Arch Linux User Repository

re sanerb / 2014-02-19 19:29 The error message captured in this comment seems impossible, hence something probably happens with undefined behavior. If you're still suspicious of the earlier memcpy() thing (overlapping regions etc), here's a debug patch, but it doesn't fire on my end. Don't forget to rebuild VfrCompile with make -C "$EDK_TOOLS_PATH". From 4680e08ee2fbb88bb37e4ab8d946ec28acce2b2f Mon Sep 17 00:00:00 2001 From: Laszlo Ersek <lersek@redhat.com> Date: Mon, 17 Mar 2014 19:14:46 +0100 Subject: [PATCH] VfrCompile: debug overlapping memcpy()s Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Laszlo Ersek <lersek@redhat.com> --- BaseTools/Source/C/VfrCompile/GNUmakefile | 2 +- BaseTools/Source/C/VfrCompile/VfrUtilityLib.cpp | 34 +++++++++++++++++++++++++ 2 files changed, 35 insertions(+), 1 deletion(-) diff --git a/BaseTools/Source/C/VfrCompile/GNUmakefile b/BaseTools/Source/C/VfrCompile/GNUmakefile index 82005e1..0d679a7 100644 --- a/BaseTools/Source/C/VfrCompile/GNUmakefile +++ b/BaseTools/Source/C/VfrCompile/GNUmakefile @@ -24,7 +24,7 @@ TOOL_INCLUDE = -I Pccts/h OBJECTS = AParser.o DLexerBase.o ATokenBuffer.o EfiVfrParser.o VfrLexer.o VfrSyntax.o \ VfrFormPkg.o VfrError.o VfrUtilityLib.o VfrCompiler.o -VFR_CPPFLAGS = -DPCCTS_USE_NAMESPACE_STD $(CPPFLAGS) +VFR_CPPFLAGS = -DPCCTS_USE_NAMESPACE_STD $(CPPFLAGS) -g3 LINKER = $(CXX) diff --git a/BaseTools/Source/C/VfrCompile/VfrUtilityLib.cpp b/BaseTools/Source/C/VfrCompile/VfrUtilityLib.cpp index 7c1383e..8d80d9b 100644 --- a/BaseTools/Source/C/VfrCompile/VfrUtilityLib.cpp +++ b/BaseTools/Source/C/VfrCompile/VfrUtilityLib.cpp @@ -18,6 +18,40 @@ WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED. #include "VfrUtilityLib.h" #include "VfrFormPkg.h" +#include <string.h> +#include <stdint.h> +#include <assert.h> +#include <execinfo.h> +#include <unistd.h> + +#define NFRAMES 32 + +void * +memcpy(void *dest, const void *src, size_t n) +{ + uintptr_t s, d; + + s = (uintptr_t)src; + d = (uintptr_t)dest; + assert(n <= (uintptr_t)-1); + assert(s <= (uintptr_t)-1 - n); + assert(d <= (uintptr_t)-1 - n); + + if (s + n > d && d + n > s) { + void *frames[NFRAMES]; + int nframes; + + nframes = backtrace(frames, NFRAMES); + assert(nframes > 0); + assert(nframes < NFRAMES); + + fprintf(stderr, "memcpy(%p, %p, 0x%zx) overlap:\n", dest, src, n); + backtrace_symbols_fd(frames, nframes, STDERR_FILENO); + } + + return memmove(dest, src, n); +} + VOID CVfrBinaryOutput::WriteLine ( IN FILE *pFile, -- 1.8.3.1

haven't tried it yet, but worth a look: https://aur.archlinux.org/packages/freeswitch-git/ comment: The mod_v8 python build error can be mitigated by creating a custom /usr/local/bin/python file as detailed in the Arch wiki https://wiki.archlinux.org/index.php/Python#Dealing_with_version_problem_in_build_scripts Modify this line case "$script" in (/path/to/project1/*|/path/to/project2/*|/path/to/project3*) So it looks something like this case "$script" in (/tmp/*|/home/username/Downloads/freeswitch-git/*) The first path (/tmp/*) to is to make it work with AUR utils such as packer-color. The second path would be to change the active version of python if you wanted to build freeswitch-git using makepkg at path /home/username/Downloads/freeswitch-git/* since it seems to be build.py that's failing, perhaps this is a python2 vs python3 thing.

this error continues to occur even with a fresh building dir and source fetch: "VfrCompile" -l -n --string-db /var/abs/local/yaourtbuild/ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/OUTPUT/BdsDxeStrDefs.hpk --output-directory /var/abs/local/yaourtbuild/ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/DEBUG/DeviceMngr /var/abs/local/yaourtbuild/ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/OUTPUT/DeviceManagerVfr.i VfrCompile: ERROR 0003: Error parsing compile error in file (null) make: *** [/var/abs/local/yaourtbuild/ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/DEBUG/DeviceMngr/DeviceManagerVfr.c] Error 2 GNUmakefile:920: recipe for target '/var/abs/local/yaourtbuild/ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/DEBUG/DeviceMngr/DeviceManagerVfr.c' failed build.py... : error 7000: Failed to execute command make tbuild [/var/abs/local/yaourtbuild/ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe] build.py... : error F002: Failed to build module /var/abs/local/yaourtbuild/ovmf-svn/src/tianocore-edk2-svn_build/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe.inf [X64, GCC48, RELEASE] - Failed - Build end time: 14:21:38, Feb.19 2014 Build total time: 00:01:33 ==> ERROR: A failure occurred in build(). Aborting... ==> ERROR: Makepkg was unable to build ovmf-svn. ==> Restart building ovmf-svn ? [y/N] ==> --------------------------------- ==> i suspect it may be due to upstream as it seems a similar issue happened with a (very much so) earlier version in the same file (DeviceManagerVfr.c): http://feishare.com/efimail/messages/20081205-2046-Re__Edk2_Dev__Trouble_building_the_TOT_tools-_Andrew_J__Fish_.html their bug tracker (via trac) is here: http://sourceforge.net/apps/trac/edk2/report/1?asc=0&sort=created&USER=anonymous i haven't submitted a report, however, as i'm not entirely sure what's going on with this failure. i'd encourage someone a bit more familiar with EDKII to submit one though.

I couldn't get this to build using yaourt or with makepkg. "VfrCompile" -l -n --string-db /var/abs/local/yaourt-tmp/yaourt-tmp-hugg/aur-ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/OUTPUT/BdsDxeStrDefs.hpk --output-directory /var/abs/local/yaourt-tmp/yaourt-tmp-hugg/aur-ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/DEBUG/DeviceMngr /var/abs/local/yaourt-tmp/yaourt-tmp-hugg/aur-ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/OUTPUT/DeviceManagerVfr.i VfrCompile: ERROR 0003: Error parsing compile error in file (null) make: *** [/var/abs/local/yaourt-tmp/yaourt-tmp-hugg/aur-ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/DEBUG/DeviceMngr/DeviceManagerVfr.c] Error 2 GNUmakefile:920: recipe for target '/var/abs/local/yaourt-tmp/yaourt-tmp-hugg/aur-ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe/DEBUG/DeviceMngr/DeviceManagerVfr.c' failed build.py... : error 7000: Failed to execute command make tbuild [/var/abs/local/yaourt-tmp/yaourt-tmp-hugg/aur-ovmf-svn/src/tianocore-edk2-svn_build/Build/OvmfX64/RELEASE_GCC48/X64/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe] build.py... : error F002: Failed to build module /var/abs/local/yaourt-tmp/yaourt-tmp-hugg/aur-ovmf-svn/src/tianocore-edk2-svn_build/IntelFrameworkModulePkg/Universal/BdsDxe/BdsDxe.inf [X64, GCC48, RELEASE] - Failed - Build end time: 22:58:42, Feb.10 2014 Build total time: 00:01:58 ==> ERROR: A failure occurred in build(). Aborting... FAIL: 2

@mazieres: There have been a few changes in UEFI Shell and OVMF recently. The help "-" issue and "map" command issues might be due to the changes. I suggest contacting tianocore devs in edk2-devel ML.

Crashing on start with qemu 1.6 and qemu-git : qemu-system-x86_64 -enable-kvm -net none -m 1024 -bios /usr/share/ovmf/x86_64/bios.bin KVM internal error. Suberror: 1 emulation failure EAX=c0000033 EBX=fffe60e4 ECX=c0000080 EDX=00000000 ESI=fffe62c4 EDI=00005042 EBP=fffe6000 ESP=00000000 EIP=ffffff26 EFL=00000082 [--S----] CPL=0 II=0 A20=1 SMM=0 HLT=0 ES =0008 00000000 ffffffff 00c09300 DPL=0 DS [-WA] CS =0010 00000000 ffffffff 00c09b00 DPL=0 CS32 [-RA] SS =0008 00000000 ffffffff 00c09300 DPL=0 DS [-WA] DS =0008 00000000 ffffffff 00c09300 DPL=0 DS [-WA] FS =0008 00000000 ffffffff 00c09300 DPL=0 DS [-WA] GS =0008 00000000 ffffffff 00c09300 DPL=0 DS [-WA] LDT=0000 00000000 0000ffff 00008200 DPL=0 LDT TR =0000 00000000 0000ffff 00008b00 DPL=0 TSS64-busy GDT= 00000000ffffff80 0000001f IDT= 0000000000000000 0000ffff CR0=c0000033 CR2=0000000000000000 CR3=00000000ffffe000 CR4=00000660 DR0=0000000000000000 DR1=0000000000000000 DR2=0000000000000000 DR3=0000000000000000 DR6=00000000ffff0ff0 DR7=0000000000000400 EFER=0000000000000500 Code=00 c0 0f 32 0f ba e8 08 0f 30 0f 20 c0 0f ba e8 1f 0f 22 c0 <ea> 2d ff ff ff 18 00 e9 93 00 00 00 fa bb 00 f0 8e db bb 6c ff 2e 66 0f 01 17 66 b8 23 00

For some reason, the help summary in the efi shell does not work with this package. If I used ovmf-bin, it works fine. But if I boot this version and get into the efi shell, typing "help" just produces a bunch of blank lines with "-" in the middle. I can still get help on individual topics, just not the list of commands. Very annoying. I suspect this has something to do with display colors or something falling back to black on black. I can say "help * -usage -b" so as to see all the commands, but that's a lot less readable than the default help, so I'm having trouble finding the new commands. In particular, compared to the shell in ovmf-bin, the "map" command does not display all the mappings. For example, I can't figure out how to discover the name of the EFI system partition. With the ovmf-bin shell, the command "map" shows fs0, but such an entry is missing from the output of this shell. I assume there's some other command now, but can't figure out what it is.

@wuischke: 'iasl' added to makedepends.