* [U-Boot-Users] Possible bug in PPC4xx dcache() function
@ 2007-10-15 16:22 Larry Johnson
0 siblings, 0 replies; only message in thread
From: Larry Johnson @ 2007-10-15 16:22 UTC (permalink / raw)
To: u-boot
Greetings!
If I call the dflush() function in cpu/ppc4xx/start.S on a PPC440EPx as
part of initializing SDRAM, I later get the error
The value of DVLIM special register is incorrect: 0x00000000
from spr_post_test(). I can make the spr_post_test() error message go
away by applying the patch
--- a/cpu/ppc4xx/start.S
+++ b/cpu/ppc4xx/start.S
@@ -1711,6 +1711,7 @@ trap_reloc:
rlwinm r8,r9,0,15,13
rlwinm r8,r8,0,17,15
mtmsr r8
+ mfspr r8,dvlim
addi r3,r0,0x0000
mtspr dvlim,r3
mfspr r3,ivpr
@@ -1725,6 +1726,7 @@ trap_reloc:
..ag: dcbf r0,r3
addi r3,r3,-32
bdnz ..ag
+ mtspr dvlim,r8
sync
mtmsr r9
blr
The routine dflush() zeros DVLIM but does not save and restore the
original value. Is this the intended behavior? The patch I tried seems
safe, although I couldn't easily find the gcc function calling
conventions for PowerPC. (Does anyone have a pointer to these?)
-- Best regards, Larry Johnson <lrj@acm.org>
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2007-10-15 16:22 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-10-15 16:22 [U-Boot-Users] Possible bug in PPC4xx dcache() function Larry Johnson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox