build: New configure option --with-libtool-modification.
rsa: Fix decoding of PKCS#1 v1.5 and OAEP padding.
const-time: Use ct_not_memequal, instead. Tested with AVR.
build: Check if arch is VAX or compiler is MSVC.
cipher: Fix ElGamal decryption.
rsa: Use memmov_independently when unpadding.
const-time: Add ct_memmov_cond, fix _gcry_mpih_set_cond.
Use single constant-time memory comparison implementation
const-time: always avoid comparison operator for byte comparison
rsa, elgamal: avoid logical not operator in constant-time code
const-time: prefix global symbols with _gcry_
mpih_set_cond: restore EM leakage mitigation
const-time: ct_memmov_cond: switch to use dual mask approach
mpih-const-time: use global vzero/vone variable
mpiutil: use global vone and vzero
ec-nist: use global vone and vzero
ec-nist: avoid unintentional conditional branch by comparison
mpih_cmp_ui: avoid unintentional conditional branch
ec: avoid unintentional condition branches for 25519, 448 and 256k1
const-time: add functions for generating masks from 0/1 input
mpih-const-time: use constant-time comparisons conditional add/sub/abs
mpih_mod: avoid unintentional conditional branch
mpi: Fix ECC computation on hppa.
random:jent: Fix for jent_rng_is_initialized.
random:jent: Fix build with address sanitizer.
cipher:kyber: No change ABI/API for gcry_kem_genkey.
mpi: Use secure MPI in _gcry_mpi_assign_limb_space.
secmem: Handle HAVE_BROKEN_MLOCK for the case with ASAN.
Merge branch 'master' into LIBGCRYPT-1.11-BRANCH
Merge commit '4876a1a4' into LIBGCRYPT-1.11-BRANCH
cipher:kem:ecc: Support secp256k1 by KEM API.
cipher:kem: Provide each enum constant as macro.
build: Allow build with no Kyber.
build: More changes to allow build with no Kyber.
cipher:ecc: Silence GCC 15 warning.
cipher:rsa: Fix missing initialization in generate_fips.
mpi: Provide the function prototype of __udiv_qrnnd.
Add missing abiversion tag for PowerPC assembly
Add missing machine tags for PowerPC assembly
poly1305-p10le: use '.rodata' section for read-only data
Mark nonstring use cases with __nonstring__ attribute.
random:jent: Fix for jent_rng_is_initialized.
random:jent: Fix build with address sanitizer.
mpi: Fix redefinition of types.
secmem: Handle HAVE_BROKEN_MLOCK for the case with ASAN.
mpi: Introduce mpi_tfr and use it for point_tfr.
t-kem: fix test loop iteration
cipher:aria: Fix compiler error on NetBSD.
Update autogen.sh and gpg-error.m4 from upstream.
Update autogen.sh also for test(1) -o adjustment.
Adjust scripts for test(1) operator -a removal
mceliece6688128f: fix stack overflow crash on win64/wine
mceliece6688128f: fix UBSAN runtime errors
cipher-xts: harden mask generation against branch optimization
rijndael: harden mask generation against branch optimization
mpih-pow: harden condition calculation against branch optimization
kyber: harden mask generation against branch optimization
sntrup761: harden mask generation against branch optimization
sntrup761: use const-time helpers for memory comparison and cond move
mceliece6688128f: harden mask generation against branch optimization
const-time: add 64-bit fast paths for const-time buffer functions
mceliece6688128f: use const-time helper for memory comparison
Add stack burning for PQC algorithms
cipher:kyber: Apply a change from upstream.
mpi: Use secure MPI in _gcry_mpi_assign_limb_space.
fips,cipher: Fix the regression with disabled public-key algo.
Thank you for your report.
This ticket was created because building static-linked shows warnings from glibc for and .
Basically, we can/should ignore the warnings from glibc at link time (for normal use cases), because it is irrelevant.
Fix build with libassuan 2.
I see your point. I am afraid adding causes a larger changes.
Here are changes to fix the behavior:
Re-opened because a regression is reported.
fips,cipher: Fix the regression with disabled public-key algo.
• gniibe renamed
T8045: Stack-based buffer overflow in TPM2 `PKDECRYPT` from
Security (internal) - Stack-based buffer overflow in TPM2 `PKDECRYPT` to
Stack-based buffer overflow in TPM2 `PKDECRYPT`.
On 2026-01-20, I found the message to of:
in may spam folder. It has a 10MB long attachment. That might be one of reasons to be identified as a spam.
Considering the current implementation (tpm2d doesn't support keyinfo like scdaemon), it would be good to check the buffer size.
(If key information is accessible easily, we can check with a specific key.)
• gniibe changed the status of
T7889: libgcrypt: HAVE_BROKEN_MLOCK from
Open to
Testing.
Backports have been done in both (1.10/1.11) branches.
build: Update build-aux and m4 from libgpg-error.
build: Update ksba.m4 from libksba.
build: Update build-aux and m4 from libgpg-error.
build: Update build-aux and m4 from libgpg-error.
build: Update build-aux and m4 from libgpg-error and libassuan.
build: Update build-aux and m4 from libgpg-error and libassuan.
build: Update build-aux and m4 from libgpg-error.
build: Fix typos in texinfo.tex.
build: Apply local change to texinfo.tex.
Fix for header files for sys/time.h and sys/select.h.
build: Update build-aux and m4 from libgpg-error.