* [PATCH] x86/ioemul: Misc improvements to ioport_emulate.c
@ 2018-01-26 14:48 Andrew Cooper
2018-01-29 11:23 ` Jan Beulich
0 siblings, 1 reply; 3+ messages in thread
From: Andrew Cooper @ 2018-01-26 14:48 UTC (permalink / raw)
To: Xen-devel; +Cc: Andrew Cooper, Jan Beulich
Put the opcode into an array and use memcpy. This allows the compiled code to
be written with two movs, rather than 10 mov $imm8's. Also, drop trailing
whitespace in the file.
Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
---
CC: Jan Beulich <JBeulich@suse.com>
---
xen/arch/x86/ioport_emulate.c | 34 +++++++++++++---------------------
1 file changed, 13 insertions(+), 21 deletions(-)
diff --git a/xen/arch/x86/ioport_emulate.c b/xen/arch/x86/ioport_emulate.c
index e80993a..c2aded7 100644
--- a/xen/arch/x86/ioport_emulate.c
+++ b/xen/arch/x86/ioport_emulate.c
@@ -1,6 +1,6 @@
/******************************************************************************
* ioport_emulate.c
- *
+ *
* Handle I/O port access quirks of various platforms.
*/
@@ -11,32 +11,24 @@
static bool ioemul_handle_proliant_quirk(
u8 opcode, char *io_emul_stub, struct cpu_user_regs *regs)
{
+ static const char stub[] = {
+ 0x9c, /* pushf */
+ 0xfa, /* cli */
+ 0xee, /* out %al, %dx */
+ 0xec, /* 1: in %dx, %al */
+ 0xa8, 0x80, /* test $0x80, %al */
+ 0x75, 0xfb, /* jnz 1b */
+ 0x9d, /* popf */
+ 0xc3, /* ret */
+ };
uint16_t port = regs->dx;
uint8_t value = regs->al;
if ( (opcode != 0xee) || (port != 0xcd4) || !(value & 0x80) )
return false;
- /* pushf */
- io_emul_stub[0] = 0x9c;
- /* cli */
- io_emul_stub[1] = 0xfa;
- /* out %al,%dx */
- io_emul_stub[2] = 0xee;
- /* 1: in %dx,%al */
- io_emul_stub[3] = 0xec;
- /* test $0x80,%al */
- io_emul_stub[4] = 0xa8;
- io_emul_stub[5] = 0x80;
- /* jnz 1b */
- io_emul_stub[6] = 0x75;
- io_emul_stub[7] = 0xfb;
- /* popf */
- io_emul_stub[8] = 0x9d;
- /* ret */
- io_emul_stub[9] = 0xc3;
-
- BUILD_BUG_ON(IOEMUL_QUIRK_STUB_BYTES < 10);
+ memcpy(io_emul_stub, stub, sizeof(stub));
+ BUILD_BUG_ON(IOEMUL_QUIRK_STUB_BYTES < sizeof(stub));
return true;
}
--
2.1.4
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] x86/ioemul: Misc improvements to ioport_emulate.c
2018-01-26 14:48 [PATCH] x86/ioemul: Misc improvements to ioport_emulate.c Andrew Cooper
@ 2018-01-29 11:23 ` Jan Beulich
2018-01-29 11:37 ` Andrew Cooper
0 siblings, 1 reply; 3+ messages in thread
From: Jan Beulich @ 2018-01-29 11:23 UTC (permalink / raw)
To: Andrew Cooper; +Cc: Xen-devel
>>> On 26.01.18 at 15:48, <andrew.cooper3@citrix.com> wrote:
> Put the opcode into an array and use memcpy. This allows the compiled code to
> be written with two movs, rather than 10 mov $imm8's. Also, drop trailing
> whitespace in the file.
>
> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
It's a clear improvement, so
Reviewed-by: Jan Beulich <jbeulich@suse.com>
but ...
> @@ -11,32 +11,24 @@
> static bool ioemul_handle_proliant_quirk(
> u8 opcode, char *io_emul_stub, struct cpu_user_regs *regs)
> {
> + static const char stub[] = {
> + 0x9c, /* pushf */
> + 0xfa, /* cli */
> + 0xee, /* out %al, %dx */
> + 0xec, /* 1: in %dx, %al */
> + 0xa8, 0x80, /* test $0x80, %al */
> + 0x75, 0xfb, /* jnz 1b */
> + 0x9d, /* popf */
> + 0xc3, /* ret */
> + };
... could you go even further and avoid this hex encoding of insns
altogether, but using a file scope asm(), or a function scope one
putting this into e.g. .fixup via .pushsection/.popsection?
Jan
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] x86/ioemul: Misc improvements to ioport_emulate.c
2018-01-29 11:23 ` Jan Beulich
@ 2018-01-29 11:37 ` Andrew Cooper
0 siblings, 0 replies; 3+ messages in thread
From: Andrew Cooper @ 2018-01-29 11:37 UTC (permalink / raw)
To: Jan Beulich; +Cc: Xen-devel
On 29/01/18 11:23, Jan Beulich wrote:
>>>> On 26.01.18 at 15:48, <andrew.cooper3@citrix.com> wrote:
>> Put the opcode into an array and use memcpy. This allows the compiled code to
>> be written with two movs, rather than 10 mov $imm8's. Also, drop trailing
>> whitespace in the file.
>>
>> Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com>
> It's a clear improvement, so
> Reviewed-by: Jan Beulich <jbeulich@suse.com>
> but ...
>
>> @@ -11,32 +11,24 @@
>> static bool ioemul_handle_proliant_quirk(
>> u8 opcode, char *io_emul_stub, struct cpu_user_regs *regs)
>> {
>> + static const char stub[] = {
>> + 0x9c, /* pushf */
>> + 0xfa, /* cli */
>> + 0xee, /* out %al, %dx */
>> + 0xec, /* 1: in %dx, %al */
>> + 0xa8, 0x80, /* test $0x80, %al */
>> + 0x75, 0xfb, /* jnz 1b */
>> + 0x9d, /* popf */
>> + 0xc3, /* ret */
>> + };
> ... could you go even further and avoid this hex encoding of insns
> altogether, but using a file scope asm(), or a function scope one
> putting this into e.g. .fixup via .pushsection/.popsection?
I tried that, but couldn't find a way of creating the blob as a static
in C's eyes, or the size out as a build-time constant.
~Andre
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2018-01-29 11:37 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-01-26 14:48 [PATCH] x86/ioemul: Misc improvements to ioport_emulate.c Andrew Cooper
2018-01-29 11:23 ` Jan Beulich
2018-01-29 11:37 ` Andrew Cooper
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.