From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, Steven Rostedt <rostedt@goodmis.org>
Subject: [ 042/115] tracing: Remove most or all of stack tracer stack size from stack_max_size
Date: Mon, 6 May 2013 13:44:38 -0700 [thread overview]
Message-ID: <20130506203059.871305620@linuxfoundation.org> (raw)
In-Reply-To: <20130506203055.537199268@linuxfoundation.org>
3.9-stable review patch. If anyone has any objections, please let me know.
------------------
From: "Steven Rostedt (Red Hat)" <rostedt@goodmis.org>
commit 4df297129f622bdc18935c856f42b9ddd18f9f28 upstream.
Currently, the depth reported in the stack tracer stack_trace file
does not match the stack_max_size file. This is because the stack_max_size
includes the overhead of stack tracer itself while the depth does not.
The first time a max is triggered, a calculation is not performed that
figures out the overhead of the stack tracer and subtracts it from
the stack_max_size variable. The overhead is stored and is subtracted
from the reported stack size for comparing for a new max.
Now the stack_max_size corresponds to the reported depth:
# cat stack_max_size
4640
# cat stack_trace
Depth Size Location (48 entries)
----- ---- --------
0) 4640 32 _raw_spin_lock+0x18/0x24
1) 4608 112 ____cache_alloc+0xb7/0x22d
2) 4496 80 kmem_cache_alloc+0x63/0x12f
3) 4416 16 mempool_alloc_slab+0x15/0x17
[...]
While testing against and older gcc on x86 that uses mcount instead
of fentry, I found that pasing in ip + MCOUNT_INSN_SIZE let the
stack trace show one more function deep which was missing before.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
kernel/trace/trace_stack.c | 75 ++++++++++++++++++++++++++++++++-------------
1 file changed, 54 insertions(+), 21 deletions(-)
--- a/kernel/trace/trace_stack.c
+++ b/kernel/trace/trace_stack.c
@@ -20,27 +20,24 @@
#define STACK_TRACE_ENTRIES 500
-/*
- * If fentry is used, then the function being traced will
- * jump to fentry directly before it sets up its stack frame.
- * We need to ignore that one and record the parent. Since
- * the stack frame for the traced function wasn't set up yet,
- * the stack_trace wont see the parent. That needs to be added
- * manually to stack_dump_trace[] as the first element.
- */
#ifdef CC_USING_FENTRY
-# define add_func 1
+# define fentry 1
#else
-# define add_func 0
+# define fentry 0
#endif
static unsigned long stack_dump_trace[STACK_TRACE_ENTRIES+1] =
{ [0 ... (STACK_TRACE_ENTRIES)] = ULONG_MAX };
static unsigned stack_dump_index[STACK_TRACE_ENTRIES];
+/*
+ * Reserve one entry for the passed in ip. This will allow
+ * us to remove most or all of the stack size overhead
+ * added by the stack tracer itself.
+ */
static struct stack_trace max_stack_trace = {
- .max_entries = STACK_TRACE_ENTRIES - add_func,
- .entries = &stack_dump_trace[add_func],
+ .max_entries = STACK_TRACE_ENTRIES - 1,
+ .entries = &stack_dump_trace[1],
};
static unsigned long max_stack_size;
@@ -58,10 +55,14 @@ check_stack(unsigned long ip, unsigned l
{
unsigned long this_size, flags;
unsigned long *p, *top, *start;
+ static int tracer_frame;
+ int frame_size = ACCESS_ONCE(tracer_frame);
int i;
this_size = ((unsigned long)stack) & (THREAD_SIZE-1);
this_size = THREAD_SIZE - this_size;
+ /* Remove the frame of the tracer */
+ this_size -= frame_size;
if (this_size <= max_stack_size)
return;
@@ -73,6 +74,10 @@ check_stack(unsigned long ip, unsigned l
local_irq_save(flags);
arch_spin_lock(&max_stack_lock);
+ /* In case another CPU set the tracer_frame on us */
+ if (unlikely(!frame_size))
+ this_size -= tracer_frame;
+
/* a race could have already updated it */
if (this_size <= max_stack_size)
goto out;
@@ -85,15 +90,12 @@ check_stack(unsigned long ip, unsigned l
save_stack_trace(&max_stack_trace);
/*
- * When fentry is used, the traced function does not get
- * its stack frame set up, and we lose the parent.
- * Add that one in manally. We set up save_stack_trace()
- * to not touch the first element in this case.
+ * Add the passed in ip from the function tracer.
+ * Searching for this on the stack will skip over
+ * most of the overhead from the stack tracer itself.
*/
- if (add_func) {
- stack_dump_trace[0] = ip;
- max_stack_trace.nr_entries++;
- }
+ stack_dump_trace[0] = ip;
+ max_stack_trace.nr_entries++;
/*
* Now find where in the stack these are.
@@ -123,6 +125,18 @@ check_stack(unsigned long ip, unsigned l
found = 1;
/* Start the search from here */
start = p + 1;
+ /*
+ * We do not want to show the overhead
+ * of the stack tracer stack in the
+ * max stack. If we haven't figured
+ * out what that is, then figure it out
+ * now.
+ */
+ if (unlikely(!tracer_frame) && i == 1) {
+ tracer_frame = (p - stack) *
+ sizeof(unsigned long);
+ max_stack_size -= tracer_frame;
+ }
}
}
@@ -149,7 +163,26 @@ stack_trace_call(unsigned long ip, unsig
if (per_cpu(trace_active, cpu)++ != 0)
goto out;
- check_stack(parent_ip, &stack);
+ /*
+ * When fentry is used, the traced function does not get
+ * its stack frame set up, and we lose the parent.
+ * The ip is pretty useless because the function tracer
+ * was called before that function set up its stack frame.
+ * In this case, we use the parent ip.
+ *
+ * By adding the return address of either the parent ip
+ * or the current ip we can disregard most of the stack usage
+ * caused by the stack tracer itself.
+ *
+ * The function tracer always reports the address of where the
+ * mcount call was, but the stack will hold the return address.
+ */
+ if (fentry)
+ ip = parent_ip;
+ else
+ ip += MCOUNT_INSN_SIZE;
+
+ check_stack(ip, &stack);
out:
per_cpu(trace_active, cpu)--;
next prev parent reply other threads:[~2013-05-06 20:44 UTC|newest]
Thread overview: 119+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-05-06 20:43 [ 000/115] 3.9.1-stable review Greg Kroah-Hartman
2013-05-06 20:43 ` [ 001/115] USB: serial: option: Added support Olivetti Olicard 145 Greg Kroah-Hartman
2013-05-06 20:43 ` [ 002/115] USB: option: add a D-Link DWM-156 variant Greg Kroah-Hartman
2013-05-06 20:43 ` [ 003/115] staging: zsmalloc: Fix link error on ARM Greg Kroah-Hartman
2013-05-06 20:44 ` [ 004/115] ARM: omap3: cpuidle: enable time keeping Greg Kroah-Hartman
2013-05-06 20:44 ` [ 005/115] ARM: u300: fix ages old copy/paste bug Greg Kroah-Hartman
2013-05-06 20:44 ` [ 006/115] ARM: at91/at91sam9260.dtsi: fix u(s)art pinctrl encoding Greg Kroah-Hartman
2013-05-06 20:44 ` [ 007/115] ARM: at91: remove partial parameter in bootargs for at91sam9x5ek.dtsi Greg Kroah-Hartman
2013-05-06 20:44 ` [ 008/115] ARM: at91: Fix typo in restart code panic message Greg Kroah-Hartman
2013-05-06 20:44 ` [ 009/115] ARM: at91/trivial: fix model name for SAM9G15-EK Greg Kroah-Hartman
2013-05-06 20:44 ` [ 010/115] ARM: at91/trivial: typos in compatible property Greg Kroah-Hartman
2013-05-06 20:44 ` [ 011/115] powerpc: Add isync to copy_and_flush Greg Kroah-Hartman
2013-05-06 20:44 ` [ 012/115] powerpc: Fix hardware IRQs with MMU on exceptions when HV=0 Greg Kroah-Hartman
2013-05-06 20:44 ` [ 013/115] powerpc/power8: Fix secondary CPUs hanging on boot for HV=0 Greg Kroah-Hartman
2013-05-06 20:44 ` [ 014/115] powerpc/spufs: Initialise inode->i_ino in spufs_new_inode() Greg Kroah-Hartman
2013-05-06 20:44 ` [ 015/115] iwlwifi: fix freeing uninitialized pointer Greg Kroah-Hartman
2013-05-06 20:44 ` [ 016/115] iwlwifi: dvm: dont send zeroed LQ cmd Greg Kroah-Hartman
2013-05-06 20:44 ` [ 017/115] mwifiex: Use pci_release_region() instead of a pci_release_regions() Greg Kroah-Hartman
2013-05-06 20:44 ` [ 018/115] mwifiex: Call pci_release_region after calling pci_disable_device Greg Kroah-Hartman
2013-05-06 20:44 ` [ 019/115] mac80211: fix station entry leak/warning while suspending Greg Kroah-Hartman
2013-05-06 20:44 ` [ 020/115] mac80211: use synchronize_rcu() with rcu_barrier() Greg Kroah-Hartman
2013-05-06 20:44 ` [ 021/115] usb/misc/appledisplay: Add 24" LED Cinema display Greg Kroah-Hartman
2013-05-06 20:44 ` [ 022/115] USB: add ftdi_sio USB ID for GDM Boost V1.x Greg Kroah-Hartman
2013-05-06 20:44 ` [ 023/115] USB: ftdi_sio: correct ST Micro Connect Lite PIDs Greg Kroah-Hartman
2013-05-06 20:44 ` [ 024/115] USB: ftdi_sio: enable two UART ports on ST Microconnect Lite Greg Kroah-Hartman
2013-05-06 20:44 ` [ 025/115] USB: io_ti: fix TIOCGSERIAL Greg Kroah-Hartman
2013-05-06 20:44 ` [ 026/115] usbfs: Always allow ctrl requests with USB_RECIP_ENDPOINT on the ctrl ep Greg Kroah-Hartman
2013-05-06 20:44 ` [ 027/115] usb: chipidea: udc: fix memory access of shared memory on armv5 machines Greg Kroah-Hartman
2013-05-06 20:44 ` [ 028/115] usb: chipidea: udc: fix memory leak in _ep_nuke Greg Kroah-Hartman
2013-05-06 20:44 ` [ 029/115] usb: remove redundant tdi_reset Greg Kroah-Hartman
2013-05-06 20:44 ` [ 030/115] usb-storage: CY7C68300A chips do not support Cypress ATACB Greg Kroah-Hartman
2013-05-06 20:44 ` [ 031/115] s390/memory hotplug: prevent offline of active memory increments Greg Kroah-Hartman
2013-05-06 20:44 ` [ 032/115] xen/time: Fix kasprintf splat when allocating timer%d IRQ line Greg Kroah-Hartman
2013-05-06 20:44 ` [ 033/115] xen/smp: Fix leakage of timer interrupt line for every CPU online/offline Greg Kroah-Hartman
2013-05-06 20:44 ` [ 034/115] xen/smp/spinlock: Fix leakage of the spinlock " Greg Kroah-Hartman
2013-05-06 20:44 ` [ 035/115] serial_core.c: add put_device() after device_find_child() Greg Kroah-Hartman
2013-05-06 20:44 ` [ 036/115] arm: set the page table freeing ceiling to TASK_SIZE Greg Kroah-Hartman
2013-05-06 20:44 ` [ 037/115] gianfar: do not advertise any alarm capability Greg Kroah-Hartman
2013-05-06 20:44 ` [ 038/115] tty: fix up atime/mtime mess, take three Greg Kroah-Hartman
2013-05-06 20:44 ` [ 039/115] fbcon: when font is freed, clear also vc_font.data Greg Kroah-Hartman
2013-05-06 20:44 ` [ 040/115] tracing: Use stack of calling function for stack tracer Greg Kroah-Hartman
2013-05-06 20:44 ` [ 041/115] tracing: Fix stack tracer with fentry use Greg Kroah-Hartman
2013-05-06 20:44 ` Greg Kroah-Hartman [this message]
2013-05-06 20:44 ` [ 043/115] tracing: Fix off-by-one on allocating stat->pages Greg Kroah-Hartman
2013-05-06 20:44 ` [ 044/115] tracing: Check return value of tracing_init_dentry() Greg Kroah-Hartman
2013-05-06 20:44 ` [ 045/115] tracing: Reset ftrace_graph_filter_enabled if count is zero Greg Kroah-Hartman
2013-05-06 20:44 ` [ 046/115] i2c: xiic: must always write 16-bit words to TX_FIFO Greg Kroah-Hartman
2013-05-06 20:44 ` [ 047/115] crypto: crc32-pclmul - Use gas macro for pclmulqdq Greg Kroah-Hartman
2013-05-06 20:44 ` [ 048/115] Drivers: hv: vmbus: Fix a bug in hv_need_to_signal() Greg Kroah-Hartman
2013-05-06 20:44 ` [ 049/115] sysfs: fix use after free in case of concurrent read/write and readdir Greg Kroah-Hartman
2013-05-06 20:44 ` [ 050/115] Fix initialization of CMCI/CMCP interrupts Greg Kroah-Hartman
2013-05-06 20:44 ` [ 051/115] PCI / ACPI: Dont query OSC support with all possible controls Greg Kroah-Hartman
2013-05-06 20:44 ` [ 052/115] PCI/PM: Fix fallback to PCI_D0 in pci_platform_power_transition() Greg Kroah-Hartman
2013-05-06 20:44 ` [ 053/115] rt2x00: Fix transmit power troubles on some Ralink RT30xx cards Greg Kroah-Hartman
2013-05-06 20:44 ` [ 054/115] Wrong asm register contraints in the futex implementation Greg Kroah-Hartman
2013-05-06 20:44 ` [ 055/115] Wrong asm register contraints in the kvm implementation Greg Kroah-Hartman
2013-05-06 20:44 ` [ 056/115] fs/fscache/stats.c: fix memory leak Greg Kroah-Hartman
2013-05-06 20:44 ` [ 057/115] mm: allow arch code to control the user page table ceiling Greg Kroah-Hartman
2013-05-06 20:44 ` [ 058/115] TPM: Retry SaveState command in suspend path Greg Kroah-Hartman
2013-05-06 20:44 ` [ 059/115] ALSA: emu10k1: Fix dock firmware loading Greg Kroah-Hartman
2013-05-06 20:44 ` [ 060/115] ALSA: snd-usb: try harder to find USB_DT_CS_ENDPOINT Greg Kroah-Hartman
2013-05-06 20:44 ` [ 061/115] ALSA: usb: Add quirk for 192KHz recording on E-Mu devices Greg Kroah-Hartman
2013-05-06 20:44 ` [ 062/115] ALSA: usb-audio: disable autopm for MIDI devices Greg Kroah-Hartman
2013-05-06 20:44 ` [ 063/115] ALSA: usb-audio: Fix autopm error during probing Greg Kroah-Hartman
2013-05-06 20:45 ` [ 064/115] ALSA: USB: adjust for changed 3.8 USB API Greg Kroah-Hartman
2013-05-06 20:45 ` [ 065/115] ALSA: hda - Fix aamix activation with loopback control on VIA codecs Greg Kroah-Hartman
2013-05-06 20:45 ` [ 066/115] ALSA: hda - Add the support for ALC286 codec Greg Kroah-Hartman
2013-05-06 20:45 ` [ 067/115] ARM: 7702/1: Set the page table freeing ceiling to TASK_SIZE Greg Kroah-Hartman
2013-05-07 0:00 ` Hugh Dickins
2013-05-07 17:23 ` Greg Kroah-Hartman
2013-05-06 20:45 ` [ 068/115] ASoC: max98088: Fix logging of hardware revision Greg Kroah-Hartman
2013-05-06 20:45 ` [ 069/115] hrtimer: Fix ktime_add_ns() overflow on 32bit architectures Greg Kroah-Hartman
2013-05-06 20:45 ` [ 070/115] hrtimer: Add expiry time overflow check in hrtimer_interrupt Greg Kroah-Hartman
2013-05-06 20:45 ` [ 071/115] swap: redirty page if page write fails on swap file Greg Kroah-Hartman
2013-05-06 20:45 ` [ 072/115] mm: swap: mark swap pages writeback before queueing for direct IO Greg Kroah-Hartman
2013-05-06 20:45 ` [ 073/115] drivers/rtc/rtc-cmos.c: dont disable hpet emulation on suspend Greg Kroah-Hartman
2013-05-06 20:45 ` [ 074/115] drivers/rtc/rtc-at91rm9200.c: fix missing iounmap Greg Kroah-Hartman
2013-05-06 20:45 ` [ 075/115] libata: acpi: make ata_ap_acpi_handle not block Greg Kroah-Hartman
2013-05-06 20:45 ` [ 076/115] ACPI: Fix wrong parameter passed to memblock_reserve Greg Kroah-Hartman
2013-05-06 20:45 ` [ 077/115] ACPI / thermal: do not always return THERMAL_TREND_RAISING for active trip points Greg Kroah-Hartman
2013-05-06 20:45 ` [ 078/115] cgroup: fix an off-by-one bug which may trigger BUG_ON() Greg Kroah-Hartman
2013-05-06 20:45 ` [ 079/115] cgroup: fix broken file xattrs Greg Kroah-Hartman
2013-05-06 20:45 ` [ 080/115] localmodconfig: Process source kconfig files as they are found Greg Kroah-Hartman
2013-05-06 20:45 ` [ 081/115] clockevents: Set dummy handler on CPU_DEAD shutdown Greg Kroah-Hartman
2013-05-06 20:45 ` [ 082/115] sata_highbank: Rename proc_name to the module name Greg Kroah-Hartman
2013-05-06 20:45 ` [ 083/115] inotify: invalid mask should return a error number but not set it Greg Kroah-Hartman
2013-05-06 20:45 ` [ 084/115] fs/dcache.c: add cond_resched() to shrink_dcache_parent() Greg Kroah-Hartman
2013-05-06 20:45 ` [ 085/115] exec: do not abuse ->cred_guard_mutex in threadgroup_lock() Greg Kroah-Hartman
2013-05-06 20:45 ` [ 086/115] LOCKD: Ensure that nlmclnt_block resets block->b_status after a server reboot Greg Kroah-Hartman
2013-05-06 20:45 ` [ 087/115] md: bad block list should default to disabled Greg Kroah-Hartman
2013-05-06 20:45 ` [ 088/115] MD: ignore discard request for hard disks of hybid raid1/raid10 array Greg Kroah-Hartman
2013-05-06 20:45 ` [ 089/115] NFSv4: Handle NFS4ERR_DELAY and NFS4ERR_GRACE in nfs4_lock_delegation_recall Greg Kroah-Hartman
2013-05-06 20:45 ` [ 090/115] NFSv4: Handle NFS4ERR_DELAY and NFS4ERR_GRACE in nfs4_open_delegation_recall Greg Kroah-Hartman
2013-05-06 20:45 ` [ 091/115] nfsd4: dont close read-write opens too soon Greg Kroah-Hartman
2013-05-06 20:45 ` [ 092/115] nfsd: dont run get_file if nfs4_preprocess_stateid_op return error Greg Kroah-Hartman
2013-05-06 20:45 ` [ 093/115] nfsd: use kmem_cache_free() instead of kfree() Greg Kroah-Hartman
2013-05-06 20:45 ` [ 094/115] nfsd: Decode and send 64bit time values Greg Kroah-Hartman
2013-05-06 20:45 ` [ 095/115] wireless: regulatory: fix channel disabling race condition Greg Kroah-Hartman
2013-05-06 20:45 ` [ 096/115] ipc: sysv shared memory limited to 8TiB Greg Kroah-Hartman
2013-05-06 20:45 ` [ 097/115] ixgbe: fix EICR write in ixgbe_msix_other Greg Kroah-Hartman
2013-05-06 20:45 ` [ 098/115] e1000e: fix numeric overflow in phc settime method Greg Kroah-Hartman
2013-05-06 20:45 ` [ 099/115] x86-64, init: Do not set NX bits on non-NX capable hardware Greg Kroah-Hartman
2013-05-06 20:45 ` [ 100/115] ext4/jbd2: dont wait (forever) for stale tid caused by wraparound Greg Kroah-Hartman
2013-05-06 20:45 ` [ 101/115] jbd2: fix race between jbd2_journal_remove_checkpoint and ->j_commit_callback Greg Kroah-Hartman
2013-05-06 20:45 ` [ 102/115] ext4: fix journal callback list traversal Greg Kroah-Hartman
2013-05-06 20:45 ` [ 103/115] ext4: unregister es_shrinker if mount failed Greg Kroah-Hartman
2013-05-06 20:45 ` [ 104/115] ext4: fix big-endian bug in metadata checksum calculations Greg Kroah-Hartman
2013-05-06 20:45 ` [ 105/115] ext4: fix online resizing for ext3-compat file systems Greg Kroah-Hartman
2013-05-06 20:45 ` [ 106/115] ext4: fix Kconfig documentation for CONFIG_EXT4_DEBUG Greg Kroah-Hartman
2013-05-06 20:45 ` [ 107/115] do_mount(): fix a leak introduced in 3.9 ("mount: consolidate permission checks") Greg Kroah-Hartman
2013-05-06 20:45 ` [ 108/115] mmc: at91/avr32/atmel-mci: fix DMA-channel leak on module unload Greg Kroah-Hartman
2013-05-06 20:45 ` [ 109/115] Give the OID registry file module info to avoid kernel tainting Greg Kroah-Hartman
2013-05-06 20:45 ` [ 110/115] KVM: X86 emulator: fix source operand decoding for 8bit mov[zs]x instructions Greg Kroah-Hartman
2013-05-06 20:45 ` [ 111/115] x86: Eliminate irq_mis_count counted in arch_irq_stat Greg Kroah-Hartman
2013-05-06 20:45 ` [ 112/115] mmc: core: Fix bit width test failing on old eMMC cards Greg Kroah-Hartman
2013-05-06 20:45 ` [ 113/115] mmc: atmel-mci: pio hang on block errors Greg Kroah-Hartman
2013-05-06 20:45 ` [ 114/115] rcutrace: single_open() leaks Greg Kroah-Hartman
2013-05-06 20:45 ` [ 115/115] mfd: adp5520: Restore mode bits on resume Greg Kroah-Hartman
[not found] ` <CAKocOOOcpAeLo=K5usRnk2w5S4tXx2Y4rXe9bff=2ba=ZWdDJA@mail.gmail.com>
2013-05-07 19:11 ` Fwd: [ 000/115] 3.9.1-stable review Shuah Khan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20130506203059.871305620@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=rostedt@goodmis.org \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox