Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 18 Jan 2008 16:58:38 GMT
From:      Warner Losh <imp@FreeBSD.org>
To:        Perforce Change Reviews <perforce@freebsd.org>
Subject:   PERFORCE change 133557 for review
Message-ID:  <200801181658.m0IGwc2K093762@repoman.freebsd.org>

next in thread | raw e-mail | index | archive | help
http://perforce.freebsd.org/chv.cgi?CH=133557

Change 133557 by imp@imp_paco-paco on 2008/01/18 16:58:22

	Add a nop to prevent macro expanding into delay slot of blt.  The
	target of the blt also loads the per cpu stuff (into t0), so there's
	no need to put it in the delay slot.  One might be able to tweak
	the code a little to load it always and have the leaf code cope,
	but I'm not sure if the leaf that's called is called only from here
	so I'm just doing the correct thing and noting the posbilty that more
	analysis could yield more optimal results.

Affected files ...

.. //depot/projects/mips2-jnpr/src/sys/mips/mips/support.S#6 edit

Differences ...

==== //depot/projects/mips2-jnpr/src/sys/mips/mips/support.S#6 (text+ko) ====

@@ -1466,6 +1466,7 @@
 
 	li	t0, VM_MAXUSER_ADDRESS /* verify address validity */
 	blt	a0, t0, fusufault		/* trap faults */
+	nop
 
 	GET_CPU_PCPU(t1)
 	lw	t1, PC_CURTHREAD(t1)



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200801181658.m0IGwc2K093762>