* [patch] simulate_ll and simulate_sc(resend)
@ 2003-03-03 10:21 Yoichi Yuasa
2003-03-04 12:07 ` Ralf Baechle
2003-03-04 20:36 ` Ralf Baechle
0 siblings, 2 replies; 5+ messages in thread
From: Yoichi Yuasa @ 2003-03-03 10:21 UTC (permalink / raw)
To: ralf; +Cc: yoichi_yuasa, linux-mips, jsun
[-- Attachment #1: Type: text/plain, Size: 331 bytes --]
Hi Ralf,
I found a bug in simulate_ll and simulate_sc.
The board that uses ll/sc emulation is not started.
When send_sig in these, in order not to operate the value of EPC correctly,
simulate_* happens continuously.
The previous patches were not perfect, I changed more.
Please apply these patches to CVS tree.
Thanks,
Yoichi
[-- Attachment #2: llsc-v24-0303.diff --]
[-- Type: text/plain, Size: 722 bytes --]
diff -aruN --exclude=CVS --exclude=.cvsignore linux.orig/arch/mips/kernel/traps.c linux/arch/mips/kernel/traps.c
--- linux.orig/arch/mips/kernel/traps.c Tue Feb 11 07:50:48 2003
+++ linux/arch/mips/kernel/traps.c Mon Mar 3 17:52:16 2003
@@ -140,6 +140,7 @@
return;
sig:
+ compute_return_epc(regs);
send_sig(signal, current, 1);
}
@@ -172,18 +173,19 @@
}
if (ll_bit == 0 || ll_task != current) {
regs->regs[reg] = 0;
- goto sig;
}
- if (put_user(regs->regs[reg], vaddr))
+ if (put_user(regs->regs[reg], vaddr)) {
signal = SIGSEGV;
- else
+ goto sig;
+ } else
regs->regs[reg] = 1;
compute_return_epc(regs);
return;
sig:
+ compute_return_epc(regs);
send_sig(signal, current, 1);
}
[-- Attachment #3: llsc-v25-0303.diff --]
[-- Type: text/plain, Size: 722 bytes --]
diff -aruN --exclude=CVS --exclude=.cvsignore linux.orig/arch/mips/kernel/traps.c linux/arch/mips/kernel/traps.c
--- linux.orig/arch/mips/kernel/traps.c Wed Feb 12 13:26:43 2003
+++ linux/arch/mips/kernel/traps.c Mon Mar 3 17:48:11 2003
@@ -135,6 +135,7 @@
return;
sig:
+ compute_return_epc(regs);
send_sig(signal, current, 1);
}
@@ -167,18 +168,19 @@
}
if (ll_bit == 0 || ll_task != current) {
regs->regs[reg] = 0;
- goto sig;
}
- if (put_user(regs->regs[reg], vaddr))
+ if (put_user(regs->regs[reg], vaddr)) {
signal = SIGSEGV;
- else
+ goto sig;
+ } else
regs->regs[reg] = 1;
compute_return_epc(regs);
return;
sig:
+ compute_return_epc(regs);
send_sig(signal, current, 1);
}
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [patch] simulate_ll and simulate_sc(resend)
2003-03-03 10:21 [patch] simulate_ll and simulate_sc(resend) Yoichi Yuasa
@ 2003-03-04 12:07 ` Ralf Baechle
2003-03-04 20:36 ` Ralf Baechle
1 sibling, 0 replies; 5+ messages in thread
From: Ralf Baechle @ 2003-03-04 12:07 UTC (permalink / raw)
To: Yoichi Yuasa; +Cc: linux-mips, jsun
On Mon, Mar 03, 2003 at 07:21:37PM +0900, Yoichi Yuasa wrote:
> sig:
> + compute_return_epc(regs);
> send_sig(signal, current, 1);
In the error case you can't advance the epc ..
It's an old bug already but the signal should be sent with force_sig,
not send_sig.
Ralf
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [patch] simulate_ll and simulate_sc(resend)
2003-03-03 10:21 [patch] simulate_ll and simulate_sc(resend) Yoichi Yuasa
2003-03-04 12:07 ` Ralf Baechle
@ 2003-03-04 20:36 ` Ralf Baechle
2003-03-06 7:53 ` Yoichi Yuasa
1 sibling, 1 reply; 5+ messages in thread
From: Ralf Baechle @ 2003-03-04 20:36 UTC (permalink / raw)
To: Yoichi Yuasa; +Cc: linux-mips, jsun
On Mon, Mar 03, 2003 at 07:21:37PM +0900, Yoichi Yuasa wrote:
> I found a bug in simulate_ll and simulate_sc.
> The board that uses ll/sc emulation is not started.
>
> When send_sig in these, in order not to operate the value of EPC correctly,
> simulate_* happens continuously.
>
> The previous patches were not perfect, I changed more.
> Please apply these patches to CVS tree.
As previously mentioned there were some problems with your fix, so I
wrote an alternative fix which is attached below. It's untested because
I don't have any ll/sc-less test platform.
Ralf
Index: arch/mips/kernel/traps.c
===================================================================
RCS file: /home/cvs/linux/arch/mips/kernel/traps.c,v
retrieving revision 1.99.2.41
diff -u -r1.99.2.41 traps.c
--- arch/mips/kernel/traps.c 10 Feb 2003 22:50:48 -0000 1.99.2.41
+++ arch/mips/kernel/traps.c 4 Mar 2003 20:32:10 -0000
@@ -134,13 +134,14 @@
ll_bit = 0;
}
ll_task = current;
+
regs->regs[(opcode & RT) >> 16] = value;
compute_return_epc(regs);
return;
sig:
- send_sig(signal, current, 1);
+ force_sig(signal, current);
}
static inline void simulate_sc(struct pt_regs *regs, unsigned int opcode)
@@ -172,19 +173,21 @@
}
if (ll_bit == 0 || ll_task != current) {
regs->regs[reg] = 0;
- goto sig;
+ return;
}
- if (put_user(regs->regs[reg], vaddr))
+ if (put_user(regs->regs[reg], vaddr)) {
signal = SIGSEGV;
- else
- regs->regs[reg] = 1;
+ goto sig;
+ }
+
+ regs->regs[reg] = 1;
compute_return_epc(regs);
return;
sig:
- send_sig(signal, current, 1);
+ force_sig(signal, current);
}
/*
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [patch] simulate_ll and simulate_sc(resend)
2003-03-04 20:36 ` Ralf Baechle
@ 2003-03-06 7:53 ` Yoichi Yuasa
2003-03-10 3:41 ` Yoichi Yuasa
0 siblings, 1 reply; 5+ messages in thread
From: Yoichi Yuasa @ 2003-03-06 7:53 UTC (permalink / raw)
To: Ralf Baechle; +Cc: yoichi_yuasa, linux-mips, jsun
Hi Ralf,
On Tue, 4 Mar 2003 21:36:05 +0100
Ralf Baechle <ralf@linux-mips.org> wrote:
> On Mon, Mar 03, 2003 at 07:21:37PM +0900, Yoichi Yuasa wrote:
>
> > I found a bug in simulate_ll and simulate_sc.
> > The board that uses ll/sc emulation is not started.
> >
> > When send_sig in these, in order not to operate the value of EPC correctly,
> > simulate_* happens continuously.
> >
> > The previous patches were not perfect, I changed more.
> > Please apply these patches to CVS tree.
>
> As previously mentioned there were some problems with your fix, so I
> wrote an alternative fix which is attached below. It's untested because
> I don't have any ll/sc-less test platform.
I tested ll/sc simulate.
Furthermore, the following patch is also needed.
Thanks,
Yoichi
diff -aruN --exclude=CVS --exclude=.cvsignore linux.orig/arch/mips/kernel/traps.c linux/arch/mips/kernel/traps.c
--- linux.orig/arch/mips/kernel/traps.c Wed Mar 5 12:05:00 2003
+++ linux/arch/mips/kernel/traps.c Thu Mar 6 16:37:40 2003
@@ -173,6 +173,7 @@
}
if (ll_bit == 0 || ll_task != current) {
regs->regs[reg] = 0;
+ compute_return_epc(regs);
return;
}
>
> Ralf
>
> Index: arch/mips/kernel/traps.c
> ===================================================================
> RCS file: /home/cvs/linux/arch/mips/kernel/traps.c,v
> retrieving revision 1.99.2.41
> diff -u -r1.99.2.41 traps.c
> --- arch/mips/kernel/traps.c 10 Feb 2003 22:50:48 -0000 1.99.2.41
> +++ arch/mips/kernel/traps.c 4 Mar 2003 20:32:10 -0000
> @@ -134,13 +134,14 @@
> ll_bit = 0;
> }
> ll_task = current;
> +
> regs->regs[(opcode & RT) >> 16] = value;
>
> compute_return_epc(regs);
> return;
>
> sig:
> - send_sig(signal, current, 1);
> + force_sig(signal, current);
> }
>
> static inline void simulate_sc(struct pt_regs *regs, unsigned int opcode)
> @@ -172,19 +173,21 @@
> }
> if (ll_bit == 0 || ll_task != current) {
> regs->regs[reg] = 0;
> - goto sig;
> + return;
> }
>
> - if (put_user(regs->regs[reg], vaddr))
> + if (put_user(regs->regs[reg], vaddr)) {
> signal = SIGSEGV;
> - else
> - regs->regs[reg] = 1;
> + goto sig;
> + }
> +
> + regs->regs[reg] = 1;
>
> compute_return_epc(regs);
> return;
>
> sig:
> - send_sig(signal, current, 1);
> + force_sig(signal, current);
> }
>
> /*
>
>
--
Yoichi Yuasa
Montavista Software Japan, Inc.
e-mail: yoichi_yuasa@montavista.co.jp
http://www.montavista.co.jp
PHONE: 03-5469-8840 FAX: 03-5469-8841
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [patch] simulate_ll and simulate_sc(resend)
2003-03-06 7:53 ` Yoichi Yuasa
@ 2003-03-10 3:41 ` Yoichi Yuasa
0 siblings, 0 replies; 5+ messages in thread
From: Yoichi Yuasa @ 2003-03-10 3:41 UTC (permalink / raw)
To: ralf; +Cc: yoichi_yuasa, linux-mips, jsun
[-- Attachment #1: Type: text/plain, Size: 3187 bytes --]
Hi Ralf,
In addition to your change, this patch is still more required.
I tested simulate_ll/sc operation with this patch, it doesn't have a problem.
Please apply this patch.
Thanks,
Yoichi
On Thu, 6 Mar 2003 16:53:51 +0900
Yoichi Yuasa <yoichi_yuasa@montavista.co.jp> wrote:
> Hi Ralf,
>
> On Tue, 4 Mar 2003 21:36:05 +0100
> Ralf Baechle <ralf@linux-mips.org> wrote:
>
> > On Mon, Mar 03, 2003 at 07:21:37PM +0900, Yoichi Yuasa wrote:
> >
> > > I found a bug in simulate_ll and simulate_sc.
> > > The board that uses ll/sc emulation is not started.
> > >
> > > When send_sig in these, in order not to operate the value of EPC correctly,
> > > simulate_* happens continuously.
> > >
> > > The previous patches were not perfect, I changed more.
> > > Please apply these patches to CVS tree.
> >
> > As previously mentioned there were some problems with your fix, so I
> > wrote an alternative fix which is attached below. It's untested because
> > I don't have any ll/sc-less test platform.
>
> I tested ll/sc simulate.
> Furthermore, the following patch is also needed.
>
> Thanks,
>
> Yoichi
>
> diff -aruN --exclude=CVS --exclude=.cvsignore linux.orig/arch/mips/kernel/traps.c linux/arch/mips/kernel/traps.c
> --- linux.orig/arch/mips/kernel/traps.c Wed Mar 5 12:05:00 2003
> +++ linux/arch/mips/kernel/traps.c Thu Mar 6 16:37:40 2003
> @@ -173,6 +173,7 @@
> }
> if (ll_bit == 0 || ll_task != current) {
> regs->regs[reg] = 0;
> + compute_return_epc(regs);
> return;
> }
>
>
>
> >
> > Ralf
> >
> > Index: arch/mips/kernel/traps.c
> > ===================================================================
> > RCS file: /home/cvs/linux/arch/mips/kernel/traps.c,v
> > retrieving revision 1.99.2.41
> > diff -u -r1.99.2.41 traps.c
> > --- arch/mips/kernel/traps.c 10 Feb 2003 22:50:48 -0000 1.99.2.41
> > +++ arch/mips/kernel/traps.c 4 Mar 2003 20:32:10 -0000
> > @@ -134,13 +134,14 @@
> > ll_bit = 0;
> > }
> > ll_task = current;
> > +
> > regs->regs[(opcode & RT) >> 16] = value;
> >
> > compute_return_epc(regs);
> > return;
> >
> > sig:
> > - send_sig(signal, current, 1);
> > + force_sig(signal, current);
> > }
> >
> > static inline void simulate_sc(struct pt_regs *regs, unsigned int opcode)
> > @@ -172,19 +173,21 @@
> > }
> > if (ll_bit == 0 || ll_task != current) {
> > regs->regs[reg] = 0;
> > - goto sig;
> > + return;
> > }
> >
> > - if (put_user(regs->regs[reg], vaddr))
> > + if (put_user(regs->regs[reg], vaddr)) {
> > signal = SIGSEGV;
> > - else
> > - regs->regs[reg] = 1;
> > + goto sig;
> > + }
> > +
> > + regs->regs[reg] = 1;
> >
> > compute_return_epc(regs);
> > return;
> >
> > sig:
> > - send_sig(signal, current, 1);
> > + force_sig(signal, current);
> > }
> >
> > /*
> >
> >
>
>
> --
> Yoichi Yuasa
> Montavista Software Japan, Inc.
> e-mail: yoichi_yuasa@montavista.co.jp
> http://www.montavista.co.jp
> PHONE: 03-5469-8840 FAX: 03-5469-8841
>
>
--
Yoichi Yuasa
Montavista Software Japan, Inc.
e-mail: yoichi_yuasa@montavista.co.jp
http://www.montavista.co.jp
PHONE: 03-5469-8840 FAX: 03-5469-8841
[-- Attachment #2: simulate_llsc.diff --]
[-- Type: text/plain, Size: 375 bytes --]
diff -aruN --exclude=CVS --exclude=.cvsignore linux.orig/arch/mips/kernel/traps.c linux/arch/mips/kernel/traps.c
--- linux.orig/arch/mips/kernel/traps.c Wed Mar 5 12:05:00 2003
+++ linux/arch/mips/kernel/traps.c Thu Mar 6 16:37:40 2003
@@ -173,6 +173,7 @@
}
if (ll_bit == 0 || ll_task != current) {
regs->regs[reg] = 0;
+ compute_return_epc(regs);
return;
}
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2003-03-10 3:41 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-03-03 10:21 [patch] simulate_ll and simulate_sc(resend) Yoichi Yuasa
2003-03-04 12:07 ` Ralf Baechle
2003-03-04 20:36 ` Ralf Baechle
2003-03-06 7:53 ` Yoichi Yuasa
2003-03-10 3:41 ` Yoichi Yuasa
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox