[BSD] LSI MegaRAID U320-2

[Gabor HALASZ]

Helo!

Van egy $subject-em, ami ilyeneket irkal:

Oct  4 13:41:32 120lh kernel: amr0: Too many retries on command 
0xffffff80002ae318.  Controller is likely dead
Oct  4 13:43:02 120lh kernel: amr0: Too many retries on command 
0xffffff80002ad668.  Controller is likely dead
Oct  4 13:43:32 120lh kernel: amr0: Too many retries on command 
0xffffff80002af708.  Controller is likely dead
Oct  4 13:44:02 120lh kernel: amr0: Too many retries on command 
0xffffff80002acbe0.  Controller is likely dead
Oct  4 13:44:32 120lh kernel: amr0: Too many retries on command 
0xffffff80002af280.  Controller is likely dead
Oct  4 13:46:02 120lh kernel: amr0: Too many retries on command 
0xffffff80002abd60.  Controller is likely dead
Oct  4 13:48:02 120lh kernel: amr0: Too many retries on command 
0xffffff80002acdb0.  Controller is likely dead
Oct  4 13:48:32 120lh kernel: amr0: Too many retries on command 
0xffffff80002ae970.  Controller is likely dead
Oct  4 13:49:02 120lh kernel: amr0: Too many retries on command 
0xffffff80002ad920.  Controller is likely dead
Oct  4 13:49:31 120lh kernel: amr0: Too many retries on command 
0xffffff80002af7f0.  Controller is likely dead
Oct  4 13:50:32 120lh kernel: amr0: Too many retries on command 
0xffffff80002abf30.  Controller is likely dead
Oct  4 13:52:03 120lh kernel: amr0: Too many retries on command 
0xffffff80002af708.  Controller is likely dead
Oct  4 13:52:32 120lh kernel: amr0: Too many retries on command 
0xffffff80002ad838.  Controller is likely dead
Oct  4 13:53:02 120lh kernel: amr0: Too many retries on command 
0xffffff80002aed10.  Controller is likely dead
Oct  4 13:54:02 120lh kernel: amr0: Too many retries on command 
0xffffff80002af538.  Controller is likely dead
Oct  4 13:54:32 120lh kernel: amr0: Too many retries on command 
0xffffff80002aea58.  Controller is likely dead
Oct  4 13:56:11 120lh kernel: amr0: Too many retries on command 
0xffffff80002abc78.  Controller is likely dead
Oct  4 14:19:11 120lh kernel: amr0: Too many retries on command 
0xffffff80002af8d8.  Controller is likely dead
Oct  4 14:58:44 120lh kernel: amr0: Too many retries on command 
0xffffff80002aef78.  Controller is likely dead

Ettol meg mukodik, de engem zavar ez a litania. A kartyan firmware 
upgrade volt, az alaplapon nem tudok, mert egy rebrandelt msi e7520-as 
hulladek, amit nem lehet firmware upgradelni, mert az msi updateloje 
hasznalhatatalan (himem.sys-t hallucinal), a rebrandelo csapatnal meg 
nincs updatelo. A kartyat mar probaltam ide-oda rakosgatni a pci-x 
slotokban, es allitgatni a mindenfelet rajta, de csak az uzenetek 
gyakorisaga valtozott.

A fenti uzenetet egy sima komparacio okozza (csak hab a tortan, hogy 
ketszer szerepel ugyanaz az uzenet), nagy vakmeroseg, ha kicsit 
megnovelem a limitet (eloszor persze atirom a hulye hibauzenetet, mert 
igy azt sem tudom, melyik rutin uzeni)?

static int
amr_quartz_submit_command(struct amr_command *ac)
{
...
     if (sc->amr_mailbox->mb_busy) {
         mtx_unlock(&sc->amr_hw_lock);
         if (ac->ac_retries++ > 1000) {
             if (ppsratecheck(&lastfail, &curfail, 1))
                 device_printf(sc->amr_dev, "Too many retries on command 
%p.  "
                               "Controller is likely dead\n", ac);
             ac->ac_retries = 0;
         }
         return (EBUSY);
     }


static int
amr_std_submit_command(struct amr_command *ac)
{
...
     if (AMR_SGET_MBSTAT(sc) & AMR_SMBOX_BUSYFLAG) {
         mtx_unlock(&sc->amr_hw_lock);
         if (ac->ac_retries++ > 1000) {
             if (ppsratecheck(&lastfail, &curfail, 1))
                 device_printf(sc->amr_dev, "Too many retries on command 
%p.  "
                               "Controller is likely dead\n", ac);
             ac->ac_retries = 0;
         }
         return (EBUSY);
     }


-- 
Gabor HALASZ <halasz.g at freemail.hu>