| Message ID | 1544073508-13720-4-git-send-email-firoz.khan@linaro.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | mips: system call table generation support | expand |
Hi Firoz, On Thu, Dec 06, 2018 at 10:48:24AM +0530, Firoz Khan wrote: > diff --git a/arch/mips/include/uapi/asm/sgidefs.h b/arch/mips/include/uapi/asm/sgidefs.h > index 26143e3..0364eec 100644 > --- a/arch/mips/include/uapi/asm/sgidefs.h > +++ b/arch/mips/include/uapi/asm/sgidefs.h > @@ -40,6 +40,6 @@ > */ > #define _MIPS_SIM_ABI32 1 > #define _MIPS_SIM_NABI32 2 > -#define _MIPS_SIM_ABI64 3 > +#define _MIPS_SIM_ABIN64 3 Whilst I agree with changing our own definitions & filenames to use n64, this macro actually reflects naming used by the toolchain. ie: $ mips-linux-gcc -mips64 -mabi=64 -dM -E - </dev/null | grep ABI64 #define _ABI64 3 #define _MIPS_SIM _ABI64 Our macro here is used to compare against _MIPS_SIM provided by the toolchain, so for consistency I think we should keep the same name for the macro that the toolchain uses. And I realise that undoing that but keeping n64 in our own filenames & macros is another type of inconsistency, but something imperfect is unavoidable at this point given that the engineers way back when decided to use "ABI64" for n64. Thanks, Paul
On Mon, 10 Dec 2018, Paul Burton wrote: > And I realise that undoing that but keeping n64 in our own filenames & > macros is another type of inconsistency, but something imperfect is > unavoidable at this point given that the engineers way back when decided > to use "ABI64" for n64. My feeling has been n32 was invented at SGI as an afterthought, hence the choice of having ABI32 or ABI64 defined for the 32-bit (now o32) and the 64-bit (now n64) ABI respectively was reasonable. FWIW, Maciej
Hi Maciej, On Mon, Dec 10, 2018 at 11:32:46PM +0000, Maciej W. Rozycki wrote: > On Mon, 10 Dec 2018, Paul Burton wrote: > > > And I realise that undoing that but keeping n64 in our own filenames & > > macros is another type of inconsistency, but something imperfect is > > unavoidable at this point given that the engineers way back when decided > > to use "ABI64" for n64. > > My feeling has been n32 was invented at SGI as an afterthought, hence the > choice of having ABI32 or ABI64 defined for the 32-bit (now o32) and the > 64-bit (now n64) ABI respectively was reasonable. I'd agree if _MIPS_SIM were defined as _ABI32 for o32, but: $ mips-linux-gcc -mabi=32 -dM -E - </dev/null | grep ABIO32 #define _ABIO32 1 #define _MIPS_SIM _ABIO32 ...so _MIPS_SIM is: _ABIO32 for o32 _ABIN32 for n32 _ABI64 for n64 That doesn't seem very consistent to me, and means that there inevitably has to be some ugliness once there are multiple 64-bit ABIs. To me it feels like the result of someone thinking "one 64-bit MIPS ABI ought to be enough for anybody". I'm undecided whether that person was shortsighted or a genius whose vision was simply incomprehensible to those of us that followed. Thanks, Paul
Hi Paul, On Tue, 11 Dec 2018 at 01:21, Paul Burton <paul.burton@mips.com> wrote: > > Hi Firoz, > > On Thu, Dec 06, 2018 at 10:48:24AM +0530, Firoz Khan wrote: > > diff --git a/arch/mips/include/uapi/asm/sgidefs.h b/arch/mips/include/uapi/asm/sgidefs.h > > index 26143e3..0364eec 100644 > > --- a/arch/mips/include/uapi/asm/sgidefs.h > > +++ b/arch/mips/include/uapi/asm/sgidefs.h > > @@ -40,6 +40,6 @@ > > */ > > #define _MIPS_SIM_ABI32 1 > > #define _MIPS_SIM_NABI32 2 > > -#define _MIPS_SIM_ABI64 3 > > +#define _MIPS_SIM_ABIN64 3 > > Whilst I agree with changing our own definitions & filenames to use n64, > this macro actually reflects naming used by the toolchain. ie: > > $ mips-linux-gcc -mips64 -mabi=64 -dM -E - </dev/null | grep ABI64 > #define _ABI64 3 > #define _MIPS_SIM _ABI64 > > Our macro here is used to compare against _MIPS_SIM provided by the > toolchain, so for consistency I think we should keep the same name for > the macro that the toolchain uses. Will this below change will help? #define _MIPS_SIM_ABI32 1 #define _MIPS_SIM_NABI32 2 #define _MIPS_SIM_ABI64 3 +#define _MIPS_SIM_ABIN64 _MIPS_SIM_ABI64 Thanks Firoz > > And I realise that undoing that but keeping n64 in our own filenames & > macros is another type of inconsistency, but something imperfect is > unavoidable at this point given that the engineers way back when decided > to use "ABI64" for n64. > > Thanks, > Paul
Hi Paul, > > My feeling has been n32 was invented at SGI as an afterthought, hence the > > choice of having ABI32 or ABI64 defined for the 32-bit (now o32) and the > > 64-bit (now n64) ABI respectively was reasonable. > > I'd agree if _MIPS_SIM were defined as _ABI32 for o32, but: > > $ mips-linux-gcc -mabi=32 -dM -E - </dev/null | grep ABIO32 > #define _ABIO32 1 > #define _MIPS_SIM _ABIO32 > > ...so _MIPS_SIM is: > > _ABIO32 for o32 > _ABIN32 for n32 > _ABI64 for n64 > > That doesn't seem very consistent to me, and means that there inevitably > has to be some ugliness once there are multiple 64-bit ABIs. The _ABI* macros were invented as a GNU feature long after _MIPS_SIM_* ones, i.e. IIRC early 2000s vs mid 1990s. Only _MIPS_SIM_* macros are referred in SGI documentation. It is inconsistent and I don't know why whoever invented the _ABI* macros didn't follow the original naming convention chosen by SGI for _MIPS_SIM_* ones, i.e. _ABI32, _NABI32, _ABI64 (but then I don't know either why some people have troubles with recognising patterns in existing code and then following those patterns when making changes or additions to that code; I guess that's just what some people are like). BTW in case anybody wondered SIM stands for Subprogram Interface Model. > To me it feels like the result of someone thinking "one 64-bit MIPS ABI > ought to be enough for anybody". I'm undecided whether that person was > shortsighted or a genius whose vision was simply incomprehensible to > those of us that followed. I think back when 64-bit processors were a new invention hardly anyone thought about proliferating 64-bit ABIs. Remember that first MIPS systems were high-end workstations and servers and n32 was invented as a means to conserve memory, which wasn't seen as a necessity at the beginning. A secondary goal might have been improving the performance of badly-written 32-bit software, by means of the better function calling convention n32 has over o32. Notice for example that for a long time x86-64 only had a single native 64-bit ABI, which is LP64 like our n64, and it was only quite recently that the x86-64 x32 ILP32 ABI was added, which is like our n32. And yet at this very moment it is being discussed on LKML whether x32 actually has any users beyond developers who maintain it and consequently whether it is worth it to keep support for it in our kernel. To me it means in the workstation and server segment there is still no need to conserve memory (and with the ubiquity of 64-bit systems nowadays the need to maintain unportable 32-bit software has largely disappeared). So if anything was not expected it was the expansion of 64-bit MIPS processors into the embedded market or indeed the scale of the expansion of the embedded market itself, but I wouldn't call the lack of such an expectation short-sightedness. It would be too bold a statement to me, because the change was simply too revolutionary to be considered some 25 years ago. Nobody in early 1990s would think of a 64-bit computer in a ballpoint pen or suchlike being a common device and the consequences for code design it would have. FWIW, Maciej
Hi Firoz, On Tue, Dec 11, 2018 at 02:15:19PM +0530, Firoz Khan wrote: > > > diff --git a/arch/mips/include/uapi/asm/sgidefs.h b/arch/mips/include/uapi/asm/sgidefs.h > > > index 26143e3..0364eec 100644 > > > --- a/arch/mips/include/uapi/asm/sgidefs.h > > > +++ b/arch/mips/include/uapi/asm/sgidefs.h > > > @@ -40,6 +40,6 @@ > > > */ > > > #define _MIPS_SIM_ABI32 1 > > > #define _MIPS_SIM_NABI32 2 > > > -#define _MIPS_SIM_ABI64 3 > > > +#define _MIPS_SIM_ABIN64 3 > > > > Whilst I agree with changing our own definitions & filenames to use n64, > > this macro actually reflects naming used by the toolchain. ie: > > > > $ mips-linux-gcc -mips64 -mabi=64 -dM -E - </dev/null | grep ABI64 > > #define _ABI64 3 > > #define _MIPS_SIM _ABI64 > > > > Our macro here is used to compare against _MIPS_SIM provided by the > > toolchain, so for consistency I think we should keep the same name for > > the macro that the toolchain uses. > > Will this below change will help? > > #define _MIPS_SIM_ABI32 1 > #define _MIPS_SIM_NABI32 2 > #define _MIPS_SIM_ABI64 3 > +#define _MIPS_SIM_ABIN64 _MIPS_SIM_ABI64 Hmm, I think I'd prefer that we just keep using _MIPS_SIM_ABI64. Side note - this snippet of code shows another inconsistency... we have _MIPS_SIM_NABI32 whilst the toolchain uses _ABIN32... No need to worry about that here though. In short let's leave _MIPS_SIM_* alone in this series. Thanks, Paul
Hi Paul, On Wed, 12 Dec 2018 at 00:29, Paul Burton <paul.burton@mips.com> wrote: > > Will this below change will help? > > > > #define _MIPS_SIM_ABI32 1 > > #define _MIPS_SIM_NABI32 2 > > #define _MIPS_SIM_ABI64 3 > > +#define _MIPS_SIM_ABIN64 _MIPS_SIM_ABI64 > > Hmm, I think I'd prefer that we just keep using _MIPS_SIM_ABI64. Sure, I think '64' to 'n64' conversion must be remove it from this patch series.I can send v5 without '64' to 'n64' conversion. If we rename '64' to 'n64' in half of the place and this _MIPS_SIM_ABI64 if we half to keep it in same looks not good (according to me). (FYI, This macro name - _MIPS_SIM_ABIN64 must be _MIPS_SIM_NABI64 and defintion will be: +#define _MIPS_SIM_NABI64 _MIPS_SIM_ABI64) So If you confirm I can send v5 without '64' to 'n64' conversion (not just above one, completely from this patch series). Or uou can take a call just keep this macro - _MIPS_SIM_ABI64 as it is and change it rest of the place. Thanks Firoz
Hi Firoz, On Wed, Dec 12, 2018 at 10:04:47AM +0530, Firoz Khan wrote: > > > Will this below change will help? > > > > > > #define _MIPS_SIM_ABI32 1 > > > #define _MIPS_SIM_NABI32 2 > > > #define _MIPS_SIM_ABI64 3 > > > +#define _MIPS_SIM_ABIN64 _MIPS_SIM_ABI64 > > > > Hmm, I think I'd prefer that we just keep using _MIPS_SIM_ABI64. > > Sure, I think '64' to 'n64' conversion must be remove it from this patch > series.I can send v5 without '64' to 'n64' conversion. > > If we rename '64' to 'n64' in half of the place and this _MIPS_SIM_ABI64 > if we half to keep it in same looks not good (according to me). > > (FYI, This macro name - _MIPS_SIM_ABIN64 must be _MIPS_SIM_NABI64 > and defintion will be: > +#define _MIPS_SIM_NABI64 _MIPS_SIM_ABI64) > > So If you confirm I can send v5 without '64' to 'n64' conversion (not just above > one, completely from this patch series). Or uou can take a call just > keep this macro - > _MIPS_SIM_ABI64 as it is and change it rest of the place. Let's just go ahead & leave everything as 64, and I'll do the 64 -> n64 rename later. I hoped whilst you were adding n64-specific code this would be an easy change, but at this point let's just prioritize getting the series applied without the naming change so it can sit in -next for a while before the merge window. Thanks, Paul
Hi Paul, On Thu, 13 Dec 2018 at 03:58, Paul Burton <paul.burton@mips.com> wrote: > > Hi Firoz, > > On Wed, Dec 12, 2018 at 10:04:47AM +0530, Firoz Khan wrote: > > > > Sure, I think '64' to 'n64' conversion must be remove it from this patch > > series.I can send v5 without '64' to 'n64' conversion. > > > > If we rename '64' to 'n64' in half of the place and this _MIPS_SIM_ABI64 > > if we half to keep it in same looks not good (according to me). > > > > (FYI, This macro name - _MIPS_SIM_ABIN64 must be _MIPS_SIM_NABI64 > > and defintion will be: > > +#define _MIPS_SIM_NABI64 _MIPS_SIM_ABI64) > > > > So If you confirm I can send v5 without '64' to 'n64' conversion (not just above > > one, completely from this patch series). Or uou can take a call just > > keep this macro - > > _MIPS_SIM_ABI64 as it is and change it rest of the place. > > Let's just go ahead & leave everything as 64, and I'll do the 64 -> n64 > rename later. I hoped whilst you were adding n64-specific code this > would be an easy change, but at this point let's just prioritize getting > the series applied without the naming change so it can sit in -next for > a while before the merge window. I'll keep the macro - _MIPS_SIM_ABI64 same and will not change the rest of the patch series. Unfortunately, reverting back to 64 from n64 has lots of work. Thanks Firoz
Hi Firoz, On Thu, Dec 13, 2018 at 02:23:21PM +0530, Firoz Khan wrote: > > > So If you confirm I can send v5 without '64' to 'n64' conversion (not just above > > > one, completely from this patch series). Or uou can take a call just > > > keep this macro - > > > _MIPS_SIM_ABI64 as it is and change it rest of the place. > > > > Let's just go ahead & leave everything as 64, and I'll do the 64 -> n64 > > rename later. I hoped whilst you were adding n64-specific code this > > would be an easy change, but at this point let's just prioritize getting > > the series applied without the naming change so it can sit in -next for > > a while before the merge window. > > I'll keep the macro - _MIPS_SIM_ABI64 same and will not change the rest > of the patch series. Unfortunately, reverting back to 64 from n64 has lots of > work. I've applied v5 but undone the change from __NR_64_* to __NR_N64_* because it's part of the UAPI & a github code search showed that it's actually used. Could you take a look at this branch & check that you're OK with it before I push it to mips-next? git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux.git test-syscalls https://git.kernel.org/pub/scm/linux/kernel/git/mips/linux.git/log/?h=test-syscalls Thanks, Paul
Hi Paul, On Fri, 14 Dec 2018 at 01:45, Paul Burton <paul.burton@mips.com> wrote: > I've applied v5 but undone the change from __NR_64_* to __NR_N64_* > because it's part of the UAPI & a github code search showed that it's > actually used. > > Could you take a look at this branch & check that you're OK with it > before I push it to mips-next? > > git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux.git test-syscalls > > https://git.kernel.org/pub/scm/linux/kernel/git/mips/linux.git/log/?h=test-syscalls This looks good to me. Please push to mips-next. Thanks Firoz
diff --git a/arch/mips/include/asm/asm.h b/arch/mips/include/asm/asm.h index c23527b..9244a87 100644 --- a/arch/mips/include/asm/asm.h +++ b/arch/mips/include/asm/asm.h @@ -110,7 +110,7 @@ #define ALSZ 7 #define ALMASK ~7 #endif -#if (_MIPS_SIM == _MIPS_SIM_NABI32) || (_MIPS_SIM == _MIPS_SIM_ABI64) +#if (_MIPS_SIM == _MIPS_SIM_NABI32) || (_MIPS_SIM == _MIPS_SIM_ABIN64) #define ALSZ 15 #define ALMASK ~15 #endif @@ -138,7 +138,7 @@ #define REG_SUBU subu #define REG_ADDU addu #endif -#if (_MIPS_SIM == _MIPS_SIM_NABI32) || (_MIPS_SIM == _MIPS_SIM_ABI64) +#if (_MIPS_SIM == _MIPS_SIM_NABI32) || (_MIPS_SIM == _MIPS_SIM_ABIN64) #define REG_S sd #define REG_L ld #define REG_SUBU dsubu @@ -291,7 +291,7 @@ #define MFC0 mfc0 #define MTC0 mtc0 #endif -#if (_MIPS_SIM == _MIPS_SIM_NABI32) || (_MIPS_SIM == _MIPS_SIM_ABI64) +#if (_MIPS_SIM == _MIPS_SIM_NABI32) || (_MIPS_SIM == _MIPS_SIM_ABIN64) #define MFC0 dmfc0 #define MTC0 dmtc0 #endif diff --git a/arch/mips/include/asm/fpregdef.h b/arch/mips/include/asm/fpregdef.h index f184ba0..762ffeb 100644 --- a/arch/mips/include/asm/fpregdef.h +++ b/arch/mips/include/asm/fpregdef.h @@ -71,7 +71,7 @@ #endif /* _MIPS_SIM == _MIPS_SIM_ABI32 */ -#if _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 #define fv0 $f0 /* return value */ #define fv1 $f2 @@ -108,6 +108,6 @@ #define fcr31 $31 -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ #endif /* _ASM_FPREGDEF_H */ diff --git a/arch/mips/include/asm/fw/arc/hinv.h b/arch/mips/include/asm/fw/arc/hinv.h index d67b6a9..c9d0523 100644 --- a/arch/mips/include/asm/fw/arc/hinv.h +++ b/arch/mips/include/asm/fw/arc/hinv.h @@ -112,7 +112,7 @@ ULONG FullKey; }; -#if _MIPS_SIM == _MIPS_SIM_ABI64 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 #define SGI_ARCS_VERS 64 /* sgi 64-bit version */ #define SGI_ARCS_REV 0 /* rev .00 */ #else diff --git a/arch/mips/include/asm/regdef.h b/arch/mips/include/asm/regdef.h index 3c687df..45911d8 100644 --- a/arch/mips/include/asm/regdef.h +++ b/arch/mips/include/asm/regdef.h @@ -60,7 +60,7 @@ #endif /* _MIPS_SIM == _MIPS_SIM_ABI32 */ -#if _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 #define zero $0 /* wired zero */ #define AT $at /* assembler temp - uppercase because of ".set at" */ @@ -101,6 +101,6 @@ #define s8 $30 /* callee saved */ #define ra $31 /* return address */ -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ #endif /* _ASM_REGDEF_H */ diff --git a/arch/mips/include/asm/sigcontext.h b/arch/mips/include/asm/sigcontext.h index eeeb0f4..e3a55d1 100644 --- a/arch/mips/include/asm/sigcontext.h +++ b/arch/mips/include/asm/sigcontext.h @@ -11,7 +11,7 @@ #include <uapi/asm/sigcontext.h> -#if _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 struct sigcontext32 { __u32 sc_regmask; /* Unused */ @@ -33,5 +33,5 @@ struct sigcontext32 { __u32 sc_hi3; __u32 sc_lo3; }; -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ #endif /* _ASM_SIGCONTEXT_H */ diff --git a/arch/mips/include/uapi/asm/fcntl.h b/arch/mips/include/uapi/asm/fcntl.h index 42e13de..34effe6 100644 --- a/arch/mips/include/uapi/asm/fcntl.h +++ b/arch/mips/include/uapi/asm/fcntl.h @@ -57,7 +57,7 @@ * contain all the same fields as struct flock. */ -#if _MIPS_SIM != _MIPS_SIM_ABI64 +#if _MIPS_SIM != _MIPS_SIM_ABIN64 #include <linux/types.h> diff --git a/arch/mips/include/uapi/asm/reg.h b/arch/mips/include/uapi/asm/reg.h index 56d15cb..57fe259 100644 --- a/arch/mips/include/uapi/asm/reg.h +++ b/arch/mips/include/uapi/asm/reg.h @@ -160,7 +160,7 @@ #define EF_UNUSED0 MIPS32_EF_UNUSED0 #define EF_SIZE MIPS32_EF_SIZE -#elif _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 +#elif _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 #define EF_R0 MIPS64_EF_R0 #define EF_R1 MIPS64_EF_R1 @@ -202,6 +202,6 @@ #define EF_CP0_CAUSE MIPS64_EF_CP0_CAUSE #define EF_SIZE MIPS64_EF_SIZE -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ #endif /* __UAPI_ASM_MIPS_REG_H */ diff --git a/arch/mips/include/uapi/asm/sgidefs.h b/arch/mips/include/uapi/asm/sgidefs.h index 26143e3..0364eec 100644 --- a/arch/mips/include/uapi/asm/sgidefs.h +++ b/arch/mips/include/uapi/asm/sgidefs.h @@ -40,6 +40,6 @@ */ #define _MIPS_SIM_ABI32 1 #define _MIPS_SIM_NABI32 2 -#define _MIPS_SIM_ABI64 3 +#define _MIPS_SIM_ABIN64 3 #endif /* __ASM_SGIDEFS_H */ diff --git a/arch/mips/include/uapi/asm/sigcontext.h b/arch/mips/include/uapi/asm/sigcontext.h index d0a540e..5e8a562 100644 --- a/arch/mips/include/uapi/asm/sigcontext.h +++ b/arch/mips/include/uapi/asm/sigcontext.h @@ -54,7 +54,7 @@ struct sigcontext { #endif /* _MIPS_SIM == _MIPS_SIM_ABI32 */ -#if _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 #include <linux/posix_types.h> /* @@ -86,6 +86,6 @@ struct sigcontext { }; -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ #endif /* _UAPI_ASM_SIGCONTEXT_H */ diff --git a/arch/mips/include/uapi/asm/stat.h b/arch/mips/include/uapi/asm/stat.h index 95416f3..c1f6433 100644 --- a/arch/mips/include/uapi/asm/stat.h +++ b/arch/mips/include/uapi/asm/stat.h @@ -87,7 +87,7 @@ struct stat64 { #endif /* _MIPS_SIM == _MIPS_SIM_ABI32 */ -#if _MIPS_SIM == _MIPS_SIM_ABI64 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 /* The memory layout is the same as of struct stat64 of the 32-bit kernel. */ struct stat { @@ -126,7 +126,7 @@ struct stat { unsigned long st_blocks; }; -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 */ #define STAT_HAVE_NSEC 1 diff --git a/arch/mips/include/uapi/asm/statfs.h b/arch/mips/include/uapi/asm/statfs.h index f4174dc..a9ca452 100644 --- a/arch/mips/include/uapi/asm/statfs.h +++ b/arch/mips/include/uapi/asm/statfs.h @@ -61,7 +61,7 @@ struct statfs64 { #endif /* _MIPS_SIM == _MIPS_SIM_ABI32 */ -#if _MIPS_SIM == _MIPS_SIM_ABI64 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 struct statfs64 { /* Same as struct statfs */ long f_type; @@ -96,6 +96,6 @@ struct compat_statfs64 { __u32 f_spare[5]; }; -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 */ #endif /* _ASM_STATFS_H */ diff --git a/arch/mips/include/uapi/asm/unistd.h b/arch/mips/include/uapi/asm/unistd.h index c897195..b997f43 100644 --- a/arch/mips/include/uapi/asm/unistd.h +++ b/arch/mips/include/uapi/asm/unistd.h @@ -400,7 +400,7 @@ #define __NR_O32_Linux 4000 #define __NR_O32_Linux_syscalls 368 -#if _MIPS_SIM == _MIPS_SIM_ABI64 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 /* * Linux 64-bit syscalls are in the range from 5000 to 5999. @@ -740,10 +740,10 @@ #define __NR_syscalls 328 #endif -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 */ -#define __NR_64_Linux 5000 -#define __NR_64_Linux_syscalls 328 +#define __NR_N64_Linux 5000 +#define __NR_N64_Linux_syscalls 328 #if _MIPS_SIM == _MIPS_SIM_NABI32 diff --git a/arch/mips/kernel/Makefile b/arch/mips/kernel/Makefile index 210c280..25af613 100644 --- a/arch/mips/kernel/Makefile +++ b/arch/mips/kernel/Makefile @@ -72,7 +72,7 @@ obj-$(CONFIG_IRQ_GT641XX) += irq-gt641xx.o obj-$(CONFIG_KPROBES) += kprobes.o obj-$(CONFIG_32BIT) += scall32-o32.o -obj-$(CONFIG_64BIT) += scall64-64.o +obj-$(CONFIG_64BIT) += scall64-n64.o obj-$(CONFIG_MIPS32_COMPAT) += linux32.o ptrace32.o signal32.o obj-$(CONFIG_MIPS32_N32) += binfmt_elfn32.o scall64-n32.o signal_n32.o obj-$(CONFIG_MIPS32_O32) += binfmt_elfo32.o scall64-o32.o signal_o32.o diff --git a/arch/mips/kernel/ftrace.c b/arch/mips/kernel/ftrace.c index b122cbb..d91a6e7 100644 --- a/arch/mips/kernel/ftrace.c +++ b/arch/mips/kernel/ftrace.c @@ -403,8 +403,8 @@ unsigned long __init arch_syscall_addr(int nr) if (nr >= __NR_N32_Linux && nr <= __NR_N32_Linux + __NR_N32_Linux_syscalls) return (unsigned long)sysn32_call_table[nr - __NR_N32_Linux]; #endif - if (nr >= __NR_64_Linux && nr <= __NR_64_Linux + __NR_64_Linux_syscalls) - return (unsigned long)sys_call_table[nr - __NR_64_Linux]; + if (nr >= __NR_N64_Linux && nr <= __NR_N64_Linux + __NR_N64_Linux_syscalls) + return (unsigned long)sys_call_table[nr - __NR_N64_Linux]; #ifdef CONFIG_MIPS32_O32 if (nr >= __NR_O32_Linux && nr <= __NR_O32_Linux + __NR_O32_Linux_syscalls) return (unsigned long)sys32_call_table[nr - __NR_O32_Linux]; diff --git a/arch/mips/kernel/scall64-64.S b/arch/mips/kernel/scall64-n64.S similarity index 99% rename from arch/mips/kernel/scall64-64.S rename to arch/mips/kernel/scall64-n64.S index 358d959..497cd62 100644 --- a/arch/mips/kernel/scall64-64.S +++ b/arch/mips/kernel/scall64-n64.S @@ -53,8 +53,8 @@ NESTED(handle_sys64, PT_SIZE, sp) bnez t0, syscall_trace_entry syscall_common: - dsubu t2, v0, __NR_64_Linux - sltiu t0, t2, __NR_64_Linux_syscalls + 1 + dsubu t2, v0, __NR_N64_Linux + sltiu t0, t2, __NR_N64_Linux_syscalls + 1 beqz t0, illegal_syscall dsll t0, t2, 3 # offset into table diff --git a/arch/mips/kvm/entry.c b/arch/mips/kvm/entry.c index 16e1c93..d30f42c 100644 --- a/arch/mips/kvm/entry.c +++ b/arch/mips/kvm/entry.c @@ -34,12 +34,12 @@ #define T3 11 #endif /* _MIPS_SIM == _MIPS_SIM_ABI32 */ -#if _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 #define T0 12 #define T1 13 #define T2 14 #define T3 15 -#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ +#endif /* _MIPS_SIM == _MIPS_SIM_ABIN64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ #define S0 16 #define S1 17 diff --git a/arch/mips/vdso/vdso.h b/arch/mips/vdso/vdso.h index cfb1be4..1fd1692 100644 --- a/arch/mips/vdso/vdso.h +++ b/arch/mips/vdso/vdso.h @@ -10,7 +10,7 @@ #include <asm/sgidefs.h> -#if _MIPS_SIM != _MIPS_SIM_ABI64 && defined(CONFIG_64BIT) +#if _MIPS_SIM != _MIPS_SIM_ABIN64 && defined(CONFIG_64BIT) /* Building 32-bit VDSO for the 64-bit kernel. Fake a 32-bit Kconfig. */ #undef CONFIG_64BIT diff --git a/arch/mips/vdso/vdso.lds.S b/arch/mips/vdso/vdso.lds.S index 8df7dd5..4d049e9 100644 --- a/arch/mips/vdso/vdso.lds.S +++ b/arch/mips/vdso/vdso.lds.S @@ -10,7 +10,7 @@ #include <asm/sgidefs.h> -#if _MIPS_SIM == _MIPS_SIM_ABI64 +#if _MIPS_SIM == _MIPS_SIM_ABIN64 OUTPUT_FORMAT("elf64-tradlittlemips", "elf64-tradbigmips", "elf64-tradlittlemips") #elif _MIPS_SIM == _MIPS_SIM_NABI32 OUTPUT_FORMAT("elf32-ntradlittlemips", "elf32-ntradbigmips", "elf32-ntradlittlemips")
When we get nanoMIPS support we'll be introducing the p32 ABI, and there's a reasonable chance that the equivalent p64 ABI may come along in the future. Using 'n64' now would avoid confusion in that case where we may have 2 different 64-bit ABIs. Suggested-by: Paul Burton <paul.burton@mips.com> Signed-off-by: Firoz Khan <firoz.khan@linaro.org> --- arch/mips/include/asm/asm.h | 6 +++--- arch/mips/include/asm/fpregdef.h | 4 ++-- arch/mips/include/asm/fw/arc/hinv.h | 2 +- arch/mips/include/asm/regdef.h | 4 ++-- arch/mips/include/asm/sigcontext.h | 4 ++-- arch/mips/include/uapi/asm/fcntl.h | 2 +- arch/mips/include/uapi/asm/reg.h | 4 ++-- arch/mips/include/uapi/asm/sgidefs.h | 2 +- arch/mips/include/uapi/asm/sigcontext.h | 4 ++-- arch/mips/include/uapi/asm/stat.h | 4 ++-- arch/mips/include/uapi/asm/statfs.h | 4 ++-- arch/mips/include/uapi/asm/unistd.h | 8 ++++---- arch/mips/kernel/Makefile | 2 +- arch/mips/kernel/ftrace.c | 4 ++-- arch/mips/kernel/{scall64-64.S => scall64-n64.S} | 4 ++-- arch/mips/kvm/entry.c | 4 ++-- arch/mips/vdso/vdso.h | 2 +- arch/mips/vdso/vdso.lds.S | 2 +- 18 files changed, 33 insertions(+), 33 deletions(-) rename arch/mips/kernel/{scall64-64.S => scall64-n64.S} (99%)