How should I proceed to get commit 891348c to 3.0.y?

How should I proceed to get commit 891348c to 3.0.y?

[Robin Holt]

I missed the email from 12/31 indicating commit 891348c was pulled from
the 3.0.y stable tree due to a build breakage.

That commit requires the enum defined in traps.h by commit c940826.  That,
in turn depends on commit 228bdaa95f which depends upon earlier commits.

How should I proceed?  Should I give you a patch that uses the constants
in xpc_main.c? Should I add a patch which just pulls the enum portion of
commit c940826 into traps.h? Any other direction you think this should go?

Thanks,
Robin

Re: How should I proceed to get commit 891348c to 3.0.y?

[Greg KH]

On Thu, Jan 24, 2013 at 09:28:51AM -0600, Robin Holt wrote:
> 
> I missed the email from 12/31 indicating commit 891348c was pulled from
> the 3.0.y stable tree due to a build breakage.
> 
> That commit requires the enum defined in traps.h by commit c940826.  That,
> in turn depends on commit 228bdaa95f which depends upon earlier commits.
> 
> How should I proceed?  Should I give you a patch that uses the constants
> in xpc_main.c? Should I add a patch which just pulls the enum portion of
> commit c940826 into traps.h? Any other direction you think this should go?

I think the last one makes the most sense as that makes it all
self-contained and sensible, right?

thanks,

greg k-h

Re: How should I proceed to get commit 891348c to 3.0.y?

[Robin Holt]

On Thu, Jan 24, 2013 at 10:15:35AM -0800, Greg KH wrote:
> On Thu, Jan 24, 2013 at 09:28:51AM -0600, Robin Holt wrote:
> > 
> > I missed the email from 12/31 indicating commit 891348c was pulled from
> > the 3.0.y stable tree due to a build breakage.
> > 
> > That commit requires the enum defined in traps.h by commit c940826.  That,
> > in turn depends on commit 228bdaa95f which depends upon earlier commits.
> > 
> > How should I proceed?  Should I give you a patch that uses the constants
> > in xpc_main.c? Should I add a patch which just pulls the enum portion of
> > commit c940826 into traps.h? Any other direction you think this should go?
> 
> I think the last one makes the most sense as that makes it all
> self-contained and sensible, right?

git diff c940826~..c940826 -- arch/x86/include/asm/traps.h | patch -p1
git commit -a -c c940826

Then fixed it up to add a Signed-off-by: Robin Holt <holt@sgi.com>

I then did a make x86_64_defconfig, make menuconfig to get SGI_XP
selected.  I ran into a compile error where traps.h line 56 was failing
on __kprobes not being defined.  I added the #include <linux/kprobes.h>
and now it builds.

Here is the munged commit I currently have.  With that 891348c can be
cherry-picked and will build.  How should I proceed?  Is this enough
for you to make progress?  Should I clear out those Signed-off-by
lines for the submitter and HPA or should those stay?

Thank you for your patience with my ignorance,
Robin Holt


commit b4ff8081e86ebf7bb8c1d4c93dfaaf81a4539ee6
Author: Kees Cook <keescook@chromium.org>
Date:   Fri Mar 9 16:07:10 2012 -0800

    x86: Use enum instead of literals for trap values [PARTIAL c940826]
    
    The traps are referred to by their numbers and it can be difficult to
    understand them while reading the code without context. This patch adds
    enumeration of the trap numbers and replaces the numbers with the correct
    enum for x86.
    
    Signed-off-by: Kees Cook <keescook@chromium.org>
    Link: http://lkml.kernel.org/r/20120310000710.GA32667@www.outflux.net
    Signed-off-by: H. Peter Anvin <hpa@zytor.com>
    Signed-off-by: Robin Holt <holt@sgi.com>

diff --git a/arch/x86/include/asm/traps.h b/arch/x86/include/asm/traps.h
index 0310da6..1d44903 100644
--- a/arch/x86/include/asm/traps.h
+++ b/arch/x86/include/asm/traps.h
@@ -1,6 +1,7 @@
 #ifndef _ASM_X86_TRAPS_H
 #define _ASM_X86_TRAPS_H
 
+#include <linux/kprobes.h>
 #include <asm/debugreg.h>
 #include <asm/siginfo.h>			/* TRAP_TRACE, ... */
 
@@ -87,4 +88,29 @@ asmlinkage void smp_thermal_interrupt(void);
 asmlinkage void mce_threshold_interrupt(void);
 #endif
 
+/* Interrupts/Exceptions */
+enum {
+	X86_TRAP_DE = 0,	/*  0, Divide-by-zero */
+	X86_TRAP_DB,		/*  1, Debug */
+	X86_TRAP_NMI,		/*  2, Non-maskable Interrupt */
+	X86_TRAP_BP,		/*  3, Breakpoint */
+	X86_TRAP_OF,		/*  4, Overflow */
+	X86_TRAP_BR,		/*  5, Bound Range Exceeded */
+	X86_TRAP_UD,		/*  6, Invalid Opcode */
+	X86_TRAP_NM,		/*  7, Device Not Available */
+	X86_TRAP_DF,		/*  8, Double Fault */
+	X86_TRAP_OLD_MF,	/*  9, Coprocessor Segment Overrun */
+	X86_TRAP_TS,		/* 10, Invalid TSS */
+	X86_TRAP_NP,		/* 11, Segment Not Present */
+	X86_TRAP_SS,		/* 12, Stack Segment Fault */
+	X86_TRAP_GP,		/* 13, General Protection Fault */
+	X86_TRAP_PF,		/* 14, Page Fault */
+	X86_TRAP_SPURIOUS,	/* 15, Spurious Interrupt */
+	X86_TRAP_MF,		/* 16, x87 Floating-Point Exception */
+	X86_TRAP_AC,		/* 17, Alignment Check */
+	X86_TRAP_MC,		/* 18, Machine Check */
+	X86_TRAP_XF,		/* 19, SIMD Floating-Point Exception */
+	X86_TRAP_IRET = 32,	/* 32, IRET Exception */
+};
+
 #endif /* _ASM_X86_TRAPS_H */

Re: How should I proceed to get commit 891348c to 3.0.y?

[Greg KH]

On Thu, Jan 24, 2013 at 12:30:10PM -0600, Robin Holt wrote:
> On Thu, Jan 24, 2013 at 10:15:35AM -0800, Greg KH wrote:
> > On Thu, Jan 24, 2013 at 09:28:51AM -0600, Robin Holt wrote:
> > > 
> > > I missed the email from 12/31 indicating commit 891348c was pulled from
> > > the 3.0.y stable tree due to a build breakage.
> > > 
> > > That commit requires the enum defined in traps.h by commit c940826.  That,
> > > in turn depends on commit 228bdaa95f which depends upon earlier commits.
> > > 
> > > How should I proceed?  Should I give you a patch that uses the constants
> > > in xpc_main.c? Should I add a patch which just pulls the enum portion of
> > > commit c940826 into traps.h? Any other direction you think this should go?
> > 
> > I think the last one makes the most sense as that makes it all
> > self-contained and sensible, right?
> 
> git diff c940826~..c940826 -- arch/x86/include/asm/traps.h | patch -p1
> git commit -a -c c940826
> 
> Then fixed it up to add a Signed-off-by: Robin Holt <holt@sgi.com>
> 
> I then did a make x86_64_defconfig, make menuconfig to get SGI_XP
> selected.  I ran into a compile error where traps.h line 56 was failing
> on __kprobes not being defined.  I added the #include <linux/kprobes.h>
> and now it builds.
> 
> Here is the munged commit I currently have.  With that 891348c can be
> cherry-picked and will build.  How should I proceed?  Is this enough
> for you to make progress?  Should I clear out those Signed-off-by
> lines for the submitter and HPA or should those stay?

They should stay.  resend this saying where the patch came from, and I
will be glad to apply it.  Also tell me what the original patch you
wanted me to apply was as well, I forgot it :)

thanks,

greg k-h

Re: How should I proceed to get commit 891348c to 3.0.y?

[Robin Holt]

On Thu, Jan 24, 2013 at 12:00:34PM -0800, Greg KH wrote:
> On Thu, Jan 24, 2013 at 12:30:10PM -0600, Robin Holt wrote:
> > On Thu, Jan 24, 2013 at 10:15:35AM -0800, Greg KH wrote:
> > > On Thu, Jan 24, 2013 at 09:28:51AM -0600, Robin Holt wrote:
> > > > 
> > > > I missed the email from 12/31 indicating commit 891348c was pulled from
> > > > the 3.0.y stable tree due to a build breakage.
> > > > 
> > > > That commit requires the enum defined in traps.h by commit c940826.  That,
> > > > in turn depends on commit 228bdaa95f which depends upon earlier commits.
> > > > 
> > > > How should I proceed?  Should I give you a patch that uses the constants
> > > > in xpc_main.c? Should I add a patch which just pulls the enum portion of
> > > > commit c940826 into traps.h? Any other direction you think this should go?
> > > 
> > > I think the last one makes the most sense as that makes it all
> > > self-contained and sensible, right?
> > 
> > git diff c940826~..c940826 -- arch/x86/include/asm/traps.h | patch -p1
> > git commit -a -c c940826
> > 
> > Then fixed it up to add a Signed-off-by: Robin Holt <holt@sgi.com>
> > 
> > I then did a make x86_64_defconfig, make menuconfig to get SGI_XP
> > selected.  I ran into a compile error where traps.h line 56 was failing
> > on __kprobes not being defined.  I added the #include <linux/kprobes.h>
> > and now it builds.
> > 
> > Here is the munged commit I currently have.  With that 891348c can be
> > cherry-picked and will build.  How should I proceed?  Is this enough
> > for you to make progress?  Should I clear out those Signed-off-by
> > lines for the submitter and HPA or should those stay?
> 
> They should stay.  resend this saying where the patch came from, and I
> will be glad to apply it.  Also tell me what the original patch you
> wanted me to apply was as well, I forgot it :)

I changed the "Author:" line to me and added my own Signed-off-by:
line.

The original commit which needed this fixup was 891348c

Thanks,
Robin

commit b4ff8081e86ebf7bb8c1d4c93dfaaf81a4539ee6
Author: Robin Holt <holt@sgi.com>
Date:   Fri Mar 9 16:07:10 2012 -0800

    x86: Use enum instead of literals for trap values [PARTIAL]
    
    The traps are referred to by their numbers and it can be difficult to
    understand them while reading the code without context. This patch adds
    enumeration of the trap numbers and replaces the numbers with the correct
    enum for x86.
    
    Signed-off-by: Kees Cook <keescook@chromium.org>
    Link: http://lkml.kernel.org/r/20120310000710.GA32667@www.outflux.net
    Signed-off-by: H. Peter Anvin <hpa@zytor.com>
    Signed-off-by: Robin Holt <holt@sgi.com>

diff --git a/arch/x86/include/asm/traps.h b/arch/x86/include/asm/traps.h
index 0310da6..1d44903 100644
--- a/arch/x86/include/asm/traps.h
+++ b/arch/x86/include/asm/traps.h
@@ -1,6 +1,7 @@
 #ifndef _ASM_X86_TRAPS_H
 #define _ASM_X86_TRAPS_H
 
+#include <linux/kprobes.h>
 #include <asm/debugreg.h>
 #include <asm/siginfo.h>			/* TRAP_TRACE, ... */
 
@@ -87,4 +88,29 @@ asmlinkage void smp_thermal_interrupt(void);
 asmlinkage void mce_threshold_interrupt(void);
 #endif
 
+/* Interrupts/Exceptions */
+enum {
+	X86_TRAP_DE = 0,	/*  0, Divide-by-zero */
+	X86_TRAP_DB,		/*  1, Debug */
+	X86_TRAP_NMI,		/*  2, Non-maskable Interrupt */
+	X86_TRAP_BP,		/*  3, Breakpoint */
+	X86_TRAP_OF,		/*  4, Overflow */
+	X86_TRAP_BR,		/*  5, Bound Range Exceeded */
+	X86_TRAP_UD,		/*  6, Invalid Opcode */
+	X86_TRAP_NM,		/*  7, Device Not Available */
+	X86_TRAP_DF,		/*  8, Double Fault */
+	X86_TRAP_OLD_MF,	/*  9, Coprocessor Segment Overrun */
+	X86_TRAP_TS,		/* 10, Invalid TSS */
+	X86_TRAP_NP,		/* 11, Segment Not Present */
+	X86_TRAP_SS,		/* 12, Stack Segment Fault */
+	X86_TRAP_GP,		/* 13, General Protection Fault */
+	X86_TRAP_PF,		/* 14, Page Fault */
+	X86_TRAP_SPURIOUS,	/* 15, Spurious Interrupt */
+	X86_TRAP_MF,		/* 16, x87 Floating-Point Exception */
+	X86_TRAP_AC,		/* 17, Alignment Check */
+	X86_TRAP_MC,		/* 18, Machine Check */
+	X86_TRAP_XF,		/* 19, SIMD Floating-Point Exception */
+	X86_TRAP_IRET = 32,	/* 32, IRET Exception */
+};
+
 #endif /* _ASM_X86_TRAPS_H */

Re: How should I proceed to get commit 891348c to 3.0.y?

[Greg KH]

On Thu, Jan 24, 2013 at 02:14:20PM -0600, Robin Holt wrote:
> On Thu, Jan 24, 2013 at 12:00:34PM -0800, Greg KH wrote:
> > On Thu, Jan 24, 2013 at 12:30:10PM -0600, Robin Holt wrote:
> > > On Thu, Jan 24, 2013 at 10:15:35AM -0800, Greg KH wrote:
> > > > On Thu, Jan 24, 2013 at 09:28:51AM -0600, Robin Holt wrote:
> > > > > 
> > > > > I missed the email from 12/31 indicating commit 891348c was pulled from
> > > > > the 3.0.y stable tree due to a build breakage.
> > > > > 
> > > > > That commit requires the enum defined in traps.h by commit c940826.  That,
> > > > > in turn depends on commit 228bdaa95f which depends upon earlier commits.
> > > > > 
> > > > > How should I proceed?  Should I give you a patch that uses the constants
> > > > > in xpc_main.c? Should I add a patch which just pulls the enum portion of
> > > > > commit c940826 into traps.h? Any other direction you think this should go?
> > > > 
> > > > I think the last one makes the most sense as that makes it all
> > > > self-contained and sensible, right?
> > > 
> > > git diff c940826~..c940826 -- arch/x86/include/asm/traps.h | patch -p1
> > > git commit -a -c c940826
> > > 
> > > Then fixed it up to add a Signed-off-by: Robin Holt <holt@sgi.com>
> > > 
> > > I then did a make x86_64_defconfig, make menuconfig to get SGI_XP
> > > selected.  I ran into a compile error where traps.h line 56 was failing
> > > on __kprobes not being defined.  I added the #include <linux/kprobes.h>
> > > and now it builds.
> > > 
> > > Here is the munged commit I currently have.  With that 891348c can be
> > > cherry-picked and will build.  How should I proceed?  Is this enough
> > > for you to make progress?  Should I clear out those Signed-off-by
> > > lines for the submitter and HPA or should those stay?
> > 
> > They should stay.  resend this saying where the patch came from, and I
> > will be glad to apply it.  Also tell me what the original patch you
> > wanted me to apply was as well, I forgot it :)
> 
> I changed the "Author:" line to me and added my own Signed-off-by:
> line.
> 
> The original commit which needed this fixup was 891348c
> 
> Thanks,
> Robin
> 
> commit b4ff8081e86ebf7bb8c1d4c93dfaaf81a4539ee6
> Author: Robin Holt <holt@sgi.com>
> Date:   Fri Mar 9 16:07:10 2012 -0800
> 
>     x86: Use enum instead of literals for trap values [PARTIAL]
>     
>     The traps are referred to by their numbers and it can be difficult to
>     understand them while reading the code without context. This patch adds
>     enumeration of the trap numbers and replaces the numbers with the correct
>     enum for x86.
>     
>     Signed-off-by: Kees Cook <keescook@chromium.org>
>     Link: http://lkml.kernel.org/r/20120310000710.GA32667@www.outflux.net
>     Signed-off-by: H. Peter Anvin <hpa@zytor.com>
>     Signed-off-by: Robin Holt <holt@sgi.com>

No, don't loose the authorship info.

{sigh}

Ok, I've edited this up by hand and included it, and the original patch
in the 3.0-stable tree, you should have an email about it now.  Please
verify I got it all right.

greg k-h

Re: How should I proceed to get commit 891348c to 3.0.y?

[Robin Holt]

On Thu, Jan 24, 2013 at 12:50:50PM -0800, Greg KH wrote:
> On Thu, Jan 24, 2013 at 02:14:20PM -0600, Robin Holt wrote:
> > On Thu, Jan 24, 2013 at 12:00:34PM -0800, Greg KH wrote:
> > > On Thu, Jan 24, 2013 at 12:30:10PM -0600, Robin Holt wrote:
> > > > On Thu, Jan 24, 2013 at 10:15:35AM -0800, Greg KH wrote:
> > > > > On Thu, Jan 24, 2013 at 09:28:51AM -0600, Robin Holt wrote:
> > > > > > 
> > > > > > I missed the email from 12/31 indicating commit 891348c was pulled from
> > > > > > the 3.0.y stable tree due to a build breakage.
> > > > > > 
> > > > > > That commit requires the enum defined in traps.h by commit c940826.  That,
> > > > > > in turn depends on commit 228bdaa95f which depends upon earlier commits.
> > > > > > 
> > > > > > How should I proceed?  Should I give you a patch that uses the constants
> > > > > > in xpc_main.c? Should I add a patch which just pulls the enum portion of
> > > > > > commit c940826 into traps.h? Any other direction you think this should go?
> > > > > 
> > > > > I think the last one makes the most sense as that makes it all
> > > > > self-contained and sensible, right?
> > > > 
> > > > git diff c940826~..c940826 -- arch/x86/include/asm/traps.h | patch -p1
> > > > git commit -a -c c940826
> > > > 
> > > > Then fixed it up to add a Signed-off-by: Robin Holt <holt@sgi.com>
> > > > 
> > > > I then did a make x86_64_defconfig, make menuconfig to get SGI_XP
> > > > selected.  I ran into a compile error where traps.h line 56 was failing
> > > > on __kprobes not being defined.  I added the #include <linux/kprobes.h>
> > > > and now it builds.
> > > > 
> > > > Here is the munged commit I currently have.  With that 891348c can be
> > > > cherry-picked and will build.  How should I proceed?  Is this enough
> > > > for you to make progress?  Should I clear out those Signed-off-by
> > > > lines for the submitter and HPA or should those stay?
> > > 
> > > They should stay.  resend this saying where the patch came from, and I
> > > will be glad to apply it.  Also tell me what the original patch you
> > > wanted me to apply was as well, I forgot it :)
> > 
> > I changed the "Author:" line to me and added my own Signed-off-by:
> > line.
> > 
> > The original commit which needed this fixup was 891348c
> > 
> > Thanks,
> > Robin
> > 
> > commit b4ff8081e86ebf7bb8c1d4c93dfaaf81a4539ee6
> > Author: Robin Holt <holt@sgi.com>
> > Date:   Fri Mar 9 16:07:10 2012 -0800
> > 
> >     x86: Use enum instead of literals for trap values [PARTIAL]
> >     
> >     The traps are referred to by their numbers and it can be difficult to
> >     understand them while reading the code without context. This patch adds
> >     enumeration of the trap numbers and replaces the numbers with the correct
> >     enum for x86.
> >     
> >     Signed-off-by: Kees Cook <keescook@chromium.org>
> >     Link: http://lkml.kernel.org/r/20120310000710.GA32667@www.outflux.net
> >     Signed-off-by: H. Peter Anvin <hpa@zytor.com>
> >     Signed-off-by: Robin Holt <holt@sgi.com>
> 
> No, don't loose the authorship info.
> 
> {sigh}
> 
> Ok, I've edited this up by hand and included it, and the original patch
> in the 3.0-stable tree, you should have an email about it now.  Please
> verify I got it all right.

That looks good to me.

Thank you,
Robin Holt