Re: [PATCH v14 08/17] splice: Do splice read from a file without using ITER_PIPE

[egorenar]

Forgot to mention another related problem we had recently.

We have seen LTP test suite failing in 'sendfile09' test case on 2023-02-15:

 [ 5157.233192] CPU: 0 PID: 2435571 Comm: sendfile09 Tainted: G           OE K  N 6.2.0-20230214.rc8.git112.3ebb0ac55efa.300.fc37.s390x+next #1
 [ 5157.233197] Hardware name: IBM 8561 T01 701 (z/VM 7.3.0)
 [ 5157.233199] Krnl PSW : 0704d00180000000 0000000000000002 (0x2)
 [ 5157.233207]            R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:3 CC:1 PM:0 RI:0 EA:3
 [ 5157.233210] Krnl GPRS: 0000000000000008 0000000000000000 00000000893d1800 00000372088bd040
 [ 5157.233213]            00000372088bd040 000000018c836000 0000000000000000 0000038000a4bb80
 [ 5157.233215]            00000372088bd040 0000000000000000 00000000893d1800 00000372088bd040
 [ 5157.233218]            0000000081182100 00000000893d1800 000000001f691596 0000038000a4b9b0
 [ 5157.233222] Krnl Code:#0000000000000000: 0000       illegal 
              >0000000000000002: 0000        illegal 
               0000000000000004: 0000       illegal 
               0000000000000006: 0000       illegal 
               0000000000000008: 0000       illegal 
               000000000000000a: 0000       illegal 
               000000000000000c: 0000       illegal 
               000000000000000e: 0000       illegal 
 [ 5157.233281] Call Trace:
 [ 5157.233284]  [<0000000000000002>] 0x2 
 [ 5157.233288]  [<000000001f694e26>] filemap_get_pages+0x276/0x3b0 
 [ 5157.233296]  [<000000001f7c7256>] generic_file_buffered_splice_read.constprop.0+0xd6/0x370 
 [ 5157.233301]  [<000000001f7c6aa0>] do_splice_to+0x98/0xc8 
 [ 5157.233304]  [<000000001f7c6eee>] splice_direct_to_actor+0xb6/0x270 
 [ 5157.233306]  [<000000001f7c714a>] do_splice_direct+0xa2/0xd8 
 [ 5157.233309]  [<000000001f77c79c>] do_sendfile+0x39c/0x4e8 
 [ 5157.233314]  [<000000001f77ca00>] __do_sys_sendfile64+0x68/0xc0 
 [ 5157.233317]  [<00000000200b4bb4>] __do_syscall+0x1d4/0x200 
 [ 5157.233322]  [<00000000200c42f2>] system_call+0x82/0xb0 
 [ 5157.233328] Last Breaking-Event-Address:
 [ 5157.233329]  [<000000001f691594>] filemap_read_folio+0x3c/0xd0
 [ 5157.233338] Kernel panic - not syncing: Fatal exception: panic_on_oops

But it was apparently fixed on the next day.

Regards
Alex