Skip to contentSkip to conversation

target/hppa: Fix boot with old Linux installation CDs

PATCH In review Helge Deller 2 participants 2 messages Aug 30, 2020

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 target/hppa: Fix boot with old Linux installation CDs

The current qemu hppa emulation emulates a PA1.1 CPU, which can only execute the 32-bit instruction set. For unknown 64-bit instructions, a instruction trap is sent to the virtual CPU.
This behaviour is correct in the sense that we emulate what the PA1.1 specification says.

But when trying to boot older Linux installation images, e.g. ftp://parisc.parisc-linux.org/debian-cd/debian-5.0/lenny-5.0.10-hppa-iso-cd/cdimage.debian.org/debian-5010-hppa-netinst.iso one finds that qemu fails to boot those images.
This can be verified with:
qemu-system-hppa -cdrom debian-5010-hppa-netinst.iso -nographic

The problem is, that in the Linux kernel (e.g. 2.6.26) of those old images 64-bit instructions were used by mistake in the fault handlers. The relevant instructions (the ",*" indicates that it's a 64-bit instruction) I see are:

   0:   09 3e 04 29     sub,* sp,r9,r9
   0:   08 3d 06 3d     add,* ret1,r1,ret1
   0:   0a 09 02 61     or,* r9,r16,r1
   0:   0a ba 00 3a     andcm,* r26,r21,r26
   0:   08 33 02 33     and,* r19,r1,r19

The interesting part is, that real physical 32-bit machines (like the 700/64 and B160L - which is the one we emulate) do boot those images and thus seem to simply ignore the 64-bit flag on those instructions.

The patch below modifies the qemu instruction decoder to ignore the 64-bit flag too - which is what real 32-bit hardware seems to do as well. With this modification qemu now successfully boots those older images too.

I suggest to apply the patch below - even if it does not reflect what the SPEC says. Instead it increases the compatibility to really existing hardware and seem to not create problems if we add real PA2.0 support anytime later.

target/hppa/insns.decode +5 -5
index f0dd71dd08..dceaad65e9 100644
@@ -149,9 +149,9 @@ lci             000001 ----- ----- -- 01001100 0 t:5
 # Arith/Log
 ####
 
-andcm           000010 ..... ..... .... 000000 0 .....  @rrr_cf
-and             000010 ..... ..... .... 001000 0 .....  @rrr_cf
-or              000010 ..... ..... .... 001001 0 .....  @rrr_cf
+andcm           000010 ..... ..... .... 000000 - .....  @rrr_cf
+and             000010 ..... ..... .... 001000 - .....  @rrr_cf
+or              000010 ..... ..... .... 001001 - .....  @rrr_cf
 xor             000010 ..... ..... .... 001010 0 .....  @rrr_cf
 uxor            000010 ..... ..... .... 001110 0 .....  @rrr_cf
 ds              000010 ..... ..... .... 010001 0 .....  @rrr_cf
@@ -161,13 +161,13 @@ uaddcm_tc       000010 ..... ..... .... 100111 0 .....  @rrr_cf
 dcor            000010 ..... 00000 .... 101110 0 .....  @rr_cf
 dcor_i          000010 ..... 00000 .... 101111 0 .....  @rr_cf
 
-add             000010 ..... ..... .... 0110.. 0 .....  @rrr_cf_sh
+add             000010 ..... ..... .... 0110.. - .....  @rrr_cf_sh
 add_l           000010 ..... ..... .... 1010.. 0 .....  @rrr_cf_sh
 add_tsv         000010 ..... ..... .... 1110.. 0 .....  @rrr_cf_sh
 add_c           000010 ..... ..... .... 011100 0 .....  @rrr_cf_sh0
 add_c_tsv       000010 ..... ..... .... 111100 0 .....  @rrr_cf_sh0
 
-sub             000010 ..... ..... .... 010000 0 .....  @rrr_cf
+sub             000010 ..... ..... .... 010000 - .....  @rrr_cf
 sub_tsv         000010 ..... ..... .... 110000 0 .....  @rrr_cf
 sub_tc          000010 ..... ..... .... 010011 0 .....  @rrr_cf
 sub_tsv_tc      000010 ..... ..... .... 110011 0 .....  @rrr_cf
Reviewed-by on v1 1/1
#1 lore raw
Quoted 11 lines · Helge Deller

The patch below modifies the qemu instruction decoder to ignore the 64-bit flag too - which is what real 32-bit hardware seems to do as well. With this modification qemu now successfully boots those older images too.

I suggest to apply the patch below - even if it does not reflect what the SPEC says. Instead it increases the compatibility to really existing hardware and seem to not create problems if we add real PA2.0 support anytime later.

r~