Seeing a crash building DAOS with icx in xiar tool on Leap 15.4 with latest kernel

[Jeffrey_O_Intel1]

Here is the error we see.  It happens on every build in our CI with a recent Leap update to this  kernel: kernel-default-5.14.21-150400.24.55.3

[2023-04-03T16:12:19.025Z] icx -o build/dev/icc/src/bio/smd/smd_store.o -c -Werror -g -Wshadow -Wall -fpic -Wno-gnu-designator -Wno-missing-braces -fstack-usage -Wno-ignored-attributes -Wno-gnu-zero-variadic-macro-arguments -Wno-tautological-constant-out-of-range-compare -Wno-unused-command-line-argument -Wframe-larger-than=4096 -fno-strict-overflow -fno-delete-null-pointer-checks -fwrapv -fstack-protector-strong -fstack-clash-protection -O2 -DAPI_VERSION=\"2.7.0\" -DDAOS_VERSION=\"2.3.106\" -DCART_VERSION=\"4.9.0\" -DD_HAS_VALGRIND -D_FORTIFY_SOURCE=2 -DFAULT_INJECTION=1 -DCMOCKA_FILTER_SUPPORTED=0 -D_GNU_SOURCE -DDAOS_PMEM_BUILD -DDAOS_PMEM_BUILD -Isrc/include -Ibuild/dev/icc/src/include -Isrc/gurt -I/opt/daos/prereq/release/mercury/include -I/opt/daos/prereq/release/argobots/include -I/opt/daos/prereq/release/isal/include -I/opt/daos/prereq/release/isal_crypto/include -I/opt/daos/prereq/release/pmdk/include src/bio/smd/smd_store.c
[2023-04-03T16:12:19.025Z] icx -o build/dev/icc/src/bio/smd/smd_pool.o -c -Werror -g -Wshadow -Wall -fpic -Wno-gnu-designator -Wno-missing-braces -fstack-usage -Wno-ignored-attributes -Wno-gnu-zero-variadic-macro-arguments -Wno-tautological-constant-out-of-range-compare -Wno-unused-command-line-argument -Wframe-larger-than=4096 -fno-strict-overflow -fno-delete-null-pointer-checks -fwrapv -fstack-protector-strong -fstack-clash-protection -O2 -DAPI_VERSION=\"2.7.0\" -DDAOS_VERSION=\"2.3.106\" -DCART_VERSION=\"4.9.0\" -DD_HAS_VALGRIND -D_FORTIFY_SOURCE=2 -DFAULT_INJECTION=1 -DCMOCKA_FILTER_SUPPORTED=0 -D_GNU_SOURCE -DDAOS_PMEM_BUILD -DDAOS_PMEM_BUILD -Isrc/include -Ibuild/dev/icc/src/include -Isrc/gurt -I/opt/daos/prereq/release/mercury/include -I/opt/daos/prereq/release/argobots/include -I/opt/daos/prereq/release/isal/include -I/opt/daos/prereq/release/isal_crypto/include -I/opt/daos/prereq/release/pmdk/include src/bio/smd/smd_pool.c
[2023-04-03T16:12:19.025Z] icx -o build/dev/icc/src/bio/smd/smd_device.o -c -Werror -g -Wshadow -Wall -fpic -Wno-gnu-designator -Wno-missing-braces -fstack-usage -Wno-ignored-attributes -Wno-gnu-zero-variadic-macro-arguments -Wno-tautological-constant-out-of-range-compare -Wno-unused-command-line-argument -Wframe-larger-than=4096 -fno-strict-overflow -fno-delete-null-pointer-checks -fwrapv -fstack-protector-strong -fstack-clash-protection -O2 -DAPI_VERSION=\"2.7.0\" -DDAOS_VERSION=\"2.3.106\" -DCART_VERSION=\"4.9.0\" -DD_HAS_VALGRIND -D_FORTIFY_SOURCE=2 -DFAULT_INJECTION=1 -DCMOCKA_FILTER_SUPPORTED=0 -D_GNU_SOURCE -DDAOS_PMEM_BUILD -DDAOS_PMEM_BUILD -Isrc/include -Ibuild/dev/icc/src/include -Isrc/gurt -I/opt/daos/prereq/release/mercury/include -I/opt/daos/prereq/release/argobots/include -I/opt/daos/prereq/release/isal/include -I/opt/daos/prereq/release/isal_crypto/include -I/opt/daos/prereq/release/pmdk/include src/bio/smd/smd_device.c
[2023-04-03T16:12:20.601Z] xiar rc build/dev/icc/src/bio/smd/libsmd.a build/dev/icc/src/bio/smd/smd_device.o build/dev/icc/src/bio/smd/smd_pool.o build/dev/icc/src/bio/smd/smd_store.o
[2023-04-03T16:12:20.601Z] mremap_chunk(): invalid pointer
[2023-04-03T16:12:20.866Z] xiar: error core_dumped: /opt/intel/oneapi/compiler/latest/linux/bin/intel64/xild: core dumped
[2023-04-03T16:12:20.866Z] xiar: warning unknown_signal: unknown signal(3)
[2023-04-03T16:12:20.866Z] xiar: error fatal_error_terminated: Fatal error in /opt/intel/oneapi/compiler/latest/linux/bin/intel64/xild, terminated by unknown

I can provide information on how to reproduce if needed.  Has anyone seen this type of failure before? 

For reference, this is building the master branch of daos which can be found here:
https://github.com/daos-stack/daos.git

[Jeffrey_O_Intel1]

For now, we have worked around it by using ar instead of xiar

[VarshaS_Intel]

Hi,

Thanks for posting in Intel Communities.

Great to know that you already have a workaround! Based on this, and our discussion about creating a priority ticket if needed, I am going ahead and closing this case for now.

Please feel free to contact us further if you need any assistance from our side. 

Thanks & Regards,

Varsha