PETSc's test doesn't obey --max-parallel

[hattom]

launched with --max-parallel=16, I'm seeing:

== testing...
  >> running shell command:
        make -j 16 test
        [started at: 2026-02-19 14:14:39]
        [working dir: /tmp/thay/eb/PETSc/3.24.0/lfoss-2025b/petsc-3.24.0]
        [output and state saved to /tmp/eb-j5kk_cq3/run-shell-cmd-output/make-e7d_t1vw]

and

top - 14:26:44 up 227 days,  6:36, 94 users,  load average: 3382.07, 1874.63, 806.53
Tasks: 6231 total,  65 running, 6118 sleeping,  47 stopped,   1 zombie
%Cpu(s): 88.5 us,  8.3 sy,  0.0 ni,  3.2 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
MiB Mem : 772820.2+total, 279959.8+free, 322814.2+used, 183894.3+buff/cache
MiB Swap:    0.000 total,    0.000 free,    0.000 used. 450006.0+avail Mem 

   PID USER      PR  NI    VIRT    RES    SHR S  %CPU  %MEM     TIME+ COMMAND               
156053 thay      20   0  0.125t 473516  12288 R  6505 0.060  18:45.34 ex62                  
157960 thay      20   0  0.125t  65536   8192 R  5450 0.008  10:28.79 ex64                  
157963 thay      20   0  0.125t  56372   8192 R  3405 0.007   9:22.26 ex64                  
154312 thay      20   0  0.125t 677876   8192 R  2732 0.086  22:56.03 ex62                  
154314 thay      20   0  0.125t 681488  12288 R  2286 0.086  11:05.80 ex62                  
156013 thay      20   0  0.125t 851436  16384 R  2277 0.108  11:20.30 ex62                  
157964 thay      20   0  0.125t 590388  12288 R  2132 0.075  13:50.10 ex66                  
157962 thay      20   0  0.125t 684768  12288 R  1995 0.087  11:22.51 ex66                  
174601 thay      20   0  0.125t 251828   4096 R  1609 0.032   0:18.60 ex71                  
157961 thay      20   0  0.125t  69632  12288 R  1600 0.009   6:18.74 ex64                  
174590 thay      20   0  0.125t  69632  12288 R  1318 0.009   0:32.63 ex71                  
154316 thay      20   0  0.125t 784248   8192 R  1268 0.099  19:17.66 ex62                  
174592 thay      20   0  0.125t  65536   8192 R  1241 0.008   0:17.15 ex71                  
157956 thay      20   0  0.125t 630148  12288 R 981.8 0.080   7:51.13 ex66                  
157958 thay      20   0  0.125t 703980  12288 R 972.7 0.089   7:29.73 ex66                  
157957 thay      20   0  0.126t  56152  12288 R 909.1 0.007   8:16.23 ex64                  
174596 thay      20   0  0.125t  69632   8192 R 890.9 0.009   0:29.44 ex71                  
156038 thay      20   0  0.125t 483680  12288 R 495.5 0.061  15:56.44 ex62                  
174600 thay      20   0  0.125t  69632   8192 R 468.2 0.009   0:12.17 ex71                  
174595 thay      20   0  0.125t 405720  12288 R 413.6 0.051   0:18.86 ex71                  
174591 thay      20   0  0.125t  77824  12288 S 345.5 0.010   0:20.36 ex71                  
174587 thay      20   0  0.125t 301204   8192 R 322.7 0.038   0:13.62 ex71                  
156025 thay      20   0  0.125t 465768  12288 R 45.45 0.059  19:22.76 ex62                  
158769 thay      20   0  0.125t 645940   8192 S 27.27 0.082  16:48.65 ex66           

So I have seemingly 24(?) processes and each one is trying to hit "many" threads.
I don't know if the threads are coming from e.g. OpenBLAS, or PETSc itself (unlikely, since I think PETSc doesn't use threads internally).