| Message ID | 1538288720-14910-2-git-send-email-bmeng.cn@gmail.com (mailing list archive) |
|---|---|
| State | Rejected |
| Headers | show |
| Series | [1/2] powerpc/64: Remove duplicated -mabi=elfv2 for little endian targets | expand |
On Sat, 29 Sep 2018 23:25:20 -0700 Bin Meng <bmeng.cn@gmail.com> wrote: > commit 573ebfa6601f ("powerpc: Increase stack redzone for 64-bit > userspace to 512 bytes") only changes stack userspace redzone size. > We need increase the kernel one to 512 bytes too per ABIv2 spec. You're right we need 512 to be compatible with ABIv2, but as the comment says, gcc limits this to 288 bytes so that's what is used to save stack space. We can use a compiler version test to change this if llvm or a new version of gcc does something different. Thanks, Nick > > Signed-off-by: Bin Meng <bmeng.cn@gmail.com> > --- > > arch/powerpc/include/asm/ptrace.h | 8 ++------ > 1 file changed, 2 insertions(+), 6 deletions(-) > > diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h > index 447cbd1..817be3f 100644 > --- a/arch/powerpc/include/asm/ptrace.h > +++ b/arch/powerpc/include/asm/ptrace.h > @@ -30,16 +30,12 @@ > #ifdef __powerpc64__ > > /* > - * Size of redzone that userspace is allowed to use below the stack > + * Size of redzone that kernel/userspace is allowed to use below the stack > * pointer. This is 288 in the 64-bit big-endian ELF ABI, and 512 in > * the new ELFv2 little-endian ABI, so we allow the larger amount. > - * > - * For kernel code we allow a 288-byte redzone, in order to conserve > - * kernel stack space; gcc currently only uses 288 bytes, and will > - * hopefully allow explicit control of the redzone size in future. > */ > #define USER_REDZONE_SIZE 512 > -#define KERNEL_REDZONE_SIZE 288 > +#define KERNEL_REDZONE_SIZE 512 > > #define STACK_FRAME_OVERHEAD 112 /* size of minimum stack frame */ > #define STACK_FRAME_LR_SAVE 2 /* Location of LR in stack frame */ > -- > 2.7.4 >
Hi Nick, On Mon, Oct 1, 2018 at 7:27 AM Nicholas Piggin <npiggin@gmail.com> wrote: > > On Sat, 29 Sep 2018 23:25:20 -0700 > Bin Meng <bmeng.cn@gmail.com> wrote: > > > commit 573ebfa6601f ("powerpc: Increase stack redzone for 64-bit > > userspace to 512 bytes") only changes stack userspace redzone size. > > We need increase the kernel one to 512 bytes too per ABIv2 spec. > > You're right we need 512 to be compatible with ABIv2, but as the > comment says, gcc limits this to 288 bytes so that's what is used > to save stack space. We can use a compiler version test to change > this if llvm or a new version of gcc does something different. > I believe what the comment says is for ABIv1. At the time when commit 573ebfa6601f was submitted, kernel had not switched to ABIv2 build yet. Regards, Bin
On Mon, 1 Oct 2018 09:11:04 +0800 Bin Meng <bmeng.cn@gmail.com> wrote: > Hi Nick, > > On Mon, Oct 1, 2018 at 7:27 AM Nicholas Piggin <npiggin@gmail.com> wrote: > > > > On Sat, 29 Sep 2018 23:25:20 -0700 > > Bin Meng <bmeng.cn@gmail.com> wrote: > > > > > commit 573ebfa6601f ("powerpc: Increase stack redzone for 64-bit > > > userspace to 512 bytes") only changes stack userspace redzone size. > > > We need increase the kernel one to 512 bytes too per ABIv2 spec. > > > > You're right we need 512 to be compatible with ABIv2, but as the > > comment says, gcc limits this to 288 bytes so that's what is used > > to save stack space. We can use a compiler version test to change > > this if llvm or a new version of gcc does something different. > > > > I believe what the comment says is for ABIv1. At the time when commit > 573ebfa6601f was submitted, kernel had not switched to ABIv2 build > yet. I see, yes you are right about that. However gcc still seems to be using 288 bytes. static inline bool offset_below_red_zone_p (HOST_WIDE_INT offset) { return offset < (DEFAULT_ABI == ABI_V4 ? 0 : TARGET_32BIT ? -220 : -288); } llvm does as well AFAIKS // DarwinABI has a 224-byte red zone. PPC32 SVR4ABI(Non-DarwinABI) has no // red zone and PPC64 SVR4ABI has a 288-byte red zone. unsigned getRedZoneSize() const { return isDarwinABI() ? 224 : (isPPC64() ? 288 : 0); } So I suspect we can get away with using 288 for the kernel. Although the ELFv2 ABI allows 512, I suspect at this point compilers won't switch over without an explicit red zone size flag. Thanks, Nick
Hi! On Mon, Oct 01, 2018 at 12:22:56PM +1000, Nicholas Piggin wrote: > On Mon, 1 Oct 2018 09:11:04 +0800 > Bin Meng <bmeng.cn@gmail.com> wrote: > > On Mon, Oct 1, 2018 at 7:27 AM Nicholas Piggin <npiggin@gmail.com> wrote: > > > On Sat, 29 Sep 2018 23:25:20 -0700 > > > Bin Meng <bmeng.cn@gmail.com> wrote: > > > > commit 573ebfa6601f ("powerpc: Increase stack redzone for 64-bit > > > > userspace to 512 bytes") only changes stack userspace redzone size. > > > > We need increase the kernel one to 512 bytes too per ABIv2 spec. > > > > > > You're right we need 512 to be compatible with ABIv2, but as the > > > comment says, gcc limits this to 288 bytes so that's what is used > > > to save stack space. We can use a compiler version test to change > > > this if llvm or a new version of gcc does something different. > > > > > > > I believe what the comment says is for ABIv1. At the time when commit > > 573ebfa6601f was submitted, kernel had not switched to ABIv2 build > > yet. > > I see, yes you are right about that. However gcc still seems to be using > 288 bytes. And that is required by the ABI! """ 2.2.2.4. Protected Zone The 288 bytes below the stack pointer are available as volatile program storage that is not preserved across function calls. Interrupt handlers and any other functions that might run without an explicit call must take care to preserve a protected zone, also referred to as the red zone, of 512 bytes that consists of: * The 288-byte volatile program storage region that is used to hold saved registers and local variables * An additional 224 bytes below the volatile program storage region that is set aside as a volatile system storage region for system functions If a function does not call other functions and does not need more stack space than is available in the volatile program storage region (that is, 288 bytes), it does not need to have a stack frame. The 224-byte volatile system storage region is not available to compilers for allocation to saved registers and local variables. """ A routine has a red zone of 288 bytes. Below there is 224 more bytes of available storage, but that is not available to the routine itself: some (asynchronous) other code (like an interrupt) can use (i.e. clobber) it. Segher
On Sat, Sep 29, 2018 at 11:25:20PM -0700, Bin Meng wrote: > /* > - * Size of redzone that userspace is allowed to use below the stack > + * Size of redzone that kernel/userspace is allowed to use below the stack > * pointer. This is 288 in the 64-bit big-endian ELF ABI, and 512 in > * the new ELFv2 little-endian ABI, so we allow the larger amount. > - * > - * For kernel code we allow a 288-byte redzone, in order to conserve > - * kernel stack space; gcc currently only uses 288 bytes, and will > - * hopefully allow explicit control of the redzone size in future. > */ Btw: patches welcome! This will never be useful for userland code, so no one in GCC land is looking at this (we did not even know it is wanted). Segher
Hi Nick, On Mon, Oct 1, 2018 at 10:23 AM Nicholas Piggin <npiggin@gmail.com> wrote: > > On Mon, 1 Oct 2018 09:11:04 +0800 > Bin Meng <bmeng.cn@gmail.com> wrote: > > > Hi Nick, > > > > On Mon, Oct 1, 2018 at 7:27 AM Nicholas Piggin <npiggin@gmail.com> wrote: > > > > > > On Sat, 29 Sep 2018 23:25:20 -0700 > > > Bin Meng <bmeng.cn@gmail.com> wrote: > > > > > > > commit 573ebfa6601f ("powerpc: Increase stack redzone for 64-bit > > > > userspace to 512 bytes") only changes stack userspace redzone size. > > > > We need increase the kernel one to 512 bytes too per ABIv2 spec. > > > > > > You're right we need 512 to be compatible with ABIv2, but as the > > > comment says, gcc limits this to 288 bytes so that's what is used > > > to save stack space. We can use a compiler version test to change > > > this if llvm or a new version of gcc does something different. > > > > > > > I believe what the comment says is for ABIv1. At the time when commit > > 573ebfa6601f was submitted, kernel had not switched to ABIv2 build > > yet. > > I see, yes you are right about that. However gcc still seems to be using > 288 bytes. > > static inline bool > offset_below_red_zone_p (HOST_WIDE_INT offset) > { > return offset < (DEFAULT_ABI == ABI_V4 > ? 0 > : TARGET_32BIT ? -220 : -288); > } > > llvm does as well AFAIKS > > // DarwinABI has a 224-byte red zone. PPC32 SVR4ABI(Non-DarwinABI) has no > // red zone and PPC64 SVR4ABI has a 288-byte red zone. > unsigned getRedZoneSize() const { > return isDarwinABI() ? 224 : (isPPC64() ? 288 : 0); > } > > So I suspect we can get away with using 288 for the kernel. Although > the ELFv2 ABI allows 512, I suspect at this point compilers won't switch > over without an explicit red zone size flag. > Thanks for the info of gcc/llvm codes. I suspect for the red zone size gcc/llvm still uses ABIv1 defined value which is 288. If we get way with kernel using 288, what's the point of having user as 512 (commit 573ebfa6601f)? Regards, Bin
On Mon, 1 Oct 2018 03:51:21 -0500 Segher Boessenkool <segher@kernel.crashing.org> wrote: > Hi! > > On Mon, Oct 01, 2018 at 12:22:56PM +1000, Nicholas Piggin wrote: > > On Mon, 1 Oct 2018 09:11:04 +0800 > > Bin Meng <bmeng.cn@gmail.com> wrote: > > > On Mon, Oct 1, 2018 at 7:27 AM Nicholas Piggin <npiggin@gmail.com> wrote: > > > > On Sat, 29 Sep 2018 23:25:20 -0700 > > > > Bin Meng <bmeng.cn@gmail.com> wrote: > > > > > commit 573ebfa6601f ("powerpc: Increase stack redzone for 64-bit > > > > > userspace to 512 bytes") only changes stack userspace redzone size. > > > > > We need increase the kernel one to 512 bytes too per ABIv2 spec. > > > > > > > > You're right we need 512 to be compatible with ABIv2, but as the > > > > comment says, gcc limits this to 288 bytes so that's what is used > > > > to save stack space. We can use a compiler version test to change > > > > this if llvm or a new version of gcc does something different. > > > > > > > > > > I believe what the comment says is for ABIv1. At the time when commit > > > 573ebfa6601f was submitted, kernel had not switched to ABIv2 build > > > yet. > > > > I see, yes you are right about that. However gcc still seems to be using > > 288 bytes. > > And that is required by the ABI! > > """ > 2.2.2.4. Protected Zone > > The 288 bytes below the stack pointer are available as volatile program > storage that is not preserved across function calls. Interrupt handlers and > any other functions that might run without an explicit call must take care > to preserve a protected zone, also referred to as the red zone, of 512 bytes > that consists of: > > * The 288-byte volatile program storage region that is used to hold saved > registers and local variables > * An additional 224 bytes below the volatile program storage region that is > set aside as a volatile system storage region for system functions > > If a function does not call other functions and does not need more stack > space than is available in the volatile program storage region (that is, 288 > bytes), it does not need to have a stack frame. The 224-byte volatile system > storage region is not available to compilers for allocation to saved > registers and local variables. > """ > > A routine has a red zone of 288 bytes. Below there is 224 more bytes of > available storage, but that is not available to the routine itself: some > (asynchronous) other code (like an interrupt) can use (i.e. clobber) it. Thanks Segher, that explains it very well and shows we are safe with 288 in the kernel. So we can leave the code as-is, but the comment could be updated. What are "system functions" exactly? Can the kernel use that, or are we talking about user mode system code like libraries? The kernel could maybe use that for scratch space for synchronous interrupts to avoid using a slow SPR for scratch. Thanks, Nick > > > Segher
On Mon, 1 Oct 2018 20:41:19 +0800 Bin Meng <bmeng.cn@gmail.com> wrote: > Hi Nick, > > On Mon, Oct 1, 2018 at 10:23 AM Nicholas Piggin <npiggin@gmail.com> wrote: > > > > On Mon, 1 Oct 2018 09:11:04 +0800 > > Bin Meng <bmeng.cn@gmail.com> wrote: > > > > > Hi Nick, > > > > > > On Mon, Oct 1, 2018 at 7:27 AM Nicholas Piggin <npiggin@gmail.com> wrote: > > > > > > > > On Sat, 29 Sep 2018 23:25:20 -0700 > > > > Bin Meng <bmeng.cn@gmail.com> wrote: > > > > > > > > > commit 573ebfa6601f ("powerpc: Increase stack redzone for 64-bit > > > > > userspace to 512 bytes") only changes stack userspace redzone size. > > > > > We need increase the kernel one to 512 bytes too per ABIv2 spec. > > > > > > > > You're right we need 512 to be compatible with ABIv2, but as the > > > > comment says, gcc limits this to 288 bytes so that's what is used > > > > to save stack space. We can use a compiler version test to change > > > > this if llvm or a new version of gcc does something different. > > > > > > > > > > I believe what the comment says is for ABIv1. At the time when commit > > > 573ebfa6601f was submitted, kernel had not switched to ABIv2 build > > > yet. > > > > I see, yes you are right about that. However gcc still seems to be using > > 288 bytes. > > > > static inline bool > > offset_below_red_zone_p (HOST_WIDE_INT offset) > > { > > return offset < (DEFAULT_ABI == ABI_V4 > > ? 0 > > : TARGET_32BIT ? -220 : -288); > > } > > > > llvm does as well AFAIKS > > > > // DarwinABI has a 224-byte red zone. PPC32 SVR4ABI(Non-DarwinABI) has no > > // red zone and PPC64 SVR4ABI has a 288-byte red zone. > > unsigned getRedZoneSize() const { > > return isDarwinABI() ? 224 : (isPPC64() ? 288 : 0); > > } > > > > So I suspect we can get away with using 288 for the kernel. Although > > the ELFv2 ABI allows 512, I suspect at this point compilers won't switch > > over without an explicit red zone size flag. > > > > Thanks for the info of gcc/llvm codes. I suspect for the red zone size > gcc/llvm still uses ABIv1 defined value which is 288. If we get way > with kernel using 288, what's the point of having user as 512 (commit > 573ebfa6601f)? See Segher's reply -- they are two different things here. 288 bytes is the red zone that compilers may use. But there is another region out to 512 bytes which can be used by other system code (not compilers). So the kernel always has to assume 512 for user. The kernel code itself knows that it does not use any red zone beyond 288 bytes, so it does not have to preserve any more for its own stack. Thanks, Nick
On Tue, Oct 02, 2018 at 09:59:29AM +1000, Nicholas Piggin wrote: > On Mon, 1 Oct 2018 03:51:21 -0500 > Segher Boessenkool <segher@kernel.crashing.org> wrote: > > And that is required by the ABI! > > > > """ > > 2.2.2.4. Protected Zone > > > > The 288 bytes below the stack pointer are available as volatile program > > storage that is not preserved across function calls. Interrupt handlers and > > any other functions that might run without an explicit call must take care > > to preserve a protected zone, also referred to as the red zone, of 512 bytes > > that consists of: > > > > * The 288-byte volatile program storage region that is used to hold saved > > registers and local variables > > * An additional 224 bytes below the volatile program storage region that is > > set aside as a volatile system storage region for system functions > > > > If a function does not call other functions and does not need more stack > > space than is available in the volatile program storage region (that is, 288 > > bytes), it does not need to have a stack frame. The 224-byte volatile system > > storage region is not available to compilers for allocation to saved > > registers and local variables. > > """ > > > > A routine has a red zone of 288 bytes. Below there is 224 more bytes of > > available storage, but that is not available to the routine itself: some > > (asynchronous) other code (like an interrupt) can use (i.e. clobber) it. > > Thanks Segher, that explains it very well and shows we are safe with > 288 in the kernel. So we can leave the code as-is, but the comment > could be updated. > > What are "system functions" exactly? That is an excellent question. I think it was left vague on purpose? "Stuff a user program cannot do itself", "ABI stuff", "whatever the OS defines as system stuff"? > Can the kernel use that, or are > we talking about user mode system code like libraries? The kernel > could maybe use that for scratch space for synchronous interrupts to > avoid using a slow SPR for scratch. If you're already using the kernel stack, sure. When does this happen? Segher
diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h index 447cbd1..817be3f 100644 --- a/arch/powerpc/include/asm/ptrace.h +++ b/arch/powerpc/include/asm/ptrace.h @@ -30,16 +30,12 @@ #ifdef __powerpc64__ /* - * Size of redzone that userspace is allowed to use below the stack + * Size of redzone that kernel/userspace is allowed to use below the stack * pointer. This is 288 in the 64-bit big-endian ELF ABI, and 512 in * the new ELFv2 little-endian ABI, so we allow the larger amount. - * - * For kernel code we allow a 288-byte redzone, in order to conserve - * kernel stack space; gcc currently only uses 288 bytes, and will - * hopefully allow explicit control of the redzone size in future. */ #define USER_REDZONE_SIZE 512 -#define KERNEL_REDZONE_SIZE 288 +#define KERNEL_REDZONE_SIZE 512 #define STACK_FRAME_OVERHEAD 112 /* size of minimum stack frame */ #define STACK_FRAME_LR_SAVE 2 /* Location of LR in stack frame */
commit 573ebfa6601f ("powerpc: Increase stack redzone for 64-bit userspace to 512 bytes") only changes stack userspace redzone size. We need increase the kernel one to 512 bytes too per ABIv2 spec. Signed-off-by: Bin Meng <bmeng.cn@gmail.com> --- arch/powerpc/include/asm/ptrace.h | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-)