sbsa_acs driver causes infinite loop in pcie/test_p005.c

[ghost]

The execution of sbsa fails because the code in the sbsa-acs driver file test_p005.c causes the following error in our testing and apparently enters an infinite loop in the driver, likely due to a unexpected configuration of SATA controllers.

      *** Starting PCIe tests ***   
[   71.368011] WARNING: CPU: 0 PID: 2477 at /usr/src/kernel/mm/page_alloc.c:4065 __alloc_pages_nodemask+0x2d0/0xe38 
[   71.369638] Modules linked in: sbsa_acs crc32_ce crct10dif_ce sm3_ce sm3_generic sha3_ce sha3_generic sha512_ce sha512_arm64 ipv6 
[   71.370616] CPU: 0 PID: 2477 Comm: sbsa Not tainted 4.18.0-luv #1 
[   71.370842] Hardware name: XXX 
[   71.371342] pstate: XXX
[   71.371521] pc : __alloc_pages_nodemask+0x2d0/0xe38 
[   71.371693] lr : __alloc_pages_nodemask+0x104/0xe38 
[   71.371858] sp : ffff00001026f850 
[   71.371987] x29: ffff00001026f850 x28: 0000000000000000  
[   71.372231] x27: ffff000001691bb0 x26: 00000000006040c0  
[   71.372411] x25: ffff000001691000 x24: 000000000000000e  
[   71.372592] x23: 0000000020800000 x22: 0000000000000000  
[   71.372762] x21: 0000000000000000 x20: 0000000000000001  
[   71.372946] x19: 0000000000000000 x18: ffffffffffffffff  
[   71.373142] x17: 0000ffff99a2e098 x16: ffff0000082f7a48  
[   71.373323] x15: ffff0000093f36c8 x14: 0000000080400009  
[   71.373503] x13: ffff000001682a34 x12: ffff003031303030  
[   71.373683] x11: ffff00000914d000 x10: 7365726464612067  
[   71.373884] x9 : 0000000000000000 x8 : 0000000000000000  
[   71.374060] x7 : 0000000000000000 x6 : 0000000000000000  
[   71.374241] x5 : 0000000000008aae x4 : 0000000000000001  
[   71.374422] x3 : 0000000000000000 x2 : 7a95d4ab2e107c00  
[   71.374605] x1 : 0000000000400000 x0 : ffff0000006040c0  
[   71.374867] Call trace: 
[   71.375041]  __alloc_pages_nodemask+0x2d0/0xe38 
[   71.375212]  alloc_pages_current+0x88/0xf0 
[   71.375369]  kmalloc_order_trace+0x38/0x140 
[   71.375524]  __kmalloc+0x2fc/0x330 
[   71.376085]  pal_print+0x6c/0x140 [sbsa_acs] 
[   71.376251]  val_print+0x40/0x50 [sbsa_acs] 
[   71.376410]  payload+0x1ac/0x358 [sbsa_acs] 
[   71.376567]  val_run_test_payload+0x3c/0xa0 [sbsa_acs] 
[   71.376750]  p005_entry+0x54/0x90 [sbsa_acs] 
[   71.376913]  val_pcie_execute_tests+0x84/0x278 [sbsa_acs] 
[   71.377107]  val_glue_execute_command+0x17c/0x2d8 [sbsa_acs] 
[   71.377316]  sbsa_proc_write+0x80/0xf0 [sbsa_acs] 
[   71.377473]  proc_reg_write+0x7c/0xc0 
[   71.377599]  __vfs_write+0x60/0x178 
[   71.377719]  vfs_write+0xac/0x1b8 
[   71.377839]  ksys_write+0x6c/0xd8 
[   71.377990]  sys_write+0x34/0x48 
[   71.378107]  __sys_trace_return+0x0/0x4 
[   71.378324] ---[ end trace 02b7f80878c00683 ]---

[gowthamsiddarthd]

Hi @bcran,

Thank you for reporting this issue. The code changes for this has been merged with the commit https://gitlab.arm.com/linux-arm/linux-acs/-/commit/ee76f93cc5a6739ae1dcf2eaf834fdebb1fbe794

Regards, Gowtham Siddarth, Enterprise ACS support team

[edhay]

Closing the issue as per the solution above.