able to easily pick
any of the following implementations: constant periodic tick,
idle dynticks, full dynticks.
As this implies a mutual exclusion, the two dynticks implementions
need to converge on the selection of a common Kconfig option in order
to ease the sharing of a common infrastructure.
It would thus seem pretty natural to reuse CONFIG_NO_HZ to
that end. It already implements all the idle dynticks code
and the full dynticks depends on all that code for now.
So ideally the choice menu would propose CONFIG_NO_HZ_IDLE and
CONFIG_NO_HZ_EXTENDED then both would select CONFIG_NO_HZ.
On the other hand we want to stay backward compatible: if
CONFIG_NO_HZ is set in an older config file, we want to
enable CONFIG_NO_HZ_IDLE by default.
But we can't afford both at the same time or we run into
a circular dependency:
1) CONFIG_NO_HZ_IDLE and CONFIG_NO_HZ_EXTENDED both select
CONFIG_NO_HZ
2) If CONFIG_NO_HZ is set, we default to CONFIG_NO_HZ_IDLE
We might be able to support that from Kconfig/Kbuild but it
may not be wise to introduce such a confusing behaviour.
So to solve this, create a new CONFIG_NO_HZ_COMMON option
which gathers the common code between idle and full dynticks
(that common code for now is simply the idle dynticks code)
and select it from their referring Kconfig.
Then we'll later create CONFIG_NO_HZ_IDLE and map CONFIG_NO_HZ
to it for backward compatibility.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Chris Metcalf <cmetcalf@tilera.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Geoff Levand <geoff@infradead.org>
Cc: Gilad Ben Yossef <gilad@benyossef.com>
Cc: Hakan Akkan <hakanakkan@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Kevin Hilman <khilman@linaro.org>
Cc: Li Zhong <zhong@linux.vnet.ibm.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
doc.2013.03.12a: Documentation changes.
fixes.2013.03.13a: Miscellaneous fixes.
idlenocb.2013.03.26b: Remove restrictions on no-CBs CPUs, make
RCU_FAST_NO_HZ take advantage of numbered callbacks, add
callback acceleration based on numbered callbacks.
If RCU's softirq handler is prevented from executing, an RCU CPU stall
warning can result. Ways to prevent RCU's softirq handler from executing
include: (1) CPU spinning with interrupts disabled, (2) infinite loop
in some softirq handler, and (3) in -rt kernels, an infinite loop in a
set of real-time threads running at priorities higher than that of RCU's
softirq handler.
Because this situation can be difficult to track down, this commit causes
the count of RCU softirq handler invocations to be printed with RCU
CPU stall warnings. This information does require some interpretation,
as now documented in Documentation/RCU/stallwarn.txt.
Reported-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Paul Gortmaker <paul.gortmaker@windriver.com>
One of the code samples in whatisRCU.txt shows a bug, but someone scanning
the document quickly might mistake it for a valid use of RCU. Add some
screaming comments to help keep speed-readers on track.
Reported-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
urgent.2012.10.27a: Fix for RCU user-mode transition (already in -tip).
doc.2012.11.08a: Documentation updates, most notably codifying the
memory-barrier guarantees inherent to grace periods.
fixes.2012.11.13a: Miscellaneous fixes.
srcu.2012.10.27a: Allow statically allocated and initialized srcu_struct
structures (courtesy of Lai Jiangshan).
stall.2012.11.13a: Add more diagnostic information to RCU CPU stall
warnings, also decrease from 60 seconds to 21 seconds.
hotplug.2012.11.08a: Minor updates to CPU hotplug handling.
tracing.2012.11.08a: Improved debugfs tracing, courtesy of Michael Wang.
idle.2012.10.24a: Updates to RCU idle/adaptive-idle handling, including
a boot parameter that maps normal grace periods to expedited.
Resolved conflict in kernel/rcutree.c due to side-by-side change.
This commit adds the documentation of the rcuexp debugfs trace file
that records statistics for expedited grace periods.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
This commit updates the tracing documentation to reflect the new
format that has per-RCU-flavor directories.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The list_for_each_continue_rcu() macro is no longer used, so this commit
removes it. The list_for_each_entry_continue_rcu() macro should be
used instead.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The approach for mixing RCU and reference counting listed in the RCU
documentation only describes one possible approach. This approach can
result in failure on the read side, which is nice if you want fresh data,
but not so good if you want simple code. This commit therefore adds
two additional approaches that feature unconditional reference-count
acquisition by RCU readers. These approaches are very similar to that
used in the security code.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Mention kfree_rcu() in the call_rcu() section. Additionally fix the
example code for list replacement that used the wrong structure element.
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Trying to go through the history of RCU (not for the weak
minded) led me to search for a non-existent paper.
Correct it to the actual reference
Signed-off-by: Dhaval Giani <dhaval.giani@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
bigrt.2012.09.23a contains additional commits to reduce scheduling latency
from RCU on huge systems (many hundrends or thousands of CPUs).
doctorture.2012.09.23a contains documentation changes and rcutorture fixes.
fixes.2012.09.23a contains miscellaneous fixes.
hotplug.2012.09.23a contains CPU-hotplug-related changes.
idle.2012.09.23a fixes architectures for which RCU no longer considered
the idle loop to be a quiescent state due to earlier
adaptive-dynticks changes. Affected architectures are alpha,
cris, frv, h8300, m32r, m68k, mn10300, parisc, score, xtensa,
and ia64.
The print_cpu_stall_fast_no_hz() function attempts to print -1 when
the ->idle_gp_timer is not pending, but unsigned arithmetic causes it
to instead print ULONG_MAX, which is 4294967295 on 32-bit systems and
18446744073709551615 on 64-bit systems. Neither of these are the most
reader-friendly values, so this commit instead causes "timer not pending"
to be printed when ->idle_gp_timer is not pending.
Reported-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The current documentation did not help someone grepping for SRCU to
learn that disabling preemption is not a replacement for srcu_read_lock(),
so upgrade the documentation to bring this out, not just for SRCU,
but also for RCU-bh. Also document the fact that SRCU readers are
respected on CPUs executing in user mode, idle CPUs, and even on
offline CPUs.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Reviewed-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Moving quiescent-state forcing into a kthread dispenses with the need
for the ->n_rp_need_fqs field, so this commit removes it.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The advent of call_srcu() and srcu_barrier() obsoleted some of the
documentation, so this commit brings that up to date.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Although rcutorture does invoke rcu_barrier() and friends, it cannot
really be called a torture test given that it invokes them only once
at the end of the test. This commit therefore introduces heavy-duty
rcutorture testing for rcu_barrier(), which may be carried out
concurrently with normal rcutorture testing.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The expedited RCU primitives can be quite useful, but they have some
high costs as well. This commit updates and creates docbook comments
calling out the costs, and updates the RCU documentation as well.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Because newly offlined CPUs continue executing after completing the
CPU_DYING notifiers, they legitimately enter the scheduler and use
RCU while appearing to be offline. This calls for a more sophisticated
approach as follows:
1. RCU marks the CPU online during the CPU_UP_PREPARE phase.
2. RCU marks the CPU offline during the CPU_DEAD phase.
3. Diagnostics regarding use of read-side RCU by offline CPUs use
RCU's accounting rather than the cpu_online_map. (Note that
__call_rcu() still uses cpu_online_map to detect illegal
invocations within CPU_DYING notifiers.)
4. Offline CPUs are prevented from hanging the system by
force_quiescent_state(), which pays attention to cpu_online_map.
Some additional work (in a later commit) will be needed to
guarantee that force_quiescent_state() waits a full jiffy before
assuming that a CPU is offline, for example, when called from
idle entry. (This commit also makes the one-jiffy wait
explicit, since the old-style implicit wait can now be defeated
by RCU_FAST_NO_HZ and by rcutorture.)
This approach avoids the false positives encountered when attempting to
use more exact classification of CPU online/offline state.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Add documentation of CONFIG_RCU_CPU_STALL_VERBOSE, CONFIG_RCU_CPU_STALL_INFO,
and RCU_STALL_DELAY_DELTA. Describe multiple stall-warning messages from
a single stall, and the timing of the subsequent messages. Add headings.
Remove RCU_SECONDS_TILL_STALL_RECHECK because this value is now computed
at runtime from RCU_CPU_STALL_TIMEOUT, so that sysfs changes to the timeout
value now directly affect the RCU_SECONDS_TILL_STALL_RECHECK value.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Add module parameters to rcutorture that induce a CPU stall.
The stall_cpu parameter specifies how long to stall in seconds,
defaulting to zero, which indicates no stalling is to be undertaken.
The stall_cpu_holdoff parameter specifies how many seconds after
insmod (or boot, if rcutorture is built into the kernel) that this
stall is to start. The default value for stall_cpu_holdoff is ten
seconds.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The torture.txt documentation gives an example rcutorture run with a
100-second duration. This is ridiculously short, unless maybe testing
a fix for a egregious bug. Use a more-realistic one-hour duration for
the example.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
When rcutorture is started automatically at boot time, it might well
also start CPU-hotplug operations at that time, which might not be
desirable. This commit therefore adds an rcutorture parameter that
allows CPU-hotplug operations to be held off for the specified number
of seconds after the start of boot.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tyler Hicks pointed me at an additional article on RCU and I figured
it should probably be mentioned with the others.
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Running CPU-hotplug operations concurrently with rcutorture has
historically been a good way to find bugs in both RCU and CPU hotplug.
This commit therefore adds an rcutorture module parameter called
"onoff_interval" that causes a randomly selected CPU-hotplug operation to
be executed at the specified interval, in seconds. The default value of
"onoff_interval" is zero, which disables rcutorture-instigated CPU-hotplug
operations.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Although it is easy to run rcutorture tests under KVM, there is currently
no nice way to run such a test for a fixed time period, collect all of
the rcutorture data, and then shut the system down cleanly. This commit
therefore adds an rcutorture module parameter named "shutdown_secs" that
specified the run duration in seconds, after which rcutorture terminates
the test and powers the system down. The default value for "shutdown_secs"
is zero, which disables shutdown.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Update various files in Documentation/RCU to reflect srcu_read_lock_raw()
and srcu_read_unlock_raw(). Credit to Peter Zijlstra for suggesting
use of the existing _raw suffix instead of the earlier bulkref names.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
One of lclaudio's systems was seeing RCU CPU stall warnings from idle.
These turned out to be caused by a bug that stopped scheduling-clock
tick interrupts from being sent to a given CPU for several hundred seconds.
This commit therefore updates the documentation to call this out as a
possible cause for RCU CPU stall warnings.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Earlier versions of RCU used the scheduling-clock tick to detect idleness
by checking for the idle task, but handled idleness differently for
CONFIG_NO_HZ=y. But there are now a number of uses of RCU read-side
critical sections in the idle task, for example, for tracing. A more
fine-grained detection of idleness is therefore required.
This commit presses the old dyntick-idle code into full-time service,
so that rcu_idle_enter(), previously known as rcu_enter_nohz(), is
always invoked at the beginning of an idle loop iteration. Similarly,
rcu_idle_exit(), previously known as rcu_exit_nohz(), is always invoked
at the end of an idle-loop iteration. This allows the idle task to
use RCU everywhere except between consecutive rcu_idle_enter() and
rcu_idle_exit() calls, in turn allowing architecture maintainers to
specify exactly where in the idle loop that RCU may be used.
Because some of the userspace upcall uses can result in what looks
to RCU like half of an interrupt, it is not possible to expect that
the irq_enter() and irq_exit() hooks will give exact counts. This
patch therefore expands the ->dynticks_nesting counter to 64 bits
and uses two separate bitfields to count process/idle transitions
and interrupt entry/exit transitions. It is presumed that userspace
upcalls do not happen in the idle loop or from usermode execution
(though usermode might do a system call that results in an upcall).
The counter is hard-reset on each process/idle transition, which
avoids the interrupt entry/exit error from accumulating. Overflow
is avoided by the 64-bitness of the ->dyntick_nesting counter.
This commit also adds warnings if a non-idle task asks RCU to enter
idle state (and these checks will need some adjustment before applying
Frederic's OS-jitter patches (http://lkml.org/lkml/2011/10/7/246).
In addition, validation of ->dynticks and ->dynticks_nesting is added.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
There has been quite a bit of confusion about what RCU-lockdep splats
mean, so this commit adds some documentation describing how to
interpret them.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Add documentation for rcu_dereference_bh_check(),
rcu_dereference_sched_check(), srcu_dereference_check(), and
rcu_dereference_index_check().
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Since ca5ecddf (rcu: define __rcu address space modifier for sparse)
rcu_dereference_check() use rcu_read_lock_held() as a part of condition
automatically. Therefore, callers of rcu_dereference_check() no longer
need to pass rcu_read_lock_held() to rcu_dereference_check().
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
There is often a delay between the time that a CPU passes through a
quiescent state and the time that this quiescent state is reported to the
RCU core. It is quite possible that the grace period ended before the
quiescent state could be reported, for example, some other CPU might have
deduced that this CPU passed through dyntick-idle mode. It is critically
important that quiescent state be counted only against the grace period
that was in effect at the time that the quiescent state was detected.
Previously, this was handled by recording the number of the last grace
period to complete when passing through a quiescent state. The RCU
core then checks this number against the current value, and rejects
the quiescent state if there is a mismatch. However, one additional
possibility must be accounted for, namely that the quiescent state was
recorded after the prior grace period completed but before the current
grace period started. In this case, the RCU core must reject the
quiescent state, but the recorded number will match. This is handled
when the CPU becomes aware of a new grace period -- at that point,
it invalidates any prior quiescent state.
This works, but is a bit indirect. The new approach records the current
grace period, and the RCU core checks to see (1) that this is still the
current grace period and (2) that this grace period has not yet ended.
This approach simplifies reasoning about correctness, and this commit
changes over to this new approach.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
It has long been the case that the architecture must call nmi_enter()
and nmi_exit() rather than irq_enter() and irq_exit() in order to
permit RCU read-side critical sections in NMIs. Catch the documentation
up with reality.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Now that the RCU API contains synchronize_rcu_bh(), synchronize_sched(),
call_rcu_sched(), and rcu_bh_expedited()...
Make rcutorture test synchronize_rcu_bh(), getting rid of the old
rcu_bh_torture_synchronize() workaround. Similarly, make rcutorture test
synchronize_sched(), getting rid of the old sched_torture_synchronize()
workaround. Make rcutorture test call_rcu_sched() instead of wrappering
synchronize_sched(). Also add testing of rcu_bh_expedited().
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Update rcutorture documentation to account for boosting, new types of
RCU torture testing that have been added over the past few years, and
the memory-barrier testing that was added an embarrassingly long time
ago.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Call out the RCU_TRACE information that is provided only in kernels
built with RCU_BOOST.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Change all "arch/i386" to "arch/x86" in Documentaion/,
since the directory has changed.
Also update the files which have changed their filename
in the meantime accordingly.
Signed-off-by: Wanlong Gao <wanlong.gao@gmail.com>
[jkosina@suse.cz: reword changelog]
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
(Note: this was reverted, and is now being re-applied in pieces, with
this being the fifth and final piece. See below for the reason that
it is now felt to be safe to re-apply this.)
Commit d09b62d fixed grace-period synchronization, but left some smp_mb()
invocations in rcu_process_callbacks() that are no longer needed, but
sheer paranoia prevented them from being removed. This commit removes
them and provides a proof of correctness in their absence. It also adds
a memory barrier to rcu_report_qs_rsp() immediately before the update to
rsp->completed in order to handle the theoretical possibility that the
compiler or CPU might move massive quantities of code into a lock-based
critical section. This also proves that the sheer paranoia was not
entirely unjustified, at least from a theoretical point of view.
In addition, the old dyntick-idle synchronization depended on the fact
that grace periods were many milliseconds in duration, so that it could
be assumed that no dyntick-idle CPU could reorder a memory reference
across an entire grace period. Unfortunately for this design, the
addition of expedited grace periods breaks this assumption, which has
the unfortunate side-effect of requiring atomic operations in the
functions that track dyntick-idle state for RCU. (There is some hope
that the algorithms used in user-level RCU might be applied here, but
some work is required to handle the NMIs that user-space applications
can happily ignore. For the short term, better safe than sorry.)
This proof assumes that neither compiler nor CPU will allow a lock
acquisition and release to be reordered, as doing so can result in
deadlock. The proof is as follows:
1. A given CPU declares a quiescent state under the protection of
its leaf rcu_node's lock.
2. If there is more than one level of rcu_node hierarchy, the
last CPU to declare a quiescent state will also acquire the
->lock of the next rcu_node up in the hierarchy, but only
after releasing the lower level's lock. The acquisition of this
lock clearly cannot occur prior to the acquisition of the leaf
node's lock.
3. Step 2 repeats until we reach the root rcu_node structure.
Please note again that only one lock is held at a time through
this process. The acquisition of the root rcu_node's ->lock
must occur after the release of that of the leaf rcu_node.
4. At this point, we set the ->completed field in the rcu_state
structure in rcu_report_qs_rsp(). However, if the rcu_node
hierarchy contains only one rcu_node, then in theory the code
preceding the quiescent state could leak into the critical
section. We therefore precede the update of ->completed with a
memory barrier. All CPUs will therefore agree that any updates
preceding any report of a quiescent state will have happened
before the update of ->completed.
5. Regardless of whether a new grace period is needed, rcu_start_gp()
will propagate the new value of ->completed to all of the leaf
rcu_node structures, under the protection of each rcu_node's ->lock.
If a new grace period is needed immediately, this propagation
will occur in the same critical section that ->completed was
set in, but courtesy of the memory barrier in #4 above, is still
seen to follow any pre-quiescent-state activity.
6. When a given CPU invokes __rcu_process_gp_end(), it becomes
aware of the end of the old grace period and therefore makes
any RCU callbacks that were waiting on that grace period eligible
for invocation.
If this CPU is the same one that detected the end of the grace
period, and if there is but a single rcu_node in the hierarchy,
we will still be in the single critical section. In this case,
the memory barrier in step #4 guarantees that all callbacks will
be seen to execute after each CPU's quiescent state.
On the other hand, if this is a different CPU, it will acquire
the leaf rcu_node's ->lock, and will again be serialized after
each CPU's quiescent state for the old grace period.
On the strength of this proof, this commit therefore removes the memory
barriers from rcu_process_callbacks() and adds one to rcu_report_qs_rsp().
The effect is to reduce the number of memory barriers by one and to
reduce the frequency of execution from about once per scheduling tick
per CPU to once per grace period.
This was reverted do to hangs found during testing by Yinghai Lu and
Ingo Molnar. Frederic Weisbecker supplied Yinghai with tracing that
located the underlying problem, and Frederic also provided the fix.
The underlying problem was that the HARDIRQ_ENTER() macro from
lib/locking-selftest.c invoked irq_enter(), which in turn invokes
rcu_irq_enter(), but HARDIRQ_EXIT() invoked __irq_exit(), which
does not invoke rcu_irq_exit(). This situation resulted in calls
to rcu_irq_enter() that were not balanced by the required calls to
rcu_irq_exit(). Therefore, after these locking selftests completed,
RCU's dyntick-idle nesting count was a large number (for example,
72), which caused RCU to to conclude that the affected CPU was not in
dyntick-idle mode when in fact it was.
RCU would therefore incorrectly wait for this dyntick-idle CPU, resulting
in hangs.
In contrast, with Frederic's patch, which replaces the irq_enter()
in HARDIRQ_ENTER() with an __irq_enter(), these tests don't ever call
either rcu_irq_enter() or rcu_irq_exit(), which works because the CPU
running the test is already marked as not being in dyntick-idle mode.
This means that the rcu_irq_enter() and rcu_irq_exit() calls and RCU
then has no problem working out which CPUs are in dyntick-idle mode and
which are not.
The reason that the imbalance was not noticed before the barrier patch
was applied is that the old implementation of rcu_enter_nohz() ignored
the nesting depth. This could still result in delays, but much shorter
ones. Whenever there was a delay, RCU would IPI the CPU with the
unbalanced nesting level, which would eventually result in rcu_enter_nohz()
being called, which in turn would force RCU to see that the CPU was in
dyntick-idle mode.
The reason that very few people noticed the problem is that the mismatched
irq_enter() vs. __irq_exit() occured only when the kernel was built with
CONFIG_DEBUG_LOCKING_API_SELFTESTS.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
This reverts commit e59fb3120b.
This reversion was due to (extreme) boot-time slowdowns on SPARC seen by
Yinghai Lu and on x86 by Ingo
.
This is a non-trivial reversion due to intervening commits.
Conflicts:
Documentation/RCU/trace.txt
kernel/rcutree.c
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Increment a per-CPU counter on each pass through rcu_cpu_kthread()'s
service loop, and add it to the rcudata trace output.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
This commit adds the age in jiffies of the current grace period along
with the duration in jiffies of the longest grace period since boot
to the rcu/rcugp debugfs file. It also adds an additional "O" state
to kthread tracing to differentiate between the kthread waiting due to
having nothing to do on the one hand and waiting due to being on the
wrong CPU on the other hand.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
This commit documents the new debugfs rcu/rcutorture and rcu/rcuboost
trace files. The description has been updated as suggested by Josh
Triplett.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
This commit adds an indication of the state of the callback queue using
a string of four characters following the "ql=" integer queue length.
The first character is "N" if there are callbacks that have been
queued that are not yet ready to be handled by the next grace period, or
"." otherwise. The second character is "R" if there are callbacks queued
that are ready to be handled by the next grace period, or "." otherwise.
The third character is "W" if there are callbacks waiting for the current
grace period, or "." otherwise. Finally, the fourth character is "D"
if there are callbacks that have been handled by a prior grace period
and are waiting to be invoked, or ".".
Note that callbacks that are in the process of being invoked are
not shown. These callbacks would have been removed from the rcu_data
structure's list by rcu_do_batch() prior to being executed. (These
callbacks are also not reflected in the "ql=" total, FWIW.)
Also, document the new callback-queue trace information.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The trace.txt file had obsolete output for the debugfs rcu/rcudata
file, so update it.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Combine the current TREE_PREEMPT_RCU ->blocked_tasks[] lists in the
rcu_node structure into a single ->blkd_tasks list with ->gp_tasks
and ->exp_tasks tail pointers. This is in preparation for RCU priority
boosting, which will add a third dimension to the combinatorial explosion
in the ->blocked_tasks[] case, but simply a third pointer in the new
->blkd_tasks case.
Also update documentation to reflect blocked_tasks[] merge
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Commit d09b62d fixed grace-period synchronization, but left some smp_mb()
invocations in rcu_process_callbacks() that are no longer needed, but
sheer paranoia prevented them from being removed. This commit removes
them and provides a proof of correctness in their absence. It also adds
a memory barrier to rcu_report_qs_rsp() immediately before the update to
rsp->completed in order to handle the theoretical possibility that the
compiler or CPU might move massive quantities of code into a lock-based
critical section. This also proves that the sheer paranoia was not
entirely unjustified, at least from a theoretical point of view.
In addition, the old dyntick-idle synchronization depended on the fact
that grace periods were many milliseconds in duration, so that it could
be assumed that no dyntick-idle CPU could reorder a memory reference
across an entire grace period. Unfortunately for this design, the
addition of expedited grace periods breaks this assumption, which has
the unfortunate side-effect of requiring atomic operations in the
functions that track dyntick-idle state for RCU. (There is some hope
that the algorithms used in user-level RCU might be applied here, but
some work is required to handle the NMIs that user-space applications
can happily ignore. For the short term, better safe than sorry.)
This proof assumes that neither compiler nor CPU will allow a lock
acquisition and release to be reordered, as doing so can result in
deadlock. The proof is as follows:
1. A given CPU declares a quiescent state under the protection of
its leaf rcu_node's lock.
2. If there is more than one level of rcu_node hierarchy, the
last CPU to declare a quiescent state will also acquire the
->lock of the next rcu_node up in the hierarchy, but only
after releasing the lower level's lock. The acquisition of this
lock clearly cannot occur prior to the acquisition of the leaf
node's lock.
3. Step 2 repeats until we reach the root rcu_node structure.
Please note again that only one lock is held at a time through
this process. The acquisition of the root rcu_node's ->lock
must occur after the release of that of the leaf rcu_node.
4. At this point, we set the ->completed field in the rcu_state
structure in rcu_report_qs_rsp(). However, if the rcu_node
hierarchy contains only one rcu_node, then in theory the code
preceding the quiescent state could leak into the critical
section. We therefore precede the update of ->completed with a
memory barrier. All CPUs will therefore agree that any updates
preceding any report of a quiescent state will have happened
before the update of ->completed.
5. Regardless of whether a new grace period is needed, rcu_start_gp()
will propagate the new value of ->completed to all of the leaf
rcu_node structures, under the protection of each rcu_node's ->lock.
If a new grace period is needed immediately, this propagation
will occur in the same critical section that ->completed was
set in, but courtesy of the memory barrier in #4 above, is still
seen to follow any pre-quiescent-state activity.
6. When a given CPU invokes __rcu_process_gp_end(), it becomes
aware of the end of the old grace period and therefore makes
any RCU callbacks that were waiting on that grace period eligible
for invocation.
If this CPU is the same one that detected the end of the grace
period, and if there is but a single rcu_node in the hierarchy,
we will still be in the single critical section. In this case,
the memory barrier in step #4 guarantees that all callbacks will
be seen to execute after each CPU's quiescent state.
On the other hand, if this is a different CPU, it will acquire
the leaf rcu_node's ->lock, and will again be serialized after
each CPU's quiescent state for the old grace period.
On the strength of this proof, this commit therefore removes the memory
barriers from rcu_process_callbacks() and adds one to rcu_report_qs_rsp().
The effect is to reduce the number of memory barriers by one and to
reduce the frequency of execution from about once per scheduling tick
per CPU to once per grace period.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The RCU CPU stall warnings can now be controlled using the
rcu_cpu_stall_suppress boot-time parameter or via the same parameter
from sysfs. There is therefore no longer any reason to have
kernel config parameters for this feature. This commit therefore
removes the RCU_CPU_STALL_DETECTOR and RCU_CPU_STALL_DETECTOR_RUNNABLE
kernel config parameters. The RCU_CPU_STALL_TIMEOUT parameter remains
to allow the timeout to be tuned and the RCU_CPU_STALL_VERBOSE parameter
remains to allow task-stall information to be suppressed if desired.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Lai's RCU-callback immediate-adoption patch changes the RCU tracing
output, so update tracing.txt. Also update a few comments to clarify
the synchronization design.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Add the required verbiage to Documentation/RCU/trace.txt.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The current tracing data is not sufficient to deduce the average time
that a callback spends waiting for a grace period to end. Add three
per-CPU counters recording the number of callbacks invoked (ci), the
number of callbacks orphaned (co), and the number of callbacks adopted
(ca). Given the existing callback queue length (ql), the average wait
time in absence of CPU hotplug operations is ql/ci. The units of wait
time will be in terms of the duration over which ci was measured.
In the presence of CPU hotplug operations, there is room for argument,
but ql/(ci-co+ca) won't steer you too far wrong.
Also fixes a typo called out by Lucas De Marchi <lucas.de.marchi@gmail.com>.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
CPU-bound real-time processes can cause RCU CPU stall warnings, and
much other trouble as well. Document the fact that they can cause
RCU CPU stall warnings.
Suggested-by: Darren Hart <dvhltc@us.ibm.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Add a section describing PROVE_RCU, DEBUG_OBJECTS_RCU_HEAD, and
the __rcu sparse checking to the RCU checklist.
Suggested-by: David Miller <davem@davemloft.net>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Below you will find an updated version from the original series bunching all patches into one big patch
updating broken web addresses that are located in Documentation/*
Some of the addresses date as far far back as 1995 etc... so searching became a bit difficult,
the best way to deal with these is to use web.archive.org to locate these addresses that are outdated.
Now there are also some addresses pointing to .spec files some are located, but some(after searching
on the companies site)where still no where to be found. In this case I just changed the address
to the company site this way the users can contact the company and they can locate them for the users.
Signed-off-by: Justin P. Mattock <justinmattock@gmail.com>
Signed-off-by: Thomas Weber <weber@corscience.de>
Signed-off-by: Mike Frysinger <vapier.adi@gmail.com>
Cc: Paulo Marques <pmarques@grupopie.com>
Cc: Randy Dunlap <rdunlap@xenotime.net>
Cc: Michael Neuling <mikey@neuling.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (49 commits)
stop_machine: Move local variable closer to the usage site in cpu_stop_cpu_callback()
sched, wait: Use wrapper functions
sched: Remove a stale comment
ondemand: Make the iowait-is-busy time a sysfs tunable
ondemand: Solve a big performance issue by counting IOWAIT time as busy
sched: Intoduce get_cpu_iowait_time_us()
sched: Eliminate the ts->idle_lastupdate field
sched: Fold updating of the last_update_time_info into update_ts_time_stats()
sched: Update the idle statistics in get_cpu_idle_time_us()
sched: Introduce a function to update the idle statistics
sched: Add a comment to get_cpu_idle_time_us()
cpu_stop: add dummy implementation for UP
sched: Remove rq argument to the tracepoints
rcu: need barrier() in UP synchronize_sched_expedited()
sched: correctly place paranioa memory barriers in synchronize_sched_expedited()
sched: kill paranoia check in synchronize_sched_expedited()
sched: replace migration_thread with cpu_stop
stop_machine: reimplement using cpu_stop
cpu_stop: implement stop_cpu[s]()
sched: Fix select_idle_sibling() logic in select_task_rq_fair()
...
The existing Documentation/RCU/stallwarn.txt has proven unhelpful, so
rework it a bit. In particular, show how to interpret the stall-warning
messages.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Lai Jiangshan noted that up to 10% of the RCU_SOFTIRQ are spurious, and
traced this down to the fact that the current grace-period machinery
will uselessly raise RCU_SOFTIRQ when a given CPU needs to go through
a quiescent state, but has not yet done so. In this situation, there
might well be nothing that RCU_SOFTIRQ can do, and the overhead can be
worth worrying about in the ksoftirqd case. This patch therefore avoids
raising RCU_SOFTIRQ in this situation.
Changes since v1 (http://lkml.org/lkml/2010/3/30/122 from Lai Jiangshan):
o Omit the rcu_qs_pending() prechecks, as they aren't that
much less expensive than the quiescent-state checks.
o Merge with the set_need_resched() patch that reduces IPIs.
o Add the new n_rp_report_qs field to the rcu_pending tracing output.
o Update the tracing documentation accordingly.
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Currently migration_thread is serving three purposes - migration
pusher, context to execute active_load_balance() and forced context
switcher for expedited RCU synchronize_sched. All three roles are
hardcoded into migration_thread() and determining which job is
scheduled is slightly messy.
This patch kills migration_thread and replaces all three uses with
cpu_stop. The three different roles of migration_thread() are
splitted into three separate cpu_stop callbacks -
migration_cpu_stop(), active_load_balance_cpu_stop() and
synchronize_sched_expedited_cpu_stop() - and each use case now simply
asks cpu_stop to execute the callback as necessary.
synchronize_sched_expedited() was implemented with private
preallocated resources and custom multi-cpu queueing and waiting
logic, both of which are provided by cpu_stop.
synchronize_sched_expedited_count is made atomic and all other shared
resources along with the mutex are dropped.
synchronize_sched_expedited() also implemented a check to detect cases
where not all the callback got executed on their assigned cpus and
fall back to synchronize_sched(). If called with cpu hotplug blocked,
cpu_stop already guarantees that and the condition cannot happen;
otherwise, stop_machine() would break. However, this patch preserves
the paranoid check using a cpumask to record on which cpus the stopper
ran so that it can serve as a bisection point if something actually
goes wrong theree.
Because the internal execution state is no longer visible,
rcu_expedited_torture_stats() is removed.
This patch also renames cpu_stop threads to from "stopper/%d" to
"migration/%d". The names of these threads ultimately don't matter
and there's no reason to make unnecessary userland visible changes.
With this patch applied, stop_machine() and sched now share the same
resources. stop_machine() is faster without wasting any resources and
sched migration users are much cleaner.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Dipankar Sarma <dipankar@in.ibm.com>
Cc: Josh Triplett <josh@freedesktop.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Dimitri Sivanich <sivanich@sgi.com>
When a slab cache uses SLAB_DESTROY_BY_RCU, we must be careful when allocating
objects, since slab allocator could give a freed object still used by lockless
readers.
In particular, nf_conntrack RCU lookups rely on ct->tuplehash[xxx].hnnode.next
being always valid (ie containing a valid 'nulls' value, or a valid pointer to next
object in hash chain.)
kmem_cache_zalloc() setups object with NULL values, but a NULL value is not valid
for ct->tuplehash[xxx].hnnode.next.
Fix is to call kmem_cache_alloc() and do the zeroing ourself.
As spotted by Patrick, we also need to make sure lookup keys are committed to
memory before setting refcount to 1, or a lockless reader could get a reference
on the old version of the object. Its key re-check could then pass the barrier.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Doc: Fix spelling in RCU/rculist_nulls.txt.
Trival spelling fixes in RCU/rculist_nulls.txt.
Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
Tested-by: Jarek Poplawski <jarkao2@gmail.com;->
Signed-off-by: David S. Miller <davem@davemloft.net>
At some point the API of call_rcu() changed from three parameters
to two parameters, correct the documentation.
One confusing thing in RCU/listRCU.txt, which is NOT fixed in this patch,
is that no reason or explaination is given for using call_rcu() instead of
the normal synchronize_rcu() call.
Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Trivial fix while reading through the RCU docs.
Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Update the RCU documentation to call out the need for callers of
primitives like call_rcu() and synchronize_rcu() to prevent subsequent RCU
readers from hazard.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'docs-next' of git://git.lwn.net/linux-2.6:
Fix a typo in the development process document.
Document handling of bad memory
Document RCU and unloadable modules
* 'core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (63 commits)
stacktrace: provide save_stack_trace_tsk() weak alias
rcu: provide RCU options on non-preempt architectures too
printk: fix discarding message when recursion_bug
futex: clean up futex_(un)lock_pi fault handling
"Tree RCU": scalable classic RCU implementation
futex: rename field in futex_q to clarify single waiter semantics
x86/swiotlb: add default swiotlb_arch_range_needs_mapping
x86/swiotlb: add default phys<->bus conversion
x86: unify pci iommu setup and allow swiotlb to compile for 32 bit
x86: add swiotlb allocation functions
swiotlb: consolidate swiotlb info message printing
swiotlb: support bouncing of HighMem pages
swiotlb: factor out copy to/from device
swiotlb: add arch hook to force mapping
swiotlb: allow architectures to override phys<->bus<->phys conversions
swiotlb: add comment where we handle the overflow of a dma mask on 32 bit
rcu: fix rcutorture behavior during reboot
resources: skip sanity check of busy resources
swiotlb: move some definitions to header
swiotlb: allow architectures to override swiotlb pool allocation
...
Fix up trivial conflicts in
arch/x86/kernel/Makefile
arch/x86/mm/init_32.c
include/linux/hardirq.h
as per Ingo's suggestions.
This patch fixes a long-standing performance bug in classic RCU that
results in massive internal-to-RCU lock contention on systems with
more than a few hundred CPUs. Although this patch creates a separate
flavor of RCU for ease of review and patch maintenance, it is intended
to replace classic RCU.
This patch still handles stress better than does mainline, so I am still
calling it ready for inclusion. This patch is against the -tip tree.
Nevertheless, experience on an actual 1000+ CPU machine would still be
most welcome.
Most of the changes noted below were found while creating an rcutiny
(which should permit ejecting the current rcuclassic) and while doing
detailed line-by-line documentation.
Updates from v9 (http://lkml.org/lkml/2008/12/2/334):
o Fixes from remainder of line-by-line code walkthrough,
including comment spelling, initialization, undesirable
narrowing due to type conversion, removing redundant memory
barriers, removing redundant local-variable initialization,
and removing redundant local variables.
I do not believe that any of these fixes address the CPU-hotplug
issues that Andi Kleen was seeing, but please do give it a whirl
in case the machine is smarter than I am.
A writeup from the walkthrough may be found at the following
URL, in case you are suffering from terminal insomnia or
masochism:
http://www.kernel.org/pub/linux/kernel/people/paulmck/tmp/rcutree-walkthrough.2008.12.16a.pdf
o Made rcutree tracing use seq_file, as suggested some time
ago by Lai Jiangshan.
o Added a .csv variant of the rcudata debugfs trace file, to allow
people having thousands of CPUs to drop the data into
a spreadsheet. Tested with oocalc and gnumeric. Updated
documentation to suit.
Updates from v8 (http://lkml.org/lkml/2008/11/15/139):
o Fix a theoretical race between grace-period initialization and
force_quiescent_state() that could occur if more than three
jiffies were required to carry out the grace-period
initialization. Which it might, if you had enough CPUs.
o Apply Ingo's printk-standardization patch.
o Substitute local variables for repeated accesses to global
variables.
o Fix comment misspellings and redundant (but harmless) increments
of ->n_rcu_pending (this latter after having explicitly added it).
o Apply checkpatch fixes.
Updates from v7 (http://lkml.org/lkml/2008/10/10/291):
o Fixed a number of problems noted by Gautham Shenoy, including
the cpu-stall-detection bug that he was having difficulty
convincing me was real. ;-)
o Changed cpu-stall detection to wait for ten seconds rather than
three in order to reduce false positive, as suggested by Ingo
Molnar.
o Produced a design document (http://lwn.net/Articles/305782/).
The act of writing this document uncovered a number of both
theoretical and "here and now" bugs as noted below.
o Fix dynticks_nesting accounting confusion, simplify WARN_ON()
condition, fix kerneldoc comments, and add memory barriers
in dynticks interface functions.
o Add more data to tracing.
o Remove unused "rcu_barrier" field from rcu_data structure.
o Count calls to rcu_pending() from scheduling-clock interrupt
to use as a surrogate timebase should jiffies stop counting.
o Fix a theoretical race between force_quiescent_state() and
grace-period initialization. Yes, initialization does have to
go on for some jiffies for this race to occur, but given enough
CPUs...
Updates from v6 (http://lkml.org/lkml/2008/9/23/448):
o Fix a number of checkpatch.pl complaints.
o Apply review comments from Ingo Molnar and Lai Jiangshan
on the stall-detection code.
o Fix several bugs in !CONFIG_SMP builds.
o Fix a misspelled config-parameter name so that RCU now announces
at boot time if stall detection is configured.
o Run tests on numerous combinations of configurations parameters,
which after the fixes above, now build and run correctly.
Updates from v5 (http://lkml.org/lkml/2008/9/15/92, bad subject line):
o Fix a compiler error in the !CONFIG_FANOUT_EXACT case (blew a
changeset some time ago, and finally got around to retesting
this option).
o Fix some tracing bugs in rcupreempt that caused incorrect
totals to be printed.
o I now test with a more brutal random-selection online/offline
script (attached). Probably more brutal than it needs to be
on the people reading it as well, but so it goes.
o A number of optimizations and usability improvements:
o Make rcu_pending() ignore the grace-period timeout when
there is no grace period in progress.
o Make force_quiescent_state() avoid going for a global
lock in the case where there is no grace period in
progress.
o Rearrange struct fields to improve struct layout.
o Make call_rcu() initiate a grace period if RCU was
idle, rather than waiting for the next scheduling
clock interrupt.
o Invoke rcu_irq_enter() and rcu_irq_exit() only when
idle, as suggested by Andi Kleen. I still don't
completely trust this change, and might back it out.
o Make CONFIG_RCU_TRACE be the single config variable
manipulated for all forms of RCU, instead of the prior
confusion.
o Document tracing files and formats for both rcupreempt
and rcutree.
Updates from v4 for those missing v5 given its bad subject line:
o Separated dynticks interface so that NMIs and irqs call separate
functions, greatly simplifying it. In particular, this code
no longer requires a proof of correctness. ;-)
o Separated dynticks state out into its own per-CPU structure,
avoiding the duplicated accounting.
o The case where a dynticks-idle CPU runs an irq handler that
invokes call_rcu() is now correctly handled, forcing that CPU
out of dynticks-idle mode.
o Review comments have been applied (thank you all!!!).
For but one example, fixed the dynticks-ordering issue that
Manfred pointed out, saving me much debugging. ;-)
o Adjusted rcuclassic and rcupreempt to handle dynticks changes.
Attached is an updated patch to Classic RCU that applies a hierarchy,
greatly reducing the contention on the top-level lock for large machines.
This passes 10-hour concurrent rcutorture and online-offline testing on
128-CPU ppc64 without dynticks enabled, and exposes some timekeeping
bugs in presence of dynticks (exciting working on a system where
"sleep 1" hangs until interrupted...), which were fixed in the
2.6.27 kernel. It is getting more reliable than mainline by some
measures, so the next version will be against -tip for inclusion.
See also Manfred Spraul's recent patches (or his earlier work from
2004 at http://marc.info/?l=linux-kernel&m=108546384711797&w=2).
We will converge onto a common patch in the fullness of time, but are
currently exploring different regions of the design space. That said,
I have already gratefully stolen quite a few of Manfred's ideas.
This patch provides CONFIG_RCU_FANOUT, which controls the bushiness
of the RCU hierarchy. Defaults to 32 on 32-bit machines and 64 on
64-bit machines. If CONFIG_NR_CPUS is less than CONFIG_RCU_FANOUT,
there is no hierarchy. By default, the RCU initialization code will
adjust CONFIG_RCU_FANOUT to balance the hierarchy, so strongly NUMA
architectures may choose to set CONFIG_RCU_FANOUT_EXACT to disable
this balancing, allowing the hierarchy to be exactly aligned to the
underlying hardware. Up to two levels of hierarchy are permitted
(in addition to the root node), allowing up to 16,384 CPUs on 32-bit
systems and up to 262,144 CPUs on 64-bit systems. I just know that I
am going to regret saying this, but this seems more than sufficient
for the foreseeable future. (Some architectures might wish to set
CONFIG_RCU_FANOUT=4, which would limit such architectures to 64 CPUs.
If this becomes a real problem, additional levels can be added, but I
doubt that it will make a significant difference on real hardware.)
In the common case, a given CPU will manipulate its private rcu_data
structure and the rcu_node structure that it shares with its immediate
neighbors. This can reduce both lock and memory contention by multiple
orders of magnitude, which should eliminate the need for the strange
manipulations that are reported to be required when running Linux on
very large systems.
Some shortcomings:
o More bugs will probably surface as a result of an ongoing
line-by-line code inspection.
Patches will be provided as required.
o There are probably hangs, rcutorture failures, &c. Seems
quite stable on a 128-CPU machine, but that is kind of small
compared to 4096 CPUs. However, seems to do better than
mainline.
Patches will be provided as required.
o The memory footprint of this version is several KB larger
than rcuclassic.
A separate UP-only rcutiny patch will be provided, which will
reduce the memory footprint significantly, even compared
to the old rcuclassic. One such patch passes light testing,
and has a memory footprint smaller even than rcuclassic.
Initial reaction from various embedded guys was "it is not
worth it", so am putting it aside.
Credits:
o Manfred Spraul for ideas, review comments, and bugs spotted,
as well as some good friendly competition. ;-)
o Josh Triplett, Ingo Molnar, Peter Zijlstra, Mathieu Desnoyers,
Lai Jiangshan, Andi Kleen, Andy Whitcroft, and Andrew Morton
for reviews and comments.
o Thomas Gleixner for much-needed help with some timer issues
(see patches below).
o Jon M. Tollefson, Tim Pepper, Andrew Theurer, Jose R. Santos,
Andy Whitcroft, Darrick Wong, Nishanth Aravamudan, Anton
Blanchard, Dave Kleikamp, and Nathan Lynch for keeping machines
alive despite my heavy abuse^Wtesting.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Adds Documentation/RCU/rculist_nulls.txt file to describe how 'nulls'
end-of-list can help in some RCU algos.
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: David S. Miller <davem@davemloft.net>
atomic_inc_not_zero(v) return 0 if *v = 0.
use spin_lock instead of write_lock for update lock.
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
All of the in-tree uses of list_for_each_rcu() have been converted to
list_for_each_entry_rcu(), so list_for_each_rcu() can now be removed.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch re-institutes the ability to build rcutorture directly into
the Linux kernel. The reason that this capability was removed was that
this could result in your kernel being pretty much useless, as rcutorture
would be running starting from early boot. This problem has been avoided
by (1) making rcutorture run only three seconds of every six by default,
(2) adding a CONFIG_RCU_TORTURE_TEST_RUNNABLE that permits rcutorture
to be quiesced at boot time, and (3) adding a sysctl in /proc named
/proc/sys/kernel/rcutorture_runnable that permits rcutorture to be
quiesced and unquiesced when built into the kernel.
Please note that this /proc file is -not- available when rcutorture
is built as a module. Please also note that to get the earlier
take-no-prisoners behavior, you must use the boot command line to set
rcutorture's "stutter" parameter to zero.
The rcutorture quiescing mechanism is currently quite crude: loops
in each rcutorture process that poll a global variable once per tick.
Suggestions for improvement are welcome. The default action will
be to reduce the polling rate to a few times per second.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Suggested-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch takes a step towards making rcutorture more brutal by allowing
the test to be automatically periodically paused, with the default being
to run the test for five seconds then pause for five seconds and repeat.
This behavior can be controlled using a new "stutter" module parameter, so
that "stutter=0" gives the old default behavior of running continuously.
Starting and stopping rcutorture more heavily stresses RCU's interaction
with the scheduler, as well as exercising more paths through the
grace-period detection code.
Note that the default to "shuffle_interval" has also been adjusted from
5 seconds to 3 seconds to provide varying overlap with the "stutter"
interval.
I am still unable to provoke the failures that Alexey has been seeing,
even with this patch, but will be doing a few additional things to beef
up rcutorture.
Suggested-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Long-delayed update to the RCU documentation, including adding the new
call_rcu_sched() and rcu_barrier_sched() APIs.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
fastcall always expands to empty, remove it.
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch updates the RCU documentation to reflect preemptible RCU as
well as recent publications.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Gautham R Shenoy <ego@in.ibm.com>
Reviewed-by: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Add Documentation/RCU/00-INDEX
Signed-off-by: Rob Landley <rob@landley.net>
Acked-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add an item to the RCU documentation checklist noting that RCU callbacks
can run in parallel.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Implement torture testing for the "sched" variant of RCU, which uses
preempt_disable, preempt_enable, and synchronize_sched.
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Acked-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Use the newly-generic synchronous deferred free function to implement torture
testing for rcu_bh using synchronize_rcu_bh rather than the asynchronous
call_rcu_bh.
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Acked-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Use the newly-generic synchronous deferred free function to implement torture
testing for RCU using synchronize_rcu rather than the asynchronous call_rcu.
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Acked-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
rcutorture currently has one writer and an arbitrary number of readers. To
better exercise some of the code paths in RCU implementations, add fake
writer threads which call the synchronize function for the RCU variant in a
loop, with a delay between calls to arrange for different numbers of
writers running in parallel.
[bunk@stusta.de: cleanup]
Acked-by: Paul McKenney <paulmck@us.ibm.com>
Cc: Dipkanar Sarma <dipankar@in.ibm.com>
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Adds SRCU operations to rcutorture and updates rcutorture documentation.
Also increases the stress imposed by the rcutorture test.
[bunk@stusta.de: make needlessly global code static]
Signed-off-by: Paul E. McKenney <paulmck@us.ibm.com>
Cc: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Updated patch adding a variant of RCU that permits sleeping in read-side
critical sections. SRCU is as follows:
o Each use of SRCU creates its own srcu_struct, and each
srcu_struct has its own set of grace periods. This is
critical, as it prevents one subsystem with a blocking
reader from holding up SRCU grace periods for other
subsystems.
o The SRCU primitives (srcu_read_lock(), srcu_read_unlock(),
and synchronize_srcu()) all take a pointer to a srcu_struct.
o The SRCU primitives must be called from process context.
o srcu_read_lock() returns an int that must be passed to
the matching srcu_read_unlock(). Realtime RCU avoids the
need for this by storing the state in the task struct,
but SRCU needs to allow a given code path to pass through
multiple SRCU domains -- storing state in the task struct
would therefore require either arbitrary space in the
task struct or arbitrary limits on SRCU nesting. So I
kicked the state-storage problem up to the caller.
Of course, it is not permitted to call synchronize_srcu()
while in an SRCU read-side critical section.
o There is no call_srcu(). It would not be hard to implement
one, but it seems like too easy a way to OOM the system.
(Hey, we have enough trouble with call_rcu(), which does
-not- permit readers to sleep!!!) So, if you want it,
please tell me why...
[josht@us.ibm.com: sparse notation]
Signed-off-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Josh Triplett <josh@freedesktop.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Remove many duplicated words under Documentation/ and do other small
cleanups.
Examples:
"and and" --> "and"
"in in" --> "in"
"the the" --> "the"
"the the" --> "to the"
...
Signed-off-by: Paolo Ornati <ornati@fastwebnet.it>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
This patch fixes typos in various Documentation txts. The patch addresses
some words starting with the letter 'S'.
Signed-off-by: Matt LaPlante <kernel1@cyberdogtech.com>
Acked-by: Alan Cox <alan@redhat.com>
Acked-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Updater should use _rcu variant of list_del().
Signed-off-by: Urs Thuermann <urs@isnogud.escape.de>
Acked-by: "Paul E. McKenney" <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Add an ops vector to rcutorture, and add the ops for Classic RCU. Update
the rcutorture documentation to reflect slight change to the dmesg formats.
Signed-off-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This just catches the RCU torture documentation up with the recent fixes
that test RCU for architectures that turn of the scheduling-clock interrupt
for idle CPUs and the addition of a SUCCESS/FAILURE indication, fixing up
an obsolete comment as well.
Signed-off-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
An update to the RCU documentation calling out the
self-limiting-update-rate advantages of synchronize_rcu(), and describing
how to use call_rcu() in a way that results in self-limiting updates.
Self-limiting updates are important to avoiding RCU-induced OOM in face of
denial-of-service attacks.
Signed-off-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Remove synchronize_kernel() (deprecated 2-APR-2005 in
http://lkml.org/lkml/2005/4/3/11) and makes the RCU API inaccessible to
non-GPL Linux kernel modules (as was announced more than one year ago in
http://lkml.org/lkml/2005/4/3/8). Tested on x86 and ppc64.
Signed-off-by: "Paul E. McKenney" <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Replace for_each_cpu with for_each_possible_cpu.
Modifies occurences in documentaion.
for_each_cpu in whatisRCU.txt should be for_each_online_cpu ???
(I'm not sure..)
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The RCU documentation uses an fp variable which is not declared in the code
snippets. Use the new_fp variable instead.
Signed-Off-By: Baruch Even <baruch@ev-en.org>
Acked-by: <paulmck@us.ibm.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Updates to in-tree RCU documentation based on comments over the past few
months.
Signed-off-by: "Paul E. McKenney" <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Use atomic_inc_not_zero for rcu files instead of special case rcuref.
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: "Paul E. McKenney" <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Remove the hlist_for_each_rcu() API, which is used only in one place, and
is trivially converted to hlist_for_each_entry_rcu(), making the code
shorter and more readable. Any out-of-tree uses may be similarly
converted.
Signed-off-by: "Paul E. McKenney" <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch is a rewrite of the one submitted on October 1st, using modules
(http://marc.theaimsgroup.com/?l=linux-kernel&m=112819093522998&w=2).
This rewrite adds a tristate CONFIG_RCU_TORTURE_TEST, which enables an
intense torture test of the RCU infratructure. This is needed due to the
continued changes to the RCU infrastructure to accommodate dynamic ticks,
CPU hotplug, realtime, and so on. Most of the code is in a separate file
that is compiled only if the CONFIG variable is set. Documentation on how
to run the test and interpret the output is also included.
This code has been tested on i386 and ppc64, and an earlier version of the
code has received extensive testing on a number of architectures as part of
the PREEMPT_RT patchset.
Signed-off-by: "Paul E. McKenney" <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Update RCU documentation based on discussions and review of RCU-based tree
patches. Add an introductory whatisRCU.txt file.
Signed-off-by: <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Adds a set of primitives to do reference counting for objects that are looked
up without locks using RCU.
Signed-off-by: Ravikiran Thirumalai <kiran_th@gmail.com>
Signed-off-by: Dipankar Sarma <dipankar@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Uses of RCU for dynamically changeable NMI handlers need to use the new
rcu_dereference() and rcu_assign_pointer() facilities. This change makes
it clear that these uses are safe from a memory-barrier viewpoint, but the
main purpose is to document exactly what operations are being protected by
RCU. This has been tested on x86 and x86-64, which are the only
architectures affected by this change.
Signed-off-by: <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Update the RCU documentation to allow for the new synchronize_rcu() and
synchronize_sched() primitives. Fix a few other nits as well.
Signed-off-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
