On Tue, Mar 24, 2009 at 16:23, Kim Phillips <kim.phillips@xxxxxxxxxxxxx> wrote: > On Sun, 15 Mar 2009 20:21:47 -0500 > Lee Nipper <lee.nipper@xxxxxxxxx> wrote: > >> Added test for null descriptor returned from current_desc. >> Also removed the diagnostic from current_desc, >> and added one instead in talitos_error >> to report an EU error without finding the descriptor. > > Hi Lee, thanks for this - I'm not clear on how the channel error gets > triggered, points to an EU, and all of this without a referencing a > current descriptor. I don't remember the exact case when I observed it, but I did. In any case, it seemed the code needed the protection against the failed look up case (prevent null pointer use). > How did you run into this case? I had a misconfigured descriptor (I think), and the problem case is not in the code. > Assuming one does > get propagated back, what error status does the callback receive? Ug. If I were still at Freescale and had access to my logs, I could answer with detail, But that's not the case. > afaict, interrupt mitigation should not be intervening in this case. > That's right. The interrupt with error status was occurring, and it was handling it during the isr.. Thanks, Lee -- To unsubscribe from this list: send the line "unsubscribe linux-crypto" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
