| Message ID | 20230424143057.27469-1-jgross@suse.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] xen: add support for crash dump analysis with xen.efi | expand |
On 24.04.2023 16:30, Juergen Gross wrote: > --- a/xen/arch/x86/Makefile > +++ b/xen/arch/x86/Makefile > @@ -226,6 +226,9 @@ endif > $(@D)/.$(@F).1r.o $(@D)/.$(@F).1s.o $(orphan-handling-y) $(note_file_option) -o $@ > $(NM) -pa --format=sysv $(@D)/$(@F) \ > | $(objtree)/tools/symbols --all-symbols --xensyms --sysv --sort >$(@D)/$(@F).map > +ifeq ($(CONFIG_DEBUG_INFO),y) > + $(if $(filter --strip-debug,$(EFI_LDFLAGS)),:,$(OBJCOPY) -O elf64-x86-64 $@ $@.elf) This only addresses one of the two earlier raised aspects, as you didn't use what I proposed: $(if $(filter --strip-debug,$(EFI_LDFLAGS)),:)$(OBJCOPY) -O elf64-x86-64 $@ $@.elf Quite possibly because there was a blank missing in there, to separate the colon from $(OBJCOPY). Preferably with the adjustment (which I'd be fine doing while committing, as long as you're okay) Reviewed-by: Jan Beulich <jbeulich@suse.com> To also mention what we have just discussed: Since we're talking about duplicating over 30Mb of data (at least according to my build), an option is going to be to then strip debug info off of xen.efi itself, getting its size into reasonable range again. Jan
On 24.04.23 16:50, Jan Beulich wrote: > On 24.04.2023 16:30, Juergen Gross wrote: >> --- a/xen/arch/x86/Makefile >> +++ b/xen/arch/x86/Makefile >> @@ -226,6 +226,9 @@ endif >> $(@D)/.$(@F).1r.o $(@D)/.$(@F).1s.o $(orphan-handling-y) $(note_file_option) -o $@ >> $(NM) -pa --format=sysv $(@D)/$(@F) \ >> | $(objtree)/tools/symbols --all-symbols --xensyms --sysv --sort >$(@D)/$(@F).map >> +ifeq ($(CONFIG_DEBUG_INFO),y) >> + $(if $(filter --strip-debug,$(EFI_LDFLAGS)),:,$(OBJCOPY) -O elf64-x86-64 $@ $@.elf) > > This only addresses one of the two earlier raised aspects, as you didn't > use what I proposed: > > $(if $(filter --strip-debug,$(EFI_LDFLAGS)),:)$(OBJCOPY) -O elf64-x86-64 $@ $@.elf > > Quite possibly because there was a blank missing in there, to separate > the colon from $(OBJCOPY). Preferably with the adjustment (which I'd > be fine doing while committing, as long as you're okay) Took some time to understand your concern here, but finally I've got it. :-) Yes, I'm fine with this change. > Reviewed-by: Jan Beulich <jbeulich@suse.com> Thanks. > To also mention what we have just discussed: Since we're talking about > duplicating over 30Mb of data (at least according to my build), an > option is going to be to then strip debug info off of xen.efi itself, > getting its size into reasonable range again. Yes, this would shrink it to a little bit above 3MB. Juergen
diff --git a/xen/Makefile b/xen/Makefile index 2710d7327e..5166403cff 100644 --- a/xen/Makefile +++ b/xen/Makefile @@ -502,9 +502,11 @@ _install: $(TARGET)$(CONFIG_XEN_INSTALL_SUFFIX) if [ -r $(TARGET).efi -a -n '$(EFI_DIR)' ]; then \ [ -d $(D)$(EFI_DIR) ] || $(INSTALL_DIR) $(D)$(EFI_DIR); \ $(INSTALL_DATA) $(TARGET).efi $(D)$(EFI_DIR)/$(T)-$(XEN_FULLVERSION).efi; \ - if [ -e $(TARGET).efi.map ]; then \ - $(INSTALL_DATA) $(TARGET).efi.map $(D)$(DEBUG_DIR)/$(T)-$(XEN_FULLVERSION).efi.map; \ - fi; \ + for x in map elf; do \ + if [ -e $(TARGET).efi.$$x ]; then \ + $(INSTALL_DATA) $(TARGET).efi.$$x $(D)$(DEBUG_DIR)/$(T)-$(XEN_FULLVERSION).efi.$$x; \ + fi; \ + done; \ ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T)-$(XEN_VERSION).$(XEN_SUBVERSION).efi; \ ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T)-$(XEN_VERSION).efi; \ ln -sf $(T)-$(XEN_FULLVERSION).efi $(D)$(EFI_DIR)/$(T).efi; \ @@ -539,6 +541,7 @@ _uninstall: rm -f $(D)$(DEBUG_DIR)/$(T)-syms-$(XEN_FULLVERSION).map rm -f $(D)$(EFI_DIR)/$(T)-$(XEN_FULLVERSION).efi rm -f $(D)$(EFI_DIR)/$(T)-$(XEN_VERSION).$(XEN_SUBVERSION).efi + rm -f $(D)$(DEBUG_DIR)/$(T)-$(XEN_FULLVERSION).efi.elf rm -f $(D)$(DEBUG_DIR)/$(T)-$(XEN_FULLVERSION).efi.map rm -f $(D)$(EFI_DIR)/$(T)-$(XEN_VERSION).efi rm -f $(D)$(EFI_DIR)/$(T).efi @@ -569,7 +572,7 @@ _clean: -o -name '*.lex.c' -o -name '*.tab.[ch]' -o -name "*.gcno" \ -o -name ".*.cmd" -o -name "lib.a" \) -exec rm -f {} \; rm -f include/asm $(TARGET) $(TARGET).gz $(TARGET)-syms $(TARGET)-syms.map - rm -f $(TARGET).efi $(TARGET).efi.map $(TARGET).efi.stripped + rm -f $(TARGET).efi $(TARGET).efi.map $(TARGET).efi.elf $(TARGET).efi.stripped rm -f asm-offsets.s arch/*/include/asm/asm-offsets.h rm -f .banner .allconfig.tmp include/xen/compile.h diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile index fc9487aa40..0c66ba9086 100644 --- a/xen/arch/x86/Makefile +++ b/xen/arch/x86/Makefile @@ -226,6 +226,9 @@ endif $(@D)/.$(@F).1r.o $(@D)/.$(@F).1s.o $(orphan-handling-y) $(note_file_option) -o $@ $(NM) -pa --format=sysv $(@D)/$(@F) \ | $(objtree)/tools/symbols --all-symbols --xensyms --sysv --sort >$(@D)/$(@F).map +ifeq ($(CONFIG_DEBUG_INFO),y) + $(if $(filter --strip-debug,$(EFI_LDFLAGS)),:,$(OBJCOPY) -O elf64-x86-64 $@ $@.elf) +endif rm -f $(@D)/.$(@F).[0-9]* $(@D)/..$(@F).[0-9]* ifeq ($(CONFIG_XEN_IBT),y) $(SHELL) $(srctree)/tools/check-endbr.sh $@
Today it is not possible to analyse crash dumps of a system in hypervisor mode when it had been booted via EFI, as the crash utility doesn't understand the file format of xen.efi. This can easily be solved by creating an ELF file from xen.efi via objcopy. Using that file as name list for crash enables the user to analyse the dump in hypervisor mode. Note that crash isn't happy with a file containing no text and data, so using --only-keep-debug is not an option. Signed-off-by: Juergen Gross <jgross@suse.com> --- V2: - drop Kconfig help text changes (Jan Beulich) - apply some Makefile changes (Jan Beulich) - add xen.efi.elf removal to _clean target --- xen/Makefile | 11 +++++++---- xen/arch/x86/Makefile | 3 +++ 2 files changed, 10 insertions(+), 4 deletions(-)