linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Cheng Jian <cj.chengjian@huawei.com>,
	"Steven Rostedt (VMware)" <rostedt@goodmis.org>,
	Sasha Levin <sashal@kernel.org>
Subject: [PATCH 4.14 05/53] ftrace: Enable trampoline when rec count returns back to one
Date: Mon,  5 Aug 2019 15:02:30 +0200	[thread overview]
Message-ID: <20190805124928.616961161@linuxfoundation.org> (raw)
In-Reply-To: <20190805124927.973499541@linuxfoundation.org>

[ Upstream commit a124692b698b00026a58d89831ceda2331b2e1d0 ]

Custom trampolines can only be enabled if there is only a single ops
attached to it. If there's only a single callback registered to a function,
and the ops has a trampoline registered for it, then we can call the
trampoline directly. This is very useful for improving the performance of
ftrace and livepatch.

If more than one callback is registered to a function, the general
trampoline is used, and the custom trampoline is not restored back to the
direct call even if all the other callbacks were unregistered and we are
back to one callback for the function.

To fix this, set FTRACE_FL_TRAMP flag if rec count is decremented
to one, and the ops that left has a trampoline.

Testing After this patch :

insmod livepatch_unshare_files.ko
cat /sys/kernel/debug/tracing/enabled_functions

	unshare_files (1) R I	tramp: 0xffffffffc0000000(klp_ftrace_handler+0x0/0xa0) ->ftrace_ops_assist_func+0x0/0xf0

echo unshare_files > /sys/kernel/debug/tracing/set_ftrace_filter
echo function > /sys/kernel/debug/tracing/current_tracer
cat /sys/kernel/debug/tracing/enabled_functions

	unshare_files (2) R I ->ftrace_ops_list_func+0x0/0x150

echo nop > /sys/kernel/debug/tracing/current_tracer
cat /sys/kernel/debug/tracing/enabled_functions

	unshare_files (1) R I	tramp: 0xffffffffc0000000(klp_ftrace_handler+0x0/0xa0) ->ftrace_ops_assist_func+0x0/0xf0

Link: http://lkml.kernel.org/r/1556969979-111047-1-git-send-email-cj.chengjian@huawei.com

Signed-off-by: Cheng Jian <cj.chengjian@huawei.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 kernel/trace/ftrace.c | 28 +++++++++++++++-------------
 1 file changed, 15 insertions(+), 13 deletions(-)

diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
index c4a0ad18c8593..7420f5f360947 100644
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -1712,6 +1712,11 @@ static bool test_rec_ops_needs_regs(struct dyn_ftrace *rec)
 	return  keep_regs;
 }
 
+static struct ftrace_ops *
+ftrace_find_tramp_ops_any(struct dyn_ftrace *rec);
+static struct ftrace_ops *
+ftrace_find_tramp_ops_next(struct dyn_ftrace *rec, struct ftrace_ops *ops);
+
 static bool __ftrace_hash_rec_update(struct ftrace_ops *ops,
 				     int filter_hash,
 				     bool inc)
@@ -1840,15 +1845,17 @@ static bool __ftrace_hash_rec_update(struct ftrace_ops *ops,
 			}
 
 			/*
-			 * If the rec had TRAMP enabled, then it needs to
-			 * be cleared. As TRAMP can only be enabled iff
-			 * there is only a single ops attached to it.
-			 * In otherwords, always disable it on decrementing.
-			 * In the future, we may set it if rec count is
-			 * decremented to one, and the ops that is left
-			 * has a trampoline.
+			 * The TRAMP needs to be set only if rec count
+			 * is decremented to one, and the ops that is
+			 * left has a trampoline. As TRAMP can only be
+			 * enabled if there is only a single ops attached
+			 * to it.
 			 */
-			rec->flags &= ~FTRACE_FL_TRAMP;
+			if (ftrace_rec_count(rec) == 1 &&
+			    ftrace_find_tramp_ops_any(rec))
+				rec->flags |= FTRACE_FL_TRAMP;
+			else
+				rec->flags &= ~FTRACE_FL_TRAMP;
 
 			/*
 			 * flags will be cleared in ftrace_check_record()
@@ -2041,11 +2048,6 @@ static void print_ip_ins(const char *fmt, const unsigned char *p)
 		printk(KERN_CONT "%s%02x", i ? ":" : "", p[i]);
 }
 
-static struct ftrace_ops *
-ftrace_find_tramp_ops_any(struct dyn_ftrace *rec);
-static struct ftrace_ops *
-ftrace_find_tramp_ops_next(struct dyn_ftrace *rec, struct ftrace_ops *ops);
-
 enum ftrace_bug_type ftrace_bug_type;
 const void *ftrace_expected;
 
-- 
2.20.1




  parent reply	other threads:[~2019-08-05 13:07 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-05 13:02 [PATCH 4.14 00/53] 4.14.137-stable review Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 01/53] ARM: riscpc: fix DMA Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 02/53] ARM: dts: rockchip: Make rk3288-veyron-minnie run at hs200 Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 03/53] ARM: dts: rockchip: Make rk3288-veyron-mickeys emmc work again Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 04/53] ARM: dts: rockchip: Mark that the rk3288 timer might stop in suspend Greg Kroah-Hartman
2019-08-05 13:02 ` Greg Kroah-Hartman [this message]
2019-08-05 13:02 ` [PATCH 4.14 06/53] kernel/module.c: Only return -EEXIST for modules that have finished loading Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 07/53] MIPS: lantiq: Fix bitfield masking Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 08/53] dmaengine: rcar-dmac: Reject zero-length slave DMA requests Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 09/53] clk: tegra210: fix PLLU and PLLU_OUT1 Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 10/53] fs/adfs: super: fix use-after-free bug Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 11/53] btrfs: fix minimum number of chunk errors for DUP Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 12/53] cifs: Fix a race condition with cifs_echo_request Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 13/53] ceph: fix improper use of smp_mb__before_atomic() Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 14/53] ceph: return -ERANGE if virtual xattr value didnt fit in buffer Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 15/53] ACPI: blacklist: fix clang warning for unused DMI table Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 16/53] scsi: zfcp: fix GCC compiler warning emitted with -Wmaybe-uninitialized Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 17/53] x86: kvm: avoid constant-conversion warning Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 18/53] ACPI: fix false-positive -Wuninitialized warning Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 19/53] be2net: Signal that the device cannot transmit during reconfiguration Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 20/53] x86/apic: Silence -Wtype-limits compiler warnings Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 21/53] x86: math-emu: Hide clang warnings for 16-bit overflow Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 22/53] mm/cma.c: fail if fixed declaration cant be honored Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 23/53] coda: add error handling for fget Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 24/53] coda: fix build using bare-metal toolchain Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 25/53] uapi linux/coda_psdev.h: move upc_req definition from uapi to kernel side headers Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 26/53] drivers/rapidio/devices/rio_mport_cdev.c: NUL terminate some strings Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 27/53] ipc/mqueue.c: only perform resource calculation if user valid Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 28/53] xen/pv: Fix a boot up hang revealed by int3 self test Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 29/53] x86/kvm: Dont call kvm_spurious_fault() from .fixup Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 30/53] x86/paravirt: Fix callee-saved function ELF sizes Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 31/53] x86, boot: Remove multiple copy of static function sanitize_boot_params() Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 32/53] drm/nouveau: fix memory leak in nouveau_conn_reset() Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 33/53] kbuild: initialize CLANG_FLAGS correctly in the top Makefile Greg Kroah-Hartman
2019-08-05 13:02 ` [PATCH 4.14 34/53] Btrfs: fix incremental send failure after deduplication Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 35/53] Btrfs: fix race leading to fs corruption after transaction abort Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 36/53] mmc: dw_mmc: Fix occasional hang after tuning on eMMC Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 37/53] gpiolib: fix incorrect IRQ requesting of an active-low lineevent Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 38/53] IB/hfi1: Fix Spectre v1 vulnerability Greg Kroah-Hartman
2019-08-05 14:16   ` Gustavo A. R. Silva
2019-08-26  9:06     ` Greg Kroah-Hartman
2019-08-26 21:23       ` Gustavo A. R. Silva
2019-08-05 13:03 ` [PATCH 4.14 39/53] selinux: fix memory leak in policydb_init() Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 40/53] s390/dasd: fix endless loop after read unit address configuration Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 41/53] parisc: Fix build of compressed kernel even with debug enabled Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 42/53] drivers/perf: arm_pmu: Fix failure path in PM notifier Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 43/53] nbd: replace kill_bdev() with __invalidate_device() again Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 44/53] xen/swiotlb: fix condition for calling xen_destroy_contiguous_region() Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 45/53] IB/mlx5: Fix unreg_umr to ignore the mkey state Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 46/53] IB/mlx5: Use direct mkey destroy command upon UMR unreg failure Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 47/53] IB/mlx5: Move MRs to a kernel PD when freeing them to the MR cache Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 48/53] IB/mlx5: Fix RSS Toeplitz setup to be aligned with the HW specification Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 49/53] IB/hfi1: Check for error on call to alloc_rsm_map_table Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 50/53] eeprom: at24: make spd world-readable again Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 51/53] objtool: Support GCC 9 cold subfunction naming scheme Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 52/53] gcc-9: properly declare the {pv,hv}clock_page storage Greg Kroah-Hartman
2019-08-05 13:03 ` [PATCH 4.14 53/53] x86/vdso: Prevent segfaults due to hoisted vclock reads Greg Kroah-Hartman
2019-08-06  1:06 ` [PATCH 4.14 00/53] 4.14.137-stable review shuah
2019-08-06  3:34 ` Naresh Kamboju
2019-08-06  7:16 ` Jack Wang
2019-08-06 15:49 ` Guenter Roeck
2019-08-06 18:29 ` Jon Hunter

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=20190805124928.616961161@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=cj.chengjian@huawei.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=sashal@kernel.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;
as well as URLs for NNTP newsgroup(s).