search for: noundef

...is then, how can we efficiently encode an assumption that a pointer variable in IR does not have undef bits? Certainly, in the front-end language, (most of) pointers won't have undef bits, and it would be great if the information is still available in IR. A pointer argument can be encoded using noundef, but, e.g., for a pointer that is loaded from memory, such information disappears. I think this information is helpful reducing the cost of fixing existing undef/poison-related optimizations, because we can conclude that we don't need to insert freeze in more cases. Juneyoung On Tue, Sep 22,...

Hi, We're looking at what may be a real-life bug encountered by our compiler related to `undef` values and function calls. The input program effectively contains the expression clamp(v, x, x) expecting that the result will be equal to `x`, even when `v` is read from uninitialized memory. In the input language, `clamp` is a built-in, so this expectation is somewhat reasonable. In our

I think we need to allow this.  Otherwise, we have to prove that addresses are non-undef before we can hoist or sink a memory instruction.  Today, aliasing can use things like known bits, and if we imposed a no-undef in address requirement, we'd either need to replace such reasoning in AA, or have passes which wish to hoist/sink check the property afterwards. Or to say it differently, I

...others): https://reviews.llvm.org/D81678 *2. Preserve no-undef/poison assumption of lvalue reads in C/C++ when valid.* Reading an indeterminate value from an lvalue that does not have char or std::byte type is UB [1]. Since reading an lvalue is lowered to `load` in IR, we suggest attaching a new !noundef metadata to such `load`s. The IR-side change is here: https://reviews.llvm.org/D89050 The clang-side change is going to be made after D81678 is reviewed, because it is likely that this patch will have a lot of changes in clang tests. *Replacing Undef with Poison* --------------------------- Sinc...

...f/poison assumption of lvalue reads in C/C++ when >> valid.* >> >> Reading an indeterminate value from an lvalue that does not have char or >> std::byte type is UB [1]. >> Since reading an lvalue is lowered to `load` in IR, we suggest attaching >> a new >> !noundef metadata to such `load`s. >> The IR-side change is here: https://reviews.llvm.org/D89050 >> The clang-side change is going to be made after D81678 is reviewed, >> because it is likely >> that this patch will have a lot of changes in clang tests. >> >> >> *R...

...t;>> valid.* >>>> >>>> Reading an indeterminate value from an lvalue that does not have char or >>>> std::byte type is UB [1]. >>>> Since reading an lvalue is lowered to `load` in IR, we suggest >>>> attaching a new >>>> !noundef metadata to such `load`s. >>>> The IR-side change is here: https://reviews.llvm.org/D89050 >>>> The clang-side change is going to be made after D81678 is reviewed, >>>> because it is likely >>>> that this patch will have a lot of changes in clang test...

...`check'. make[3]: Leaving directory `/root/rpmbuild/BUILD/dovecot-2.2.29.1/src/plugins/imap-stats' Making check in mail-crypt make[3]: Entering directory `/root/rpmbuild/BUILD/dovecot-2.2.29.1/src/plugins/mail-crypt' for bin in test-mail-global-key test-mail-key; do \ if ! env NOUNDEF=1 ./$bin; then exit 1; fi; \ done try_load_keys ........................................................ : ok test_empty_keyset .................................................... : ok 0 / 2 tests failed generate user key .................................................... : ok generate...

...ing an indeterminate value from an lvalue that does not have char >>>>>> or >>>>>> std::byte type is UB [1]. >>>>>> Since reading an lvalue is lowered to `load` in IR, we suggest >>>>>> attaching a new >>>>>> !noundef metadata to such `load`s. >>>>>> The IR-side change is here: https://reviews.llvm.org/D89050 >>>>>> The clang-side change is going to be made after D81678 is reviewed, >>>>>> because it is likely >>>>>> that this patch will hav...

...e that does not have >>>>>>>> char or >>>>>>>> std::byte type is UB [1]. >>>>>>>> Since reading an lvalue is lowered to `load` in IR, we suggest >>>>>>>> attaching a new >>>>>>>> !noundef metadata to such `load`s. >>>>>>>> The IR-side change is here: https://reviews.llvm.org/D89050 >>>>>>>> The clang-side change is going to be made after D81678 is reviewed, >>>>>>>> because it is likely >>>>>>&...

...Error: Cannot lookup test user: Mail access for users with UID 89 not permitted (see first_valid_uid in config file, uid from mail_uid setting). ../../../run-test.sh: line 27: 29930 Segmentation fault????? (core dumped) valgrind -q --trace-children=yes --leak-check=full --log-file=test.out.$$ $noundef $* ==29930== Invalid read of size 8 ==29930==??? at 0x404699: mail_raw_user_create (mail-raw.c:142) ==29930==??? by 0x40325B: test_init (test-edit-mail.c:105) ==29930==??? by 0x4036DA: test_edit_mail_concatenated (test-edit-mail.c:187) ==29930==??? by 0x524D8B1: test_run_funcs (test-common.c:244) =...

...test user: Mail access for > users with UID 89 not permitted (see first_valid_uid in config file, uid > from mail_uid setting). > ../../../run-test.sh: line 27: 29930 Segmentation fault????? (core > dumped) valgrind -q --trace-children=yes --leak-check=full > --log-file=test.out.$$ $noundef $* > ==29930== Invalid read of size 8 > ==29930==??? at 0x404699: mail_raw_user_create (mail-raw.c:142) > ==29930==??? by 0x40325B: test_init (test-edit-mail.c:105) > ==29930==??? by 0x4036DA: test_edit_mail_concatenated > (test-edit-mail.c:187) > ==29930==??? by 0x524D8B1: test_...

Hello list, During the 'make check' (pigeonhole) portion of a building (rpmbuild) a dovecot.spec file I'm getting the following error: testsuite(root): Fatal: Couldn't drop privileges: User is missing UID (see mail_uid setting) I'm not sure how to resolve this, and any help would be appreciated to help me finish the creating the rpm -- Eric Broch White Horse Technical