[valgrind] [Bug 471036] disInstr_AMD64: disInstr miscalculated next %rip on RORX imm8, m32/64, r32/64
https://bugs.kde.org/show_bug.cgi?id=471036 Matthias Schwarzott changed: What|Removed |Added Attachment #165688|0 |1 is obsolete|| --- Comment #5 from Matthias Schwarzott --- Created attachment 165691 --> https://bugs.kde.org/attachment.cgi?id=165691=edit Show more context (rip, instr bytes, vex-disasm trace) for this panic case Even more context for the error case (idea how to rerun disInstr_AMD64_WRK taken from the expect_CAS case below): current %rip = 0x109ABE assumed next %rip = 0x109AC7 actual next %rip = 0x109AC8 instruction bytes: 0xC4 0xE3 0xFB 0xF0 0xD 0x60 0x45 0x0 0x0 0x43 0x109ABE: rorx 67,17760(%rip),%rcx vex: the `impossible' happened: disInstr_AMD64: disInstr miscalculated next %rip -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 471036] disInstr_AMD64: disInstr miscalculated next %rip on RORX imm8, m32/64, r32/64
https://bugs.kde.org/show_bug.cgi?id=471036 --- Comment #4 from Matthias Schwarzott --- I tested the attachment "Example patch for guest_amd64_toIR" https://bugs.kde.org/attachment.cgi?id=159662 It perfectly fixes the problem. Without fix (but extended context-printing) the extended bmi testcase fails like this: shrx32 -> current %rip = 0x109ABE assumed next %rip = 0x109AC7 actual next %rip = 0x109AC8 instruction bytes: 0xC4 0xE3 0xFB 0xF0 0xD 0x60 0x45 0x0 0x0 0x43 vex: the `impossible' happened: disInstr_AMD64: disInstr miscalculated next %rip vex storage: T total 171126616 bytes allocated vex storage: P total 512 bytes allocated valgrind: the 'impossible' happened: LibVEX called failure_exit(). host stacktrace: ==20396==at 0x5804383A: show_sched_status_wrk (m_libcassert.c:407) ==20396==by 0x58043957: report_and_quit (m_libcassert.c:478) ==20396==by 0x58043BAB: panic (m_libcassert.c:554) ==20396==by 0x58043BAB: vgPlain_core_panic_at (m_libcassert.c:559) ==20396==by 0x58043BCA: vgPlain_core_panic (m_libcassert.c:564) ==20396==by 0x58058034: failure_exit (m_translate.c:761) ==20396==by 0x5813068A: vpanic (main_util.c:253) ==20396==by 0x581BBDDD: disInstr_AMD64 (guest_amd64_toIR.c:32714) ==20396==by 0x58148E76: disassemble_basic_block_till_stop.constprop.0 (guest_generic_bb_to_IR.c:956) ==20396==by 0x5814965C: bb_to_IR (guest_generic_bb_to_IR.c:1365) ==20396==by 0x5812D6AF: LibVEX_FrontEnd (main_main.c:583) ==20396==by 0x5812E00C: LibVEX_Translate (main_main.c:1235) ==20396==by 0x5805A791: vgPlain_translate (m_translate.c:1831) ==20396==by 0x58097F3B: handle_chain_me (scheduler.c:1164) ==20396==by 0x5809A42B: vgPlain_scheduler (scheduler.c:1531) ==20396==by 0x580E5569: thread_wrapper (syswrap-linux.c:102) ==20396==by 0x580E5569: run_a_thread_NORETURN (syswrap-linux.c:155) sched status: running_tid=1 Thread 1: status = VgTs_Runnable (lwpid 20396) ==20396==at 0x1099FC: do_rorx64 (bmi.c:379) ==20396==by 0x10AEDC: main (bmi.c:1012) client stack range: [0x1FFEFFD000 0x1FFF000FFF] client SP: 0x1FFEFFF258 valgrind stack range: [0x1002DEB000 0x1002EEAFFF] top usage: 10960 of 1048576 -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 471036] disInstr_AMD64: disInstr miscalculated next %rip on RORX imm8, m32/64, r32/64
https://bugs.kde.org/show_bug.cgi?id=471036 --- Comment #3 from Matthias Schwarzott --- Created attachment 165690 --> https://bugs.kde.org/attachment.cgi?id=165690=edit Add rorx rip-relative testcase to bmi.c Adding these instructions to the bmi testcase: 1abe: c4 e3 fb f0 0d 60 45rorx $0x43,0x4560(%rip),%rcx# 6028 1ac5: 00 00 43 1bf9: c4 e3 7b f0 0d 2d 44rorx $0x43,0x442d(%rip),%ecx# 6030 1c00: 00 00 43 -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 471036] disInstr_AMD64: disInstr miscalculated next %rip on RORX imm8, m32/64, r32/64
https://bugs.kde.org/show_bug.cgi?id=471036 Matthias Schwarzott changed: What|Removed |Added CC||z...@gentoo.org --- Comment #2 from Matthias Schwarzott --- Created attachment 165688 --> https://bugs.kde.org/attachment.cgi?id=165688=edit Show more context (rip and instr bytes) for this panic case -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 79362] Debug info is lost for .so files when they are dlclose'd
https://bugs.kde.org/show_bug.cgi?id=79362 Matthias Schwarzott <z...@gentoo.org> changed: What|Removed |Added CC||z...@gentoo.org -- You are receiving this mail because: You are watching all bug changes.
[k3b] [Bug 382941] Segfault from getSupportedWriteSpeedsVia2A
https://bugs.kde.org/show_bug.cgi?id=382941
--- Comment #14 from Matthias Schwarzott <z...@gentoo.org> ---
(In reply to Leslie Zhai from comment #12)
> Git commit 16576d6fd33fd4fd81f66904d29a4478a91a561f by Leslie Zhai.
> Committed on 14/08/2017 at 03:35.
> Pushed by lesliezhai into branch 'master'.
>
> Revert from{2,4}Byte.
>
> Hi Thomas:
> Please help me! I have no idea why num_wr_speed_des (numDesc) might
> be wrong? and how to make it correct if bigger than (32 - 8) / 4,
> and where the *magic numbers* about 32 and 8 come from.
>
> Testcase by Matthias Schwarzott!
>
> CCMAIL: scdbac...@gmx.net
>
> M +11 -8libk3bdevice/k3bdevice.cpp
> M +2-2libk3bdevice/k3bdeviceglobals.cpp
> M +19 -5tests/k3bdeviceglobalstest.cpp
>
> https://commits.kde.org/k3b/16576d6fd33fd4fd81f66904d29a4478a91a561f
This testcase should be deleted (for from2Byte and from4Byte):
+const unsigned char buf0[] = { '\0' };
+QCOMPARE(K3b::Device::from4Byte(buf0), (quint32)0);
It reads data after the array, so will observe undefined behaviour.
--
You are receiving this mail because:
You are watching all bug changes.
[k3b] [Bug 382941] Segfault from getSupportedWriteSpeedsVia2A
https://bugs.kde.org/show_bug.cgi?id=382941
Matthias Schwarzott <z...@gentoo.org> changed:
What|Removed |Added
CC||z...@gentoo.org
--- Comment #7 from Matthias Schwarzott <z...@gentoo.org> ---
Commit 7f0be6a33b8260f7789c6aeed58be8d1c844229a seems to have broken detection
of the burn device / existing empty media to burn on.
I get a lot of lines printing "Invalid Byte!" the other errors:
[...]
Invalid Byte!
(K3b::Device::Device) "/dev/sr0" : Missing modepage 0x05 data.
(K3b::Device::Device) "/dev/sr0" : Cannot check write modes.
[...]
Looking at the code:
from{2,4}Byte seems to be used to read 16/32bit big endian integer values from
raw data.
So calling strlen is just wrong. This leads to rejecting valid byte-streams
containing zero bytes.
Second: I suggest to improve the unit test with a test parsing good 16 and
32bit values (at least one with all values != 0 and one where 0 values are
part).
e.g. for from2Byte
+unsigned const char buf0[] = { 0x00, 0x00 };
+QCOMPARE(K3b::Device::from2Byte(buf0), (quint16)0x);
+unsigned const char buf1[] = { 0x00, 0x70 };
+QCOMPARE(K3b::Device::from2Byte(buf1), (quint16)0x0070);
+unsigned const char buf2[] = { 0x05, 0x00 };
+QCOMPARE(K3b::Device::from2Byte(buf2), (quint16)0x0500);
+unsigned const char buf3[] = { 0xF0, 0x03 };
+QCOMPARE(K3b::Device::from2Byte(buf2), (quint16)0xF003);
the same should be done for from4Byte
+unsigned const char buf0[] = { 0x00, 0x00, 0x00, 0x00 };
+QCOMPARE(K3b::Device::from4Byte(buf0), (quint32)0x);
+unsigned const char buf0[] = { 0x00, 0x00, 0x00, 0x01 };
+QCOMPARE(K3b::Device::from4Byte(buf0), (quint32)0x0001);
+unsigned const char buf0[] = { 0x12, 0x34, 0x56, 0x78 };
+QCOMPARE(K3b::Device::from4Byte(buf0), (quint32)0x12345678);
--
You are receiving this mail because:
You are watching all bug changes.
[valgrind] [Bug 380200] xtree generated callgrind files refer to files without directory name
https://bugs.kde.org/show_bug.cgi?id=380200 --- Comment #3 from Matthias Schwarzott <z...@gentoo.org> --- (In reply to Philippe Waroquiers from comment #2) > (In reply to Matthias Schwarzott from comment #1) > > Created attachment 105714 [details] > > valgrind-xtree-add-regtests.patch > > > > This attachment adds two experimental tests for xtree-leak and xtree-memory. > > xtree-leak uses callgrind_annotate to show the cost values per function and > > sorts the output afterwards to make it stable. > Why is the sort needed ? Do you see variations between runs ? > Or variations between different systems/platforms ? > I run it under amd64. There are variations between runs. The order of functions differs. I have not yet checked if the xtree files are stable or not. -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 380200] xtree generated callgrind files refer to files without directory name
https://bugs.kde.org/show_bug.cgi?id=380200 --- Comment #1 from Matthias Schwarzott <z...@gentoo.org> --- Created attachment 105714 --> https://bugs.kde.org/attachment.cgi?id=105714=edit valgrind-xtree-add-regtests.patch This attachment adds two experimental tests for xtree-leak and xtree-memory. xtree-leak uses callgrind_annotate to show the cost values per function and sorts the output afterwards to make it stable. The other for xtree-memory only checks the generated file. I wonder if this one is stable enough. -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 380200] New: xtree generated callgrind files refer to files without directory name
https://bugs.kde.org/show_bug.cgi?id=380200 Bug ID: 380200 Summary: xtree generated callgrind files refer to files without directory name Product: valgrind Version: 3.13 SVN Platform: Other OS: Linux Status: UNCONFIRMED Severity: normal Priority: NOR Component: general Assignee: jsew...@acm.org Reporter: z...@gentoo.org Target Milestone: --- Created attachment 105712 --> https://bugs.kde.org/attachment.cgi?id=105712=edit valgrind-xtree-add-directory-to-filename The files generated by the new xtree code in callgrind format contain only filenames without directory. The attached patch fixes it. -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 358697] valgrind.h: Some code remains even when defining NVALGRIND
https://bugs.kde.org/show_bug.cgi?id=358697 --- Comment #5 from Matthias Schwarzott <z...@gentoo.org> --- Patch valgrind-3.12.0-valgrind_printf-simple-nvalgrind-fix-plus-testcase.patch has been tested with MSVC 2008 configured to warning level 4. -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 358697] valgrind.h: Some code remains even when defining NVALGRIND
https://bugs.kde.org/show_bug.cgi?id=358697 --- Comment #4 from Matthias Schwarzott <z...@gentoo.org> --- Created attachment 104809 --> https://bugs.kde.org/attachment.cgi?id=104809=edit valgrind-3.12.0-valgrind_printf-simple-nvalgrind-fix-plus-testcase.patch This new patch fixes the problem in the simplest way: "(void)format;" Additionally it adds a testcase to compile and run the existing vgprintf.c with NVALGRIND -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 358697] valgrind.h: Some code remains even when defining NVALGRIND
https://bugs.kde.org/show_bug.cgi?id=358697 --- Comment #3 from Matthias Schwarzott <z...@gentoo.org> --- e.g. QT has a macro like this: #define Q_UNUSED(x) (void)x; To be used inside functions: Q_UNUSED(format) -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 358697] valgrind.h: Some code remains even when defining NVALGRIND
https://bugs.kde.org/show_bug.cgi?id=358697 --- Comment #2 from Matthias Schwarzott <z...@gentoo.org> --- The simplest solution could be to use "(void)format" and protect this with an ifdef checking that we are not running under the problematic static code checker. -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 358697] valgrind.h: Some code remains even when defining NVALGRIND
https://bugs.kde.org/show_bug.cgi?id=358697 --- Comment #1 from Matthias Schwarzott <z...@gentoo.org> --- Created attachment 96891 --> https://bugs.kde.org/attachment.cgi?id=96891=edit valgrind-improve-unused-parameter-on-r15763.patch This patch implements the __attribute__ usage. But I am not sure about the ifdef code. It is just what was originally around the __attribute__ usage. -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 356817] New: valgrind.h triggers compiler errors on MSVC when defining NVALGRIND
https://bugs.kde.org/show_bug.cgi?id=356817 Bug ID: 356817 Summary: valgrind.h triggers compiler errors on MSVC when defining NVALGRIND Product: valgrind Version: unspecified Platform: Compiled Sources OS: MS Windows Status: UNCONFIRMED Severity: normal Priority: NOR Component: general Assignee: jsew...@acm.org Reporter: z...@gentoo.org When compiling sources on MSVC with NVALGRIND being defined, it shows this error: 1>include\Valgrind/valgrind.h(4493) : error C2220: warning treated as error - no 'object' file generated 1>include\Valgrind/valgrind.h(4493) : warning C4100: 'format' : unreferenced formal parameter 1>include\Valgrind/valgrind.h(4531) : warning C4100: 'format' : unreferenced formal parameter Reproducible: Always -- You are receiving this mail because: You are watching all bug changes.
