src/powerpc


Log

Author Commit Date CI Message
Clément Chigot 80912f2f 2022-09-01T14:47:16 Makefile: enable target overridings. (#661) This patch allows target to provide extra files enabling the override of Makefile rules. This patch is not needed for libffi itself but only for GCC on AIX. The t-aix file which is here empty will be replaced in GCC repository. We cannot include GCC version directly here because it has no sense for a standalone libffi. Co-authored-by: Clement <clement.chigot@atos.net>
Sergei Trofimovich 000ecd01 2022-06-02T18:43:53 powerpc64: fix build on gcc-12 (assembly requires AltiVec) (#719) Without the change build fails on powerpc64-gcc-12 as: src/powerpc/linux64_closure.S: Assembler messages: src/powerpc/linux64_closure.S:363: Error: unrecognized opcode: `lvx' It's a 90205f67 "rs6000: Fix bootstrap (libffi)" patch by Segher Boessenkool from gcc upstream repository. It's enough to get libffi build on powerpc64.
Dan Horák 464b4b66 2022-01-31T14:08:09 powerpc64: fix handling of homogeneous float128 structs (#689) If there is a homogeneous struct with float128 members, they should be copied to vector register save area. The current code incorrectly copies only the value of the first member, not increasing the pointer with each iteration. Fix this.
Ken 484c1a0d 2020-10-27T07:07:07 darwin powerpc: use go closures only if enabled (#583) SHA: 0ff9419f2e75652426469e256cb7d0748064ad58 added go closures for darwin powerpc AIX, but these have not yet been implemented for non-AIX systems use the go closures only if enabled
Alan Modra f4435980 2020-10-27T23:36:21 Power10 libffi fixes (#585) Power10 pc-relative code doesn't use or preserve r2 as a TOC pointer. That means calling between pc-relative and TOC using code can't be done without intervening linker stubs, and a call from TOC code to pc-relative code must have a nop after the bl in order to restore r2. Now the PowerPC libffi assembly code doesn't use r2 except for the implicit use when making calls back to C, ffi_closure_helper_LINUX64 and ffi_prep_args64. So changing the assembly to interoperate with pc-relative code without stubs is easily done. * src/powerpc/linux64.S (ffi_call_LINUX64): Don't emit global entry when __PCREL__. Call using @notoc. Add nops. * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Likewise. (ffi_go_closure_linux64): Likewise.
Andrew Geissler 4f9e20ac 2020-05-01T06:58:30 ffi_powerpc.h: fix build failure with powerpc7 (#561) This is a patch pulled down from the following: https://github.com/buildroot/buildroot/blob/78926f610b1411b03464152472fd430012deb9ac/package/libffi/0004-ffi_powerpc.h-fix-build-failure-with-powerpc7.patch This issue is being hit on OpenBMC code when pulling the latest libffi tag and building on a P8 ppc64le machine. I verified this patch fixes the issue we are seeing. Below is the original commit message: Sicne commit 73dd43afc8a447ba98ea02e9aad4c6898dc77fb0, build on powerpc7 fails on: In file included from ../src/powerpc/ffi.c:33:0: ../src/powerpc/ffi_powerpc.h:61:9: error: '_Float128' is not supported on this target typedef _Float128 float128; ^~~~~~~~~ Fix this build failure by checking for __HAVE_FLOAT128 before using _Float128, as _Float128 is enabled only on specific conditions, see output/host/powerpc64-buildroot-linux-gnu/sysroot/usr/include/bits/floatn.h: /* Defined to 1 if the current compiler invocation provides a floating-point type with the IEEE 754 binary128 format, and this glibc includes corresponding *f128 interfaces for it. */ #if defined _ARCH_PWR8 && defined __LITTLE_ENDIAN__ && (_CALL_ELF == 2) \ && defined __FLOAT128__ && !defined __NO_LONG_DOUBLE_MATH # define __HAVE_FLOAT128 1 #else # define __HAVE_FLOAT128 0 #endif Fixes: - http://autobuild.buildroot.org/results/5c9dd8fb3b6a128882b6250f197c80232d8a3b53 Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com> Signed-off-by: Andrew Geissler <geissonator@yahoo.com> Co-authored-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Samuel Holland 4d6d2866 2020-02-21T21:06:15 Update powerpc sysv assembly for ffi_powerpc.h changes (#541) Some of the flag bits were moved when adding powerpc64 vector support. Fixes #536
Khem Raj e50b9ef8 2019-12-07T02:34:14 powerpc64: Use memcpy to help platforms with no __int128. (#534) Signed-off-by: Khem Raj <raj.khem@gmail.com>
Anthony Green 6663047f 2019-11-29T07:00:35 Address platforms with no __int128.
Sergei Trofimovich 01a75ed7 2019-11-28T12:42:41 powerpc: fix build failure on power7 and older (#532) Build failure looks as: ``` libtool: compile: powerpc-unknown-linux-gnu-gcc \ -O2 -mcpu=powerpc -mtune=powerpc -pipe ... -c src/powerpc/ffi.c ... In file included from src/powerpc/ffi.c:33: src/powerpc/ffi_powerpc.h:65:9: error: '__int128' is not supported on this target 65 | typedef __int128 float128; | ^~~~~~~~ ``` The fix avoids using __int128 in favour of aligned char[16]. Closes: https://github.com/libffi/libffi/issues/531 Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>
Samuel Holland 73dd43af 2019-11-17T07:22:25 IEEE754 binary128 long double support for PowerPC64 (#526) * powerpc: Adjust flags to make room for vector types * powerpc64 ELFv2 IEEE128 long double support
Samuel Holland ea9b6639 2019-10-08T05:57:28 PowerPC bugfixes (#520) * powerpc: Silence warnings about unused labels * powerpc: Fix a couple of comments * powerpc: Fix alignment after float structs * powerpc: Don't pad rvalues copied from FP regs * powerpc: Add missing check in struct alignment * powerpc: Support homogeneous long double structs
Alan Modra ebf24166 2018-05-02T13:55:29 PowerPC64 ELFv1 fp arg fixes The ELFv1 ABI says: "Single precision floating point values are mapped to the second word in a single doubleword" and also "Floating point registers f1 through f13 are used consecutively to pass up to 13 floating point values, one member aggregates passed by value containing a floating point value, and to pass complex floating point values". libffi wasn't expecting float args in the second word, and wasn't passing one member aggregates in fp registers. This patch fixes those problems, making use of the existing ELFv2 homogeneous aggregate support since a one element fp struct is a special case of an homogeneous aggregate. I've also set a flag when returning pointers that might be used one day. This is just a tidy since the ppc64 assembly support code currently doesn't test FLAG_RETURNS_64BITS for integer types.. * src/powerpc/ffi_linux64.c (discover_homogeneous_aggregate): Compile for ELFv1 too, handling single element aggregates. (ffi_prep_cif_linux64_core): Call discover_homogeneous_aggregate for ELFv1. Set FLAG_RETURNS_64BITS for FFI_TYPE_POINTER return. (ffi_prep_args64): Call discover_homogeneous_aggregate for ELFv1, and handle single element structs containing float or double as if the element wasn't wrapped in a struct. Store floats in second word of doubleword slot when big-endian. (ffi_closure_helper_LINUX64): Similarly.
Tony Reix 0ff9419f 2017-05-17T14:57:53 This patch enables FFI Go Closure on AIX.
Gregory Pakosz bd72848c 2017-04-27T13:20:36 Prefix ALIGN macros with FFI_
Iain Sandoe 52fbd12d 2016-05-26T09:00:00 [Darwin-ppc, build] Fixes for clang build. clang is experimental on powerpc-darwin, but the changes are appropriate to cctools as as well. Use the 'official' simplified cmpwi insn, rather than the implied one accepted by cctools. Do not re-use a set value.
Alan Modra 43fc5bca 2015-08-03T23:34:05 Correct powerpc sysv stack argument accounting ppc32 starts using the stack for integer arg passing when we run out of integer arg passing registers. Similarly, we start using the stack for floating point args when we run out of floating point registers. The decision on where an integer arg goes does not depend on number of floating point args, nor does the decision on where a floating point arg goes depend on number of integer args. Alignment of stack args also simply depends on number of stack args. This patch untangles the horrible mess we had, with intarg_count being wrongly used to count both integer args and stack words. * src/powerpc/ffi_sysv.c (ffi_prep_cif_sysv_core): Count fprs, gprs, and stack words separately. (ffi_prep_args_SYSV): Similarly.
Richard Henderson 590663b3 2014-11-18T12:56:58 powerpc: Fix ffi_go_closure_linux64 Unlike ffi_closure_LINUX64, this entry point is called normally, so we already have the TOC in R2 and the closure in R11. * powerpc/linux64_closure.S (ffi_closure_LINUX64): Remove a register dependency chain. (ffi_go_closure_linux64): Don't load r11 or r2.
Alan Modra fa1040c1 2014-11-10T09:42:31 GO closures for powerpc linux Plus .cfi async unwind info, rearrangement of ffi_call_linux64 and ffi_call_SYSV function params to avoid register copies, tweaks to trampolines. * src/powerpc/ffitarget.h (FFI_GO_CLOSURES): Define. * src/powerpc/ffi.c (ffi_call_int): New function with extra closure param, and args rearranged on ffi_call_linux64 and ffi_call_SYSV calls, extracted from .. (ffi_call): ..here. (ffi_call_go, ffi_prep_go_closure): New functions. * src/powerpc/ffi_linux64.c (ffi_prep_closure_loc_linux64): Make hidden. Only flush insn part of ELFv2 trampoline. Don't shuffle ELFv1 trampoline. (ffi_closure_helper_LINUX64): Replace closure param with cif, fun, user_data params. * src/powerpc/ffi_powerpc.h (ffi_go_closure_sysv): Declare. (ffi_go_closure_linux64): Declare. (ffi_call_SYSV, fi_call_LINUX64): Update. (ffi_prep_closure_loc_sysv, ffi_prep_closure_loc_linux64): Declare. (ffi_closure_helper_SYSV, ffi_closure_helper_LINUX64): Update. * src/powerpc/ffi_sysv.c (ASM_NEEDS_REGISTERS): Increase to 6. (ffi_prep_closure_loc_sysv): Use bcl in trampoline, put data words last, flush just the insn part. (ffi_closure_helper_SYSV): Replace closure param with cif, fun and user_data params. * src/powerpc/linux64.S (ffi_call_LINUX64): Replace hand-written .eh_frame with .cfi directives. Adjust for changed param order. Pass extra "closure" param to user function in static chain. Add .cfi directives to describe epilogue. Don't provide traceback table for ELFv2 or _CALL_LINUX. * src/powerpc/linux64_closure.S (ffi_closure_LINUX64): Replace hand-written .eh_frame with .cfi directives. Adjust for changed ffi_closure_helper_LINUX64 params. Add .cfi directives to describe epilogue. Don't provide traceback table for ELFv2 or _CALL_LINUX. (ffi_go_closure_linux64): New function. * src/powerpc/sysv.S: Remove redundant .globl ffi_prep_args_SYSV. (ffi_call_SYSV): Make hidden. Replace hand-written .eh_frame with .cfi directives. Adjust for changed params. Pass extra "closure" param to user function in static chain. Add .cfi directives to describe epilogue. * src/powerpc/ppc_closure.S (ffi_closure_SYSV): Make hidden. Replace hand-written .eh_frame with .cfi directives. Adjust for changed ffi_closure_helper_SYSV params. Add .cfi directives to describe epilogue. Don't just use nops in the dead __NO_FPRS__ epilogues. (ffi_go_closure_sysv): New function.
Alan Modra d3d06f4c 2014-11-10T13:50:05 Fix powerpc breakage from 6e8a4460 * src/powerpc/ffitarget.h: #error on unexpected FFI_TYPE_LAST. (FFI_PPC_TYPE_LAST): Define. (FFI_TYPE_UINT128): Define in terms of FFI_PPC_TYPE_LAST. (FFI_SYSV_TYPE_SMALL_STRUCT, FFI_V2_TYPE_FLOAT_HOMOG): Likewise. (FFI_V2_TYPE_DOUBLE_HOMOG, FFI_V2_TYPE_SMALL_STRUCT): Likewise.
Richard Henderson f8632815 2014-11-13T12:32:35 powerpc: Delete patch output Clearly added by mistake.
Jakub Jelinek 32cb2ce8 2014-09-20T06:39:55 2014-09-10 Jakub Jelinek <jakub@redhat.com> * src/powerpc/linux64.S: Emit .note.GNU-stack even when POWERPC64 is not defined. * src/powerpc/linux64_closure.S: Likewise. Also test _CALL_ELF == 2.
Marcus Comstedt fa5f25c2 2014-01-04T19:00:08 Linux/ppc64: Remove assumption on contents of r11 in closure
Alan Modra ab79d6e2 2013-11-21T06:12:35 This separates the 32-bit sysv/linux/bsd code from the 64-bit linux code, and makes it possible to link code compiled with different options to those used to compile libffi. For example, a -mlong-double-128 libffi can be used with -mlong-double-64 code. Using the return value area as a place to pass parameters wasn't such a good idea, causing a failure of cls_ulonglong.c. I didn't see this when running the mainline gcc libffi testsuite because that version of the test is inferior to the upstreamm libffi test. Using NUM_FPR_ARG_REGISTERS rather than NUM_FPR_ARG_REGISTERS64 meant that a parameter save area could be allocated before it was strictly necessary. Wrong but harmless. Found when splitting apart ffi.c into 32-bit and 64-bit support.
Alan Modra ac753688 2013-11-21T06:12:35 This separates the 32-bit sysv/linux/bsd code from the 64-bit linux code, and makes it possible to link code compiled with different options to those used to compile libffi. For example, a -mlong-double-128 libffi can be used with -mlong-double-64 code. Using the return value area as a place to pass parameters wasn't such a good idea, causing a failure of cls_ulonglong.c. I didn't see this when running the mainline gcc libffi testsuite because that version of the test is inferior to the upstreamm libffi test. Using NUM_FPR_ARG_REGISTERS rather than NUM_FPR_ARG_REGISTERS64 meant that a parameter save area could be allocated before it was strictly necessary. Wrong but harmless. Found when splitting apart ffi.c into 32-bit and 64-bit support.
Iain Sandoe aa1f62c0 2013-11-18T13:11:56 Fix PowerPC Darwin FDE encodings to use pcrel correctly. Modernise the picbase labels.
Alan Modra 16d56c51 2013-11-18T06:36:03 An #endif in the wrong place would cause compile failure on powerpcle. Using bl instead of b doesn't cause runtime failures as you might think, but does mess the processor branch prediction.
Alan Modra 83f65b63 2013-11-16T06:53:50 Finally, this adds _CALL_ELF == 2 support. ELFv1 objects can't be linked with ELFv2 objects, so this is one case where preprocessor tests in ffi.c are fine. Also, there is no need to define a new FFI_ELFv2 or somesuch value in enum ffi_abi. FFI_LINUX64 will happily serve both ABIs.
Alan Modra 1fd04578 2013-11-16T06:53:50 Finally, this adds _CALL_ELF == 2 support. ELFv1 objects can't be linked with ELFv2 objects, so this is one case where preprocessor tests in ffi.c are fine. Also, there is no need to define a new FFI_ELFv2 or somesuch value in enum ffi_abi. FFI_LINUX64 will happily serve both ABIs.
Alan Modra 36285137 2013-11-16T06:52:43 Andreas' 2013-02-08 change reverted some breakage for struct return values from 2011-11-12, but in so doing reintroduced string instructions to sysv.S that are not supported on all powerpc variants. This patch properly copies the bounce buffer to destination in C code rather than in asm. I have tested this on powerpc64-linux, powerpc-linux and powerpc-freebsd. Well, the last on powerpc-linux by lying to configure with CC="gcc -m32 -msvr4-struct-return -mlong-double-64" \ CXX="g++ -m32 -msvr4-struct-return -mlong-double-64" \ /src/libffi-current/configure --build=powerpc-freebsd and then make && make CC="gcc -m32" CXX="g++ -m32" \ RUNTESTFLAGS=--target_board=unix/-m32/-msvr4-struct-return/-mlong-double-64\ check
Alan Modra 1c06515d 2013-11-16T06:41:36 The powerpc64 ABIs align structs passed by value, a fact ignored by gcc for quite some time. Since gcc now does the correct alignment, libffi needs to follow suit. This ought to be made selectable via a new abi value, and the #ifdefs removed from ffi.c along with many other #ifdefs present there and in assembly. I'll do that with a followup patch sometime. This is a revised version of https://sourceware.org/ml/libffi-discuss/2013/msg00162.html
Alan Modra a97cf1fa 2013-11-16T06:40:13 This patch prepares for ELFv2, where sizes of these areas change. It also makes some minor changes to improve code efficiency.
Alan Modra 164283f4 2013-11-16T06:38:55 The powerpc64 support opted to pass floating point values both in the fpr area and the parameter save area, necessary when the backend doesn't know if a function argument corresponds to the ellipsis arguments of a variadic function. This patch adds powerpc support for variadic functions, and changes the code to only pass fp in the ABI mandated area. ELFv2 needs this change since the parameter save area may not exist there. This also fixes two faulty tests that used a non-variadic function cast to call a variadic function, and spuriously reasoned that this is somehow necessary for static functions..
Anthony Green 128cd1d2 2013-10-08T06:45:51 Fix spelling errors
Anthony Green d3d099b4 2013-07-02T16:11:38 little-endian ppc64 support
Anthony Green 4acf0056 2013-03-16T08:18:45 Build fix for soft-float power targets
Anthony Green 70b11b47 2013-02-08T16:12:19 Fix small struct passing on ppc
Anthony Green 8bd15d13 2013-02-08T13:56:37 Fix many.c testcase for ppc
Anthony Green 05fbe1fa 2013-01-11T16:54:40 xlc compiler support
rofl0r f6b58d2b 2012-11-22T16:26:21 fix build on ppc when long double == double
Anthony Green dfadfb19 2012-10-31T06:46:41 Rebase for ppc64 fix
Anthony Green 853cc722 2012-04-10T06:33:33 Fix typo for darwin targets
Anthony Green 65f40c35 2012-04-03T07:35:59 Repair ppc build regression.
Anthony Green bd78c9c3 2012-03-21T08:09:30 More cygwin fixes
Anthony Green 84d3253f 2012-03-19T23:07:35 Rebase post GCC merge
Anthony Green 8360bf1c 2012-02-23T07:01:13 Ensure that users don't include ffitarget.h directly
Anthony Green ea14ae85 2011-11-12T16:36:59 clean up
Anthony Green 52891f8a 2011-11-12T16:35:55 Add powerpc soft float support
Anthony Green af18df2b 2011-11-12T15:52:08 Remove use of ppc string instructions
Anthony Green 3d56106b 2011-11-12T07:20:24 Rebase
Anthony Green 2d3fb364 2011-03-30T16:54:42 Fix darwin EH
Anthony Green 18dd85d6 2011-02-25T16:23:04 rc8. fix last patch.
Anthony Green 74ee6ea8 2011-02-25T15:52:14 rc7. More AIX fixes.
Anthony Green 2541679d 2011-02-25T15:09:13 Fix ppc32 bug
Anthony Green 1fbf9dc4 2011-02-13T08:06:39 Fix bad_abi test. rc5.
Anthony Green 6972a4ff 2011-02-11T07:32:51 Fix xlc build on AIX
Anthony Green 19ce7131 2011-02-08T10:34:23 grammar fix
Anthony Green 0e584399 2011-02-08T07:52:40 Refresh from GCC
Anthony Green 2db72615 2010-11-21T10:50:56 Rebase
Anthony Green 5feacad4 2010-08-05T08:30:04 define generic symbols carefully
Anthony Green 3f5b1375 2010-07-12T14:39:18 rebase
Anthony Green 9dc9a293 2010-04-13T10:33:52 Rebase to latest GCC sources
Anthony Green f2c2a4fc 2010-04-13T10:19:28 Remove warnings and add OS/2 support
Anthony Green 7b7a42f2 2010-01-12T09:14:14 Rebase from GCC
Anthony Green c3042afa 2010-01-01T08:08:02 Reset quilt patches post 3.0.9 merge with GCC
Anthony Green 2e7e03d0 2009-12-31T07:43:22 Final updates before 3.0.9
Anthony Green aea706c5 2009-12-29T10:09:31 really 3.0.9rc12
Anthony Green 0cfe60e9 2009-12-29T10:06:04 3.0.9rc12
Anthony Green 9458d88f 2009-12-26T07:02:27 Rebase from GCC
Anthony Green 6a341241 2009-12-26T06:51:33 Add Andreas Schwab's powerpc fix
Anthony Green da11bece 2009-12-24T05:34:46 Release 3.0.9rc5
Anthony Green 115ab36f 2009-12-24T00:22:00 Update missing changes for 3.0.9r4.
Anthony Green cd98813d 2009-10-05T00:25:29 From Jens Rehsack. Fix for 64-bit AIX.
Anthony Green c6dddbd0 2009-10-04T08:11:33 Initial commit