Misc minor fixes #182
Merged
Misc minor fixes #182
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This adds: .gdb_history - gdb command history .C - sometimes the output of gcc -E /.cache/ - clangd cache .cer - x509 certs .esl - efi signature lists compile_commands.json - bear / clangd / etc helper Signed-off-by: Peter Jones <pjones@redhat.com>
vathpela
force-pushed
the
misc-minor-fixes
branch
from
fa7feab
to
bdf28ab
Compare
frozencemetery
requested changes
Nov 9, 2021
There was a problem hiding this comment.
- "make: Temporarily disable gcc -flto": not sure I see how making efivar/efisecdb separate is related to LTO. Just removing the
-fltofrom OPTIMIZE_GCC seems to avoid the issue for me. - "make: remove makeguids dependency from 'make clean'": this seems to have other changes folded in, including some abixml? Possibly I'm misunderstanding the origin of the problem.
All else seems okay to me.
Currently, with binutils 2.37, linking with -flto is broken: > gcc -Og -flto -g3 -Wall -Wextra -Werror -std=gnu11 -funsigned-char -fvisibility=hidden -specs=/home/pjones/devel/github.com/efivar/security/src/include/gcc.specs -fno-merge-constants -L. -Wl,--add-needed -Wl,--build-id -Wl,--no-allow-shlib-undefined -Wl,--no-undefined-version -Wl,-z,now -Wl,-z,muldefs -Wl,-z,relro -Wl,--fatal-warnings -DLIBEFIVAR_VERSION=37 -D_GNU_SOURCE -I/home/pjones/devel/github.com/efivar/security/src/include/ -shared -Wl,-soname,libefivar.so.1 -Wl,--version-script=libefivar.map \ > -o libefivar.so crc32.o dp.o dp-acpi.o dp-hw.o dp-media.o dp-message.o efivarfs.o error.o export.o guid.o guids.o guid-symbols.o lib.o vars.o -ldl > collect2: fatal error: ld terminated with signal 11 [Segmentation fault], core dumped > compilation terminated. > make[1]: *** [/home/pjones/devel/github.com/efivar/security/src/include/rules.mk:32: libefivar.so] Error 1 This commit disables -flto temporarily until the fix[0] is more widely deployed, and forces the compiler to make the .o files for efivar and efisecdb independently. It also switches to specifying .o instead of .c on the command line, because gcc won't allow multiple .c files without -flto, and makes the "% : %.o" rule use $(sort $^) instead of just $^ in order to de-duplicate, so we're not trying to link makeguids.o into makeguids twice. [0] https://sourceware.org/bugzilla/show_bug.cgi?id=28264 Signed-off-by: Peter Jones <pjones@redhat.com>
EFIVAR_OBJECTS was just for consistency, but without doing the MAKEGUIDS_OBJECTS change, I was getting gcc command lines like:
Part of the problem is the way the stuff in guid.c gets used in makeguids, with the horrible EFIVAR_BUILD_ENVIRONMENT hack to make it work. Moving those to being static inlines that are then wrapped by our ABI stubs gets us out of needing that hack there (we still need it to avoid including the missing header file from makeguids). But you're right in that I accidentally merged the memcmp->efi_guid_cmp fix in at the same time; I'll split that out. |
vathpela
force-pushed
the
misc-minor-fixes
branch
2 times, most recently
from
f237c61
to
f4eb757
Compare
This just updates the files to a newer version of abidw. Signed-off-by: Peter Jones <pjones@redhat.com>
For a long time, we've compared guids and their symbolic names with memcmp(). That provides a stable sort order, and very little (if anything) depends on the sorted order of a list of GUIDs for anything other than searching. Unfortunately, when you're actually looking at the sorted output of a list of guids, this winds up being incredibly confusing, because the first element of a guid is a little endian uint23_t, which means memcmp() is evaluating the bytes in a different order than sorting a text file full of guids would. This patch switches to comparing the internal integral members of the guid as unsigned integers, which is still slightly confusing when it comes to the 4th one (it is compared opposite of how it appears, because GUIDs are terrible), and to comparing the symbolic names with strncmp(). Signed-off-by: Peter Jones <pjones@redhat.com>
This will keep makeguids from being built when you run 'make clean', which was annoying. Signed-off-by: Peter Jones <pjones@redhat.com>
Signed-off-by: Peter Jones <pjones@redhat.com>
Signed-off-by: Peter Jones <pjones@redhat.com>
Signed-off-by: Peter Jones <pjones@redhat.com>
Signed-off-by: Peter Jones <pjones@redhat.com>
Signed-off-by: Peter Jones <pjones@redhat.com>
Signed-off-by: Peter Jones <pjones@redhat.com>
Signed-off-by: Peter Jones <pjones@redhat.com>
Signed-off-by: Peter Jones <pjones@redhat.com>
Some of the systems we build in now make us explicitly list gcc and make, so here they are. Signed-off-by: Peter Jones <pjones@redhat.com>
This adds a couple of guids to our libefivar ABI, and also adds a checker so builds will fail if more are incorrectly added. Signed-off-by: Peter Jones <pjones@redhat.com> [rharwood@redhat.com: commit message tweak]
frozencemetery
force-pushed
the
misc-minor-fixes
branch
from
f4eb757
to
b4b14d4
Compare
frozencemetery
approved these changes
Nov 11, 2021
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Here are a bunch of relatively minor fixes.