AOCL cannot be linked in toolchain 202503 when compling ELPA

[QuantumMisaka]

Describe the bug

Using toolchain_aocc-aocl.sh in AMD AOCC-AOCL server, and :

configure: error: could not link with blas: specify path
make: *** No targets specified and no makefile found.  Stop.
make: *** No rule to make target 'install'.  Stop.
ln: failed to create symbolic link '/data/test_admin/abacus-develop/toolchain/install/elpa-2025.06.001/cpu/include/elpa': No such file or directory
ERROR: (./scripts/stage3/install_elpa.sh, line 192) Non-zero exit code detected.

Seems that the lapack in AOCL is not be linked correctly.

Expected behavior

Work well

To Reproduce

In AMD-CPU server of TMC Group:

git clone https://github.com/deepmodeling/abacus-develop
cd abacus-develop/toolchain
module load aocl-5.0.0-aocc aocc-5.0.0
./toolchain_aocc-aocl.sh

Environment

🖥️  System Information:
   ├─ OS: Linux x86_64
   ├─ Kernel: 6.8.0-85-generic
   ├─ glibc: 2.35
   ├─ CPU Cores: 128
   ├─ Memory: 503GB
   └─ GPU: AMD GPU (2 devices)

[Toolchain 202503 can directly give server environment]

Additional Context

The gcc-AOCL toolchain may have same problem ,and I suspect that this problem is directly from install_aocl.sh in stage 2

Task list for Issue attackers (only for developers)

• Verify the issue is not a duplicate.
• Describe the bug.
• Steps to reproduce.
• Expected behavior.
• Error message.
• Environment details.
• Additional context.
• Assign a priority level (low, medium, high, urgent).
• Assign the issue to a team member.
• Label the issue with relevant tags.
• Identify possible related issues.
• Create a unit test or automated test to reproduce the bug (if applicable).
• Fix the bug.
• Test the fix.
• Update documentation (if necessary).
• Close the issue and inform the reporter (if applicable).