aboutsummaryrefslogtreecommitdiffstats
path: root/kernel/rcutree_trace.c
diff options
context:
space:
mode:
authorPaul E. McKenney <paulmck@linux.vnet.ibm.com>2008-12-18 15:55:32 -0500
committerIngo Molnar <mingo@elte.hu>2008-12-18 15:56:04 -0500
commit64db4cfff99c04cd5f550357edcc8780f96b54a2 (patch)
tree4856e788d21f0e31ed78a22b70b4521f7237705e /kernel/rcutree_trace.c
parentd110ec3a1e1f522e2e9dfceb9c36d6590c26d2d4 (diff)
"Tree RCU": scalable classic RCU implementation
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>
Diffstat (limited to 'kernel/rcutree_trace.c')
-rw-r--r--kernel/rcutree_trace.c271
1 files changed, 271 insertions, 0 deletions
diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c
new file mode 100644
index 000000000000..d6db3e837826
--- /dev/null
+++ b/kernel/rcutree_trace.c
@@ -0,0 +1,271 @@
1/*
2 * Read-Copy Update tracing for classic implementation
3 *
4 * This program is free software; you can redistribute it and/or modify
5 * it under the terms of the GNU General Public License as published by
6 * the Free Software Foundation; either version 2 of the License, or
7 * (at your option) any later version.
8 *
9 * This program is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 * GNU General Public License for more details.
13 *
14 * You should have received a copy of the GNU General Public License
15 * along with this program; if not, write to the Free Software
16 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
17 *
18 * Copyright IBM Corporation, 2008
19 *
20 * Papers: http://www.rdrop.com/users/paulmck/RCU
21 *
22 * For detailed explanation of Read-Copy Update mechanism see -
23 * Documentation/RCU
24 *
25 */
26#include <linux/types.h>
27#include <linux/kernel.h>
28#include <linux/init.h>
29#include <linux/spinlock.h>
30#include <linux/smp.h>
31#include <linux/rcupdate.h>
32#include <linux/interrupt.h>
33#include <linux/sched.h>
34#include <asm/atomic.h>
35#include <linux/bitops.h>
36#include <linux/module.h>
37#include <linux/completion.h>
38#include <linux/moduleparam.h>
39#include <linux/percpu.h>
40#include <linux/notifier.h>
41#include <linux/cpu.h>
42#include <linux/mutex.h>
43#include <linux/debugfs.h>
44#include <linux/seq_file.h>
45
46static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp)
47{
48 if (!rdp->beenonline)
49 return;
50 seq_printf(m, "%3d%cc=%ld g=%ld pq=%d pqc=%ld qp=%d rpfq=%ld rp=%x",
51 rdp->cpu,
52 cpu_is_offline(rdp->cpu) ? '!' : ' ',
53 rdp->completed, rdp->gpnum,
54 rdp->passed_quiesc, rdp->passed_quiesc_completed,
55 rdp->qs_pending,
56 rdp->n_rcu_pending_force_qs - rdp->n_rcu_pending,
57 (int)(rdp->n_rcu_pending & 0xffff));
58#ifdef CONFIG_NO_HZ
59 seq_printf(m, " dt=%d/%d dn=%d df=%lu",
60 rdp->dynticks->dynticks,
61 rdp->dynticks->dynticks_nesting,
62 rdp->dynticks->dynticks_nmi,
63 rdp->dynticks_fqs);
64#endif /* #ifdef CONFIG_NO_HZ */
65 seq_printf(m, " of=%lu ri=%lu", rdp->offline_fqs, rdp->resched_ipi);
66 seq_printf(m, " ql=%ld b=%ld\n", rdp->qlen, rdp->blimit);
67}
68
69#define PRINT_RCU_DATA(name, func, m) \
70 do { \
71 int _p_r_d_i; \
72 \
73 for_each_possible_cpu(_p_r_d_i) \
74 func(m, &per_cpu(name, _p_r_d_i)); \
75 } while (0)
76
77static int show_rcudata(struct seq_file *m, void *unused)
78{
79 seq_puts(m, "rcu:\n");
80 PRINT_RCU_DATA(rcu_data, print_one_rcu_data, m);
81 seq_puts(m, "rcu_bh:\n");
82 PRINT_RCU_DATA(rcu_bh_data, print_one_rcu_data, m);
83 return 0;
84}
85
86static int rcudata_open(struct inode *inode, struct file *file)
87{
88 return single_open(file, show_rcudata, NULL);
89}
90
91static struct file_operations rcudata_fops = {
92 .owner = THIS_MODULE,
93 .open = rcudata_open,
94 .read = seq_read,
95 .llseek = seq_lseek,
96 .release = single_release,
97};
98
99static void print_one_rcu_data_csv(struct seq_file *m, struct rcu_data *rdp)
100{
101 if (!rdp->beenonline)
102 return;
103 seq_printf(m, "%d,%s,%ld,%ld,%d,%ld,%d,%ld,%ld",
104 rdp->cpu,
105 cpu_is_offline(rdp->cpu) ? "\"Y\"" : "\"N\"",
106 rdp->completed, rdp->gpnum,
107 rdp->passed_quiesc, rdp->passed_quiesc_completed,
108 rdp->qs_pending,
109 rdp->n_rcu_pending_force_qs - rdp->n_rcu_pending,
110 rdp->n_rcu_pending);
111#ifdef CONFIG_NO_HZ
112 seq_printf(m, ",%d,%d,%d,%lu",
113 rdp->dynticks->dynticks,
114 rdp->dynticks->dynticks_nesting,
115 rdp->dynticks->dynticks_nmi,
116 rdp->dynticks_fqs);
117#endif /* #ifdef CONFIG_NO_HZ */
118 seq_printf(m, ",%lu,%lu", rdp->offline_fqs, rdp->resched_ipi);
119 seq_printf(m, ",%ld,%ld\n", rdp->qlen, rdp->blimit);
120}
121
122static int show_rcudata_csv(struct seq_file *m, void *unused)
123{
124 seq_puts(m, "\"CPU\",\"Online?\",\"c\",\"g\",\"pq\",\"pqc\",\"pq\",\"rpfq\",\"rp\",");
125#ifdef CONFIG_NO_HZ
126 seq_puts(m, "\"dt\",\"dt nesting\",\"dn\",\"df\",");
127#endif /* #ifdef CONFIG_NO_HZ */
128 seq_puts(m, "\"of\",\"ri\",\"ql\",\"b\"\n");
129 seq_puts(m, "\"rcu:\"\n");
130 PRINT_RCU_DATA(rcu_data, print_one_rcu_data_csv, m);
131 seq_puts(m, "\"rcu_bh:\"\n");
132 PRINT_RCU_DATA(rcu_bh_data, print_one_rcu_data_csv, m);
133 return 0;
134}
135
136static int rcudata_csv_open(struct inode *inode, struct file *file)
137{
138 return single_open(file, show_rcudata_csv, NULL);
139}
140
141static struct file_operations rcudata_csv_fops = {
142 .owner = THIS_MODULE,
143 .open = rcudata_csv_open,
144 .read = seq_read,
145 .llseek = seq_lseek,
146 .release = single_release,
147};
148
149static void print_one_rcu_state(struct seq_file *m, struct rcu_state *rsp)
150{
151 int level = 0;
152 struct rcu_node *rnp;
153
154 seq_printf(m, "c=%ld g=%ld s=%d jfq=%ld j=%x "
155 "nfqs=%lu/nfqsng=%lu(%lu) fqlh=%lu\n",
156 rsp->completed, rsp->gpnum, rsp->signaled,
157 (long)(rsp->jiffies_force_qs - jiffies),
158 (int)(jiffies & 0xffff),
159 rsp->n_force_qs, rsp->n_force_qs_ngp,
160 rsp->n_force_qs - rsp->n_force_qs_ngp,
161 rsp->n_force_qs_lh);
162 for (rnp = &rsp->node[0]; rnp - &rsp->node[0] < NUM_RCU_NODES; rnp++) {
163 if (rnp->level != level) {
164 seq_puts(m, "\n");
165 level = rnp->level;
166 }
167 seq_printf(m, "%lx/%lx %d:%d ^%d ",
168 rnp->qsmask, rnp->qsmaskinit,
169 rnp->grplo, rnp->grphi, rnp->grpnum);
170 }
171 seq_puts(m, "\n");
172}
173
174static int show_rcuhier(struct seq_file *m, void *unused)
175{
176 seq_puts(m, "rcu:\n");
177 print_one_rcu_state(m, &rcu_state);
178 seq_puts(m, "rcu_bh:\n");
179 print_one_rcu_state(m, &rcu_bh_state);
180 return 0;
181}
182
183static int rcuhier_open(struct inode *inode, struct file *file)
184{
185 return single_open(file, show_rcuhier, NULL);
186}
187
188static struct file_operations rcuhier_fops = {
189 .owner = THIS_MODULE,
190 .open = rcuhier_open,
191 .read = seq_read,
192 .llseek = seq_lseek,
193 .release = single_release,
194};
195
196static int show_rcugp(struct seq_file *m, void *unused)
197{
198 seq_printf(m, "rcu: completed=%ld gpnum=%ld\n",
199 rcu_state.completed, rcu_state.gpnum);
200 seq_printf(m, "rcu_bh: completed=%ld gpnum=%ld\n",
201 rcu_bh_state.completed, rcu_bh_state.gpnum);
202 return 0;
203}
204
205static int rcugp_open(struct inode *inode, struct file *file)
206{
207 return single_open(file, show_rcugp, NULL);
208}
209
210static struct file_operations rcugp_fops = {
211 .owner = THIS_MODULE,
212 .open = rcugp_open,
213 .read = seq_read,
214 .llseek = seq_lseek,
215 .release = single_release,
216};
217
218static struct dentry *rcudir, *datadir, *datadir_csv, *hierdir, *gpdir;
219static int __init rcuclassic_trace_init(void)
220{
221 rcudir = debugfs_create_dir("rcu", NULL);
222 if (!rcudir)
223 goto out;
224
225 datadir = debugfs_create_file("rcudata", 0444, rcudir,
226 NULL, &rcudata_fops);
227 if (!datadir)
228 goto free_out;
229
230 datadir_csv = debugfs_create_file("rcudata.csv", 0444, rcudir,
231 NULL, &rcudata_csv_fops);
232 if (!datadir_csv)
233 goto free_out;
234
235 gpdir = debugfs_create_file("rcugp", 0444, rcudir, NULL, &rcugp_fops);
236 if (!gpdir)
237 goto free_out;
238
239 hierdir = debugfs_create_file("rcuhier", 0444, rcudir,
240 NULL, &rcuhier_fops);
241 if (!hierdir)
242 goto free_out;
243 return 0;
244free_out:
245 if (datadir)
246 debugfs_remove(datadir);
247 if (datadir_csv)
248 debugfs_remove(datadir_csv);
249 if (gpdir)
250 debugfs_remove(gpdir);
251 debugfs_remove(rcudir);
252out:
253 return 1;
254}
255
256static void __exit rcuclassic_trace_cleanup(void)
257{
258 debugfs_remove(datadir);
259 debugfs_remove(datadir_csv);
260 debugfs_remove(gpdir);
261 debugfs_remove(hierdir);
262 debugfs_remove(rcudir);
263}
264
265
266module_init(rcuclassic_trace_init);
267module_exit(rcuclassic_trace_cleanup);
268
269MODULE_AUTHOR("Paul E. McKenney");
270MODULE_DESCRIPTION("Read-Copy Update tracing for hierarchical implementation");
271MODULE_LICENSE("GPL");
generated by cgit v1.2.2 (git 2.25.0) at 2024-11-07 01:34:21 -0500