Scenario 1, ARMv7
=================
If code in arch/arm/kernel/ftrace.c would operate on mcount() pointer
the following may be generated:
00000230 <prealloc_fixed_plts>:
230: b5f8 push {r3, r4, r5, r6, r7, lr}
232: b500 push {lr}
234: f7ff fffe bl 0 <__gnu_mcount_nc>
234: R_ARM_THM_CALL __gnu_mcount_nc
238: f240 0600 movw r6, #0
238: R_ARM_THM_MOVW_ABS_NC __gnu_mcount_nc
23c: f8d0 1180 ldr.w r1, [r0, #384] ; 0x180
FTRACE currently is not able to deal with it:
WARNING: CPU: 0 PID: 0 at .../kernel/trace/ftrace.c:1979 ftrace_bug+0x1ad/0x230()
...
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.4.116-... #1
...
[<c0314e3d>] (unwind_backtrace) from [<c03115e9>] (show_stack+0x11/0x14)
[<c03115e9>] (show_stack) from [<c051a7f1>] (dump_stack+0x81/0xa8)
[<c051a7f1>] (dump_stack) from [<c0321c5d>] (warn_slowpath_common+0x69/0x90)
[<c0321c5d>] (warn_slowpath_common) from [<c0321cf3>] (warn_slowpath_null+0x17/0x1c)
[<c0321cf3>] (warn_slowpath_null) from [<c038ee9d>] (ftrace_bug+0x1ad/0x230)
[<c038ee9d>] (ftrace_bug) from [<c038f1f9>] (ftrace_process_locs+0x27d/0x444)
[<c038f1f9>] (ftrace_process_locs) from [<c08915bd>] (ftrace_init+0x91/0xe8)
[<c08915bd>] (ftrace_init) from [<c0885a67>] (start_kernel+0x34b/0x358)
[<c0885a67>] (start_kernel) from [<00308095>] (0x308095)
---[ end trace cb88537fdc8fa200 ]---
ftrace failed to modify [<c031266c>] prealloc_fixed_plts+0x8/0x60
actual: 44:f2:e1:36
ftrace record flags: 0
(0) expected tramp: c03143e9
Scenario 2, ARMv4T
==================
ftrace: allocating 14435 entries in 43 pages
------------[ cut here ]------------
WARNING: CPU: 0 PID: 0 at kernel/trace/ftrace.c:2029 ftrace_bug+0x204/0x310
CPU: 0 PID: 0 Comm: swapper Not tainted 4.19.5 #1
Hardware name: Cirrus Logic EDB9302 Evaluation Board
[<c0010a24>] (unwind_backtrace) from [<c000ecb0>] (show_stack+0x20/0x2c)
[<c000ecb0>] (show_stack) from [<c03c72e8>] (dump_stack+0x20/0x30)
[<c03c72e8>] (dump_stack) from [<c0021c18>] (__warn+0xdc/0x104)
[<c0021c18>] (__warn) from [<c0021d7c>] (warn_slowpath_null+0x4c/0x5c)
[<c0021d7c>] (warn_slowpath_null) from [<c0095360>] (ftrace_bug+0x204/0x310)
[<c0095360>] (ftrace_bug) from [<c04dabac>] (ftrace_init+0x3b4/0x4d4)
[<c04dabac>] (ftrace_init) from [<c04cef4c>] (start_kernel+0x20c/0x410)
[<c04cef4c>] (start_kernel) from [<00000000>] ( (null))
---[ end trace 0506a2f5dae6b341 ]---
ftrace failed to modify
[<c000c350>] perf_trace_sys_exit+0x5c/0xe8
actual: 1e:ff:2f:e1
Initializing ftrace call sites
ftrace record flags: 0
(0)
expected tramp: c000fb24
The analysis for this problem has been already performed previously,
refer to the link below.
Fix the above problems by allowing only selected reloc types in
__mcount_loc. The list itself comes from the legacy recordmcount.pl
script.
Link: https://lore.kernel.org/lkml/56961010.6000806@pengutronix.de/
Cc: stable@vger.kernel.org
Fixes: ed60453fa8 ("ARM: 6511/1: ftrace: add ARM support for C version of recordmcount")
Signed-off-by: Alexander Sverdlin <alexander.sverdlin@nokia.com>
Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
cleanup() mostly frees/unmaps the malloc'd/privately-mapped
copy of the ELF file recordmcount is working on, which is
set up in mmap_file(). It also deals with positioning within
the pseduo prive-mapping of the file and appending to the ELF
file.
Split into two steps:
mmap_cleanup() for the mapping itself
file_append_cleanup() for allocations storing the
appended ELF data.
Also, move the global variable initializations out of the main,
per-object-file loop and nearer to the alloc/init (mmap_file())
and two cleanup functions so we can more clearly see how they're
related.
Link: http://lkml.kernel.org/r/2a387ac86d133d22c68f57b9933c32bab1d09a2d.1564596289.git.mhelsley@vmware.com
Signed-off-by: Matt Helsley <mhelsley@vmware.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Redundant cleanup calls were introduced when transitioning from
the old error/success handling via setjmp/longjmp -- the longjmp
ensured the cleanup() call only happened once but replacing
the success_file()/fail_file() calls with cleanup() meant that
multiple cleanup() calls can happen as we return from function
calls.
In do_file(), looking just before and after the "goto out" jumps we
can see that multiple cleanups() are being performed. We remove
cleanup() calls from the nested functions because it makes the code
easier to review -- the resources being cleaned up are generally
allocated and initialized in the callers so freeing them there
makes more sense.
Other redundant cleanup() calls:
mmap_file() is only called from do_file() and, if mmap_file() fails,
then we goto out and do cleanup() there too.
write_file() is only called from do_file() and do_file()
calls cleanup() unconditionally after returning from write_file()
therefore the cleanup() calls in write_file() are not necessary.
find_secsym_ndx(), called from do_func()'s for-loop, when we are
cleaning up here it's obvious that we break out of the loop and
do another cleanup().
__has_rel_mcount() is called from two parts of do_func()
and calls cleanup(). In theory we move them into do_func(), however
these in turn prove redundant so another simplification step
removes them as well.
Link: http://lkml.kernel.org/r/de197e17fc5426623a847ea7cf3a1560a7402a4b.1564596289.git.mhelsley@vmware.com
Signed-off-by: Matt Helsley <mhelsley@vmware.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Recordmcount uses setjmp/longjmp to manage control flow as
it reads and then writes the ELF file. This unusual control
flow is hard to follow and check in addition to being unlike
kernel coding style.
So we rewrite these paths to use regular return values to
indicate error/success. When an error or previously-completed object
file is found we return an error code following kernel
coding conventions -- negative error values and 0 for success when
we're not returning a pointer. We return NULL for those that fail
and return non-NULL pointers otherwise.
One oddity is already_has_rel_mcount -- there we use pointer comparison
rather than string comparison to differentiate between
previously-processed object files and returning the name of a text
section.
Link: http://lkml.kernel.org/r/8ba8633d4afe444931f363c8d924bf9565b89a86.1564596289.git.mhelsley@vmware.com
Signed-off-by: Matt Helsley <mhelsley@vmware.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
uwrite() works within the pseudo-mapping and extends it as necessary
without needing the file descriptor (fd) parameter passed to it.
Similarly, ulseek() doesn't need its fd parameter. These parameters
were only added because the functions bear a conceptual resemblance
to write() and lseek(). Worse, they obscure the fact that at the time
uwrite() and ulseek() are called fd_map is not a valid file descriptor.
Remove the unused file descriptor parameters that make it look like
fd_map is still valid.
Link: http://lkml.kernel.org/r/2a136e820ee208469d375265c7b8eb28570749a0.1563992889.git.mhelsley@vmware.com
Signed-off-by: Matt Helsley <mhelsley@vmware.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
uread() is only used to initialize the ELF file's pseudo
private-memory mapping while uwrite() and ulseek() work within
the pseudo-mapping and extend it as necessary. Thus it is not
a complementary function to uwrite() and ulseek(). It also makes
no sense to do cleanups inside uread() when its only caller,
mmap_file(), is doing the relevant allocations and associated
initializations.
Therefore it's clearer to use a plain read() call to initialize the
data in mmap_file() and remove uread().
Link: http://lkml.kernel.org/r/31a87c22b19150cec1c8dc800c8b0873a2741703.1563992889.git.mhelsley@vmware.com
Signed-off-by: Matt Helsley <mhelsley@vmware.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Based on 1 normalized pattern(s):
licensed under the gnu general public license version 2 gplv2
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-only
has been chosen to replace the boilerplate/reference in 5 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Armijn Hemel <armijn@tjaldur.nl>
Reviewed-by: Allison Randal <allison@lohutok.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190531081036.993848054@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When building with -ffunction-sections, the compiler will place each
function into its own ELF section, prefixed with ".text". For example,
a simple test module with functions test_module_do_work() and
test_module_wq_func():
% objdump --section-headers test_module.o | awk '/\.text/{print $2}'
.text
.text.test_module_do_work
.text.test_module_wq_func
.init.text
.exit.text
Adjust the recordmcount scripts to look for ".text" as a section name
prefix. This will ensure that those functions will be included in the
__mcount_loc relocations:
% objdump --reloc --section __mcount_loc test_module.o
OFFSET TYPE VALUE
0000000000000000 R_X86_64_64 .text.test_module_do_work
0000000000000008 R_X86_64_64 .text.test_module_wq_func
0000000000000010 R_X86_64_64 .init.text
Link: http://lkml.kernel.org/r/1542745158-25392-2-git-send-email-joe.lawrence@redhat.com
Signed-off-by: Joe Lawrence <joe.lawrence@redhat.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
scripts/kallsyms.c: function write_src:
"printf", the #1 format specifier "d" need arg type "int",
but the according arg "table_cnt" has type "unsigned int"
scripts/recordmcount.c: function do_file:
"fprintf", the #1 format specifier "d" need arg type "int",
but the according arg "(*w2)(ehdr->e_machine)" has type "unsigned int"
scripts/recordmcount.h: function find_secsym_ndx:
"fprintf", the #1 format specifier "d" need arg type "int",
but the according arg "txtndx" has type "unsigned int"
Signed-off-by: nixiaoming <nixiaoming@huawei.com>
Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Now that arch/metag/ has been removed, drop a bunch of metag references
in various codes across the whole tree:
- VM_GROWSUP and __VM_ARCH_SPECIFIC_1.
- MT_METAG_* ELF note types.
- METAG Kconfig dependencies (FRAME_POINTER) and ranges
(MAX_STACK_SIZE_MB).
- metag cases in tools (checkstack.pl, recordmcount.c, perf).
Signed-off-by: James Hogan <jhogan@kernel.org>
Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: linux-mm@kvack.org
Cc: linux-metag@vger.kernel.org
Adding a hook into free_reserve_area() that informs ftrace that boot up init
text is being free, lets ftrace safely remove those init functions from its
records, which keeps ftrace from trying to modify text that no longer
exists.
Note, this still does not allow for tracing .init text of modules, as
modules require different work for freeing its init code.
Link: http://lkml.kernel.org/r/1488502497.7212.24.camel@linux.intel.com
Cc: linux-mm@kvack.org
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Peter Zijlstra <peterz@infradead.org>
Requested-by: Todd Brandt <todd.e.brandt@linux.intel.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
In similar spirit to x86 and arm64 support, add a make_nop_arm()
to replace calls to mcount with a nop in sections that aren't
traced.
Link: http://lkml.kernel.org/r/20161018234200.5804-1-sboyd@codeaurora.org
Acked-by: Russell King <rmk+kernel@armlinux.org.uk>
Acked-by: Rabin Vincent <rabin@rab.in>
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When doing an nmi backtrace of many cores, most of which are idle, the
output is a little overwhelming and very uninformative. Suppress
messages for cpus that are idling when they are interrupted and just
emit one line, "NMI backtrace for N skipped: idling at pc 0xNNN".
We do this by grouping all the cpuidle code together into a new
.cpuidle.text section, and then checking the address of the interrupted
PC to see if it lies within that section.
This commit suitably tags x86 and tile idle routines, and only adds in
the minimal framework for other architectures.
Link: http://lkml.kernel.org/r/1472487169-14923-5-git-send-email-cmetcalf@mellanox.com
Signed-off-by: Chris Metcalf <cmetcalf@mellanox.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Daniel Thompson <daniel.thompson@linaro.org> [arm]
Tested-by: Petr Mladek <pmladek@suse.com>
Cc: Aaron Tomlin <atomlin@redhat.com>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
be7635e728 ("arch, ftrace: for KASAN put hard/soft IRQ entries into
separate sections") added .softirqentry.text section, but it was not added
to recordmcount. So functions in the section are untracable. Add the
section to scripts/recordmcount.c and scripts/recordmcount.pl.
Fixes: be7635e728 ("arch, ftrace: for KASAN put hard/soft IRQ entries into separate sections")
Link: http://lkml.kernel.org/r/1474902626-73468-1-git-send-email-dvyukov@google.com
Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Acked-by: Steve Rostedt <rostedt@goodmis.org>
Cc: <stable@vger.kernel.org> [4.6+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
glibc recently did a sync up (94e73c95d9b5 "elf.h: Sync with the gabi
webpage") that added a #define for EM_METAG but did not add relocations
This triggers build errors:
scripts/recordmcount.c: In function 'do_file':
scripts/recordmcount.c:466:28: error: 'R_METAG_ADDR32' undeclared (first use in this function)
case EM_METAG: reltype = R_METAG_ADDR32;
^~~~~~~~~~~~~~
scripts/recordmcount.c:466:28: note: each undeclared identifier is reported only once for each function it appears in
scripts/recordmcount.c:468:20: error: 'R_METAG_NONE' undeclared (first use in this function)
rel_type_nop = R_METAG_NONE;
^~~~~~~~~~~~
Work around this change with some more #ifdefery for the relocations.
Fedora Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1354034
Link: http://lkml.kernel.org/r/1468005530-14757-1-git-send-email-labbott@redhat.com
Cc: stable@vger.kernel.org # v3.9+
Cc: James Hogan <james.hogan@imgtec.com>
Fixes: 00512bdd45 ("metag: ftrace support")
Reported-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Fix build warning:
scripts/recordmcount.c:589:4: warning: format not a string
literal and no format arguments [-Wformat-security]
sprintf("%s: failed\n", file);
Fixes: a50bd43935 ("ftrace/scripts: Have recordmcount copy the object file")
Link: http://lkml.kernel.org/r/1451516801-16951-1-git-send-email-colin.king@canonical.com
Cc: Li Bin <huawei.libin@huawei.com>
Cc: Russell King <rmk+kernel@arm.linux.org.uk>
Cc: Will Deacon <will.deacon@arm.com>
Cc: stable@vger.kernel.org # 2.6.37+
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Russell King found that he had weird side effects when compiling the kernel
with hard linked ccache. The reason was that recordmcount modified the
kernel in place via mmap, and when a file gets modified twice by
recordmcount, it will complain about it. To fix this issue, Russell wrote a
patch that checked if the file was hard linked more than once and would
unlink it if it was.
Linus Torvalds was not happy with the fact that recordmcount does this in
place modification. Instead of doing the unlink only if the file has two or
more hard links, it does the unlink all the time. In otherwords, it always
does a copy if it changed something. That is, it does the write out if a
change was made.
Cc: stable@vger.kernel.org # 2.6.37+
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
recordmcount edits the file in-place, which can cause problems when
using ccache in hardlink mode. Arrange for recordmcount to break a
hardlinked object.
Link: http://lkml.kernel.org/r/E1a7MVT-0000et-62@rmk-PC.arm.linux.org.uk
Cc: stable@vger.kernel.org # 2.6.37+
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
By now, the recordmcount only records the function that in
following sections:
.text/.ref.text/.sched.text/.spinlock.text/.irqentry.text/
.kprobes.text/.text.unlikely
For the function that not in these sections, the call mcount
will be in place and not be replaced when kernel boot up. And
it will bring performance overhead, such as do_mem_abort (in
.exception.text section). This patch make the call mcount to
nop for this case in recordmcount.
Link: http://lkml.kernel.org/r/1446019445-14421-1-git-send-email-huawei.libin@huawei.com
Link: http://lkml.kernel.org/r/1446193864-24593-4-git-send-email-huawei.libin@huawei.com
Cc: <lkp@intel.com>
Cc: <catalin.marinas@arm.com>
Cc: <takahiro.akashi@linaro.org>
Cc: <stable@vger.kernel.org> # 3.18+
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Li Bin <huawei.libin@huawei.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Although, the default value of rel_type_nop is zero, and the value
of R_386_NONE/R_X86_64_NONE is zero too, but it should be assigned
a meaningful value explicitly, otherwise it looks confused.
Assign R_386_NONE to rel_type_nop for 386, assign R_X86_64_NONE
to rel_type_nop for x86_64.
Link: http://lkml.kernel.org/r/1446020606-16352-1-git-send-email-huawei.libin@huawei.com
Signed-off-by: Li Bin <huawei.libin@huawei.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
If the function tracer is enabled, allow to set kprobes on the first
instruction of a function (which is the function trace caller):
If no kprobe is set handling of enabling and disabling function tracing
of a function simply patches the first instruction. Either it is a nop
(right now it's an unconditional branch, which skips the mcount block),
or it's a branch to the ftrace_caller() function.
If a kprobe is being placed on a function tracer calling instruction
we encode if we actually have a nop or branch in the remaining bytes
after the breakpoint instruction (illegal opcode).
This is possible, since the size of the instruction used for the nop
and branch is six bytes, while the size of the breakpoint is only
two bytes.
Therefore the first two bytes contain the illegal opcode and the last
four bytes contain either "0" for nop or "1" for branch. The kprobes
code will then execute/simulate the correct instruction.
Instruction patching for kprobes and function tracer is always done
with stop_machine(). Therefore we don't have any races where an
instruction is patched concurrently on a different cpu.
Besides that also the program check handler which executes the function
trace caller instruction won't be executed concurrently to any
stop_machine() execution.
This allows to keep full fault based kprobes handling which generates
correct pt_regs contents automatically.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
31 bit and 64 bit diverge more and more and it is rather painful
to keep both parts running.
To make things simpler just remove the 31 bit support which nobody
uses anyway.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Pull kbuild misc updates from Michal Marek:
"This is the non-critical part of kbuild for v3.16-rc1:
- make deb-pkg can do s390x and arm64
- new patterns in scripts/tags.sh
- scripts/tags.sh skips userspace tools' sources (which sometimes
have copies of kernel structures) and symlinks
- improvements to the objdiff tool
- two new coccinelle patches
- other minor fixes"
* 'misc' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
scripts: objdiff: support directories for the augument of record command
scripts: objdiff: fix a comment
scripts: objdiff: change the extension of disassembly from .o to .dis
scripts: objdiff: improve path flexibility for record command
scripts: objdiff: remove unnecessary code
scripts: objdiff: direct error messages to stderr
scripts: objdiff: get the path to .tmp_objdiff more simply
deb-pkg: Add automatic support for s390x architecture
coccicheck: Add unneeded return variable test
kbuild: Fix a typo in documentation
kbuild: trivial - use tabs for code indent where possible
kbuild: trivial - remove trailing empty lines
coccinelle: Check for missing NULL terminators in of_device_id tables
scripts/tags.sh: ignore symlink'ed source files
scripts/tags.sh: add regular expression replacement pattern for memcg
builddeb: add arm64 in the supported architectures
builddeb: use $OBJCOPY variable instead of objcopy
scripts/tags.sh: ignore code of user space tools
scripts/tags.sh: add pattern for DEFINE_HASHTABLE
.gitignore: ignore Module.symvers in all directories
Recordmcount utility under scripts is run, after compiling each object,
to find out all the locations of calling _mcount() and put them into
specific seciton named __mcount_loc.
Then linker collects all such information into a table in the kernel image
(between __start_mcount_loc and __stop_mcount_loc) for later use by ftrace.
This patch adds arm64 specific definitions to identify such locations.
There are two types of implementation, C and Perl. On arm64, only C version
is used to build the kernel now that CONFIG_HAVE_C_RECORDMCOUNT is on.
But Perl version is also maintained.
This patch also contains a workaround just in case where a header file,
elf.h, on host machine doesn't have definitions of EM_AARCH64 nor
R_AARCH64_ABS64. Without them, compiling C version of recordmcount will
fail.
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Do the mcount offset adjustment in the recordmcount.pl/recordmcount.[ch]
at compile time and not in ftrace_call_adjust at run time.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Do the mcount offset adjustment in the recordmcount.pl/recordmcount.[ch]
at compile time and not in ftrace_call_adjust at run time.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
There's some sections that should not have mcount recorded and should not have
modifications to the that code. But currently they waste some time by calling
mcount anyway (which simply returns). As the real answer should be to
either whitelist the section or have gcc ignore it fully.
This change adds a option to recordmcount to warn when it finds a section
that is ignored by ftrace but still contains mcount callers. This is not on
by default as developers may not know if the section should be completely
ignored or added to the whitelist.
Cc: John Reiser <jreiser@bitwagon.com>
Link: http://lkml.kernel.org/r/20110421023738.476989377@goodmis.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
There are sections that are ignored by ftrace for the function tracing because
the text is in a section that can be removed without notice. The mcount calls
in these sections are ignored and ftrace never sees them. The downside of this
is that the functions in these sections still call mcount. Although the mcount
function is defined in assembly simply as a return, this added overhead is
unnecessary.
The solution is to convert these callers into nops at compile time.
A better solution is to add 'notrace' to the section markers, but as new sections
come up all the time, it would be nice that they are delt with when they
are created.
Later patches will deal with finding these sections and doing the proper solution.
Thanks to H. Peter Anvin for giving me the right nops to use for x86.
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: John Reiser <jreiser@bitwagon.com>
Link: http://lkml.kernel.org/r/20110421023738.237101176@goodmis.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The .kprobe.text section is safe to modify mcount to nop and tracing.
Add it to the whitelist in recordmcount.c and recordmcount.pl.
Cc: John Reiser <jreiser@bitwagon.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Link: http://lkml.kernel.org/r/20110421023737.743350547@goodmis.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The Linux style for switch statements is:
switch (var) {
case x:
[...]
break;
}
Not:
switch (var) {
case x: {
[...]
} break;
Cc: John Reiser <jreiser@bitwagon.com>
Link: http://lkml.kernel.org/r/20110421023737.523968644@goodmis.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The Linux ftrace subsystem style for comparing is:
var == 1
var > 0
and not:
1 == var
0 < var
It is considered that Linux developers are smart enough not to do the
if (var = 1)
mistake.
Cc: John Reiser <jreiser@bitwagon.com>
Link: http://lkml.kernel.org/r/20110421023737.290712238@goodmis.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The section .ref.text will not go away unexpectedly and is
safe to trace. Add it to the safe list of sections to allow
tracing.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Depending on the compiler version, ARM GCC calls the mcount function
either __gnu_mcount_nc or mcount.
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Rabin Vincent <rabin@rab.in>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
arch/arm/kernel/ftrace.c references mcount like kernel/tracing/ftrace.c,
so change the exclusion filter to match any ftrace.o.
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Rabin Vincent <rabin@rab.in>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Since MIPS modules' address space differs from the core kernel space, to access
the _mcount in the core kernel, the kernel functions in modules must use long
call (-mlong-calls): load the _mcount address into one register and jump to the
address stored by the register:
c: 3c030000 lui v1,0x0 <--------> b label
c: R_MIPS_HI16 _mcount
c: R_MIPS_NONE *ABS*
c: R_MIPS_NONE *ABS*
10: 64630000 daddiu v1,v1,0
10: R_MIPS_LO16 _mcount
10: R_MIPS_NONE *ABS*
10: R_MIPS_NONE *ABS*
14: 03e0082d move at,ra
18: 0060f809 jalr v1
label:
In the old Perl version of recordmcount, we only need to record the position of
the 1st R_MIPS_HI16 type of _mcount, and later, in ftrace_make_nop(), replace
the instruction in this position by a "b label" and in ftrace_make_call(),
replace it back.
But, the default C version of recordmcount records all of the _mcount symbols,
so, we must filter the 2nd _mcount like the Perl version of recordmcount does.
The C version of recordmcount copes with the symbols before they are linked, So
It doesn't know the type of the symbols and therefore can not filter the
symbols as the Perl version of recordmcount does. But as we can see above, the
2nd _mcount symbols of the long call alawys follows the 1st _mcount symbol of
the same long call, which means the offset from the 1st to the 2nd is fixed, it
is 0x10-0xc = 4 here, 4 is the length of the 1st load instruciton, for MIPS has
fixed length of instructions, this offset is always 4.
And as we know, the _mcount is inserted into the entry of every kernel
function, the offset between the other _mcount's is expected to be always
bigger than 4. So, to filter the 2ns _mcount symbol of the long call, we can
simply check the offset between two _mcount symbols, If it is 4, then, filter
the 2nd _mcount symbol.
To avoid touching too much code, an 'empty' function fn_is_fake_mcount() is
added for all of the archs, and the specific archs can override it via chaning
the function pointer: is_fake_mcount in do_file() with the e_machine. e.g. This
patch adds MIPS_is_fake_mcount() to override the default fn_is_fake_mcount()
pointed by is_fake_mcount.
This fn_is_fake_mcount() checks if the _mcount symbol is fake, e.g. the 2nd
_mcount symbol of the long call is fake, for there are 2 _mcount symbols mapped
to one real mcount call, so, one of them is fake and must be filtered.
This fn_is_fake_mcount() is called in sift_rel_mcount() after finding the
_mcount symbols and before adding the _mcount symbol into mrelp, so, it can
prevent the fake mcount symbol going into the last __mcount_loc table.
Signed-off-by: Wu Zhangjin <wuzhangjin@gmail.com>
LKML-Reference: <b866f0138224340a132d31861fa3f9300dee30ac.1288176026.git.wuzhangjin@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
MIPS64 has 'weird' Elf64_Rel.r_info[1,2], which must be used instead of
the generic Elf64_Rel.r_info, otherwise, the C version of recordmcount
will not work for "segmentation fault".
Usage of "union mips_r_info" and the functions MIPS64_r_sym() and
MIPS64_r_info() written by Maciej W. Rozycki <macro@linux-mips.org>
----
[1] http://techpubs.sgi.com/library/manuals/4000/007-4658-001/pdf/007-4658-001.pdf
[2] arch/mips/include/asm/module.h
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Signed-off-by: John Reiser <jreiser@BitWagon.com>
Signed-off-by: Maciej W. Rozycki <macro@linux-mips.org>
LKML-Reference: <AANLkTinwXjLAYACUfhLYaocHD_vBbiErLN3NjwN8JqSy@mail.gmail.com>
LKML-Reference: <910dc2d5ae1ed042df4f96815fe4a433078d1c2a.1288176026.git.wuzhangjin@gmail.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The file kernel/trace/ftrace.c references the mcount() call to
convert the mcount() callers to nops. But because it references
mcount(), the mcount() address is placed in the relocation table.
The C version of recordmcount reads the relocation table of all
object files, and it will add all references to mcount to the
__mcount_loc table that is used to find the places that call mcount()
and change the call to a nop. When recordmcount finds the mcount reference
in kernel/trace/ftrace.o, it saves that location even though the code
is not a call, but references mcount as data.
On boot up, when all calls are converted to nops, the code has a safety
check to determine what op code it is actually replacing before it
replaces it. If that op code at the address does not match, then
a warning is printed and the function tracer is disabled.
The reference to mcount in ftrace.c, causes this warning to trigger,
since the reference is not a call to mcount(). The ftrace.c file is
not compiled with the -pg flag, so no calls to mcount() should be
expected.
This patch simply makes recordmcount.c skip the kernel/trace/ftrace.c
file. This was the same solution used by the perl version of
recordmcount.
Reported-by: Ingo Molnar <mingo@elte.hu>
Cc: John Reiser <jreiser@bitwagon.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The elf reader for recordmcount.c had duplicate functions for both
32 bit and 64 bit elf handling. This was due to the need of using
the 32 and 64 bit elf structures.
This patch consolidates the two by using macros to define the 32
and 64 bit names in a recordmcount.h file, and then by just defining
a RECORD_MCOUNT_64 macro and including recordmcount.h twice we
create the funtions for both the 32 bit version as well as the
64 bit version using one code source.
Cc: John Reiser <jreiser@bitwagon.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Currently, the mcount callers are found with a perl script that does
an objdump on every file in the kernel. This is a C version of that
same code which should increase the performance time of compiling
the kernel with dynamic ftrace enabled.
Signed-off-by: John Reiser <jreiser@bitwagon.com>
[ Updated the code to include .text.unlikely section as well as
changing the format to follow Linux coding style. ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
