* [Qemu-devel] win32 build broken @ 2005-01-04 13:03 Ronald 2005-01-04 20:55 ` René Korthaus 0 siblings, 1 reply; 14+ messages in thread From: Ronald @ 2005-01-04 13:03 UTC (permalink / raw) To: qemu-devel Hi, build of cvs for win32 abort with: /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c: In function `op_goto_tb0': /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c:1266: warning: implicit declaration of function `GOTO_TB' ../dyngen.exe -o op.h op.o dyngen: ret or jmp expected at the end of op_movslq_T0_T0 make[1]: *** [op.h] Erreur 1 GOTO_TB is not defined in exec-all.h because of undefined USE_DIRECT_JUMP. Best wishes for this new year. ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] win32 build broken 2005-01-04 13:03 [Qemu-devel] win32 build broken Ronald @ 2005-01-04 20:55 ` René Korthaus 2005-01-06 22:35 ` Renzo Davoli 0 siblings, 1 reply; 14+ messages in thread From: René Korthaus @ 2005-01-04 20:55 UTC (permalink / raw) To: daimon55, qemu-devel ------------------------------------------------------------- The signature is currently on vacations. Am 04.01.2005 um 14:03 schrieb Ronald: > Hi, > > build of cvs for win32 abort with: > > /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c: In function > `op_goto_tb0': > /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c:1266: > warning: implicit declaration of function `GOTO_TB' > ../dyngen.exe -o op.h op.o > dyngen: ret or jmp expected at the end of op_movslq_T0_T0 > make[1]: *** [op.h] Erreur 1 > > GOTO_TB is not defined in exec-all.h because of undefined > USE_DIRECT_JUMP. Same issue on Mac OS X: /Users/cordney/cocoa/qemu-cvs/target-i386/op.c: In function `op_goto_tb0': /Users/cordney/cocoa/qemu-cvs/target-i386/op.c:1266: warning: implicit declaration of function `GOTO_TB' /Users/cordney/cocoa/qemu-cvs/target-i386/op.c: At top level: /Users/cordney/cocoa/qemu-cvs/dyngen-exec.h:201: warning: `__op_param3' defined but not used ../dyngen -o op.h op.o dyngen: empty code for op_movslq_T0_T0 make[1]: *** [op.h] Error 1 make: *** [all] Error 1 May somebody provide a fix? > > > > Best wishes for this new year. > > > > _______________________________________________ > Qemu-devel mailing list > Qemu-devel@nongnu.org > http://lists.nongnu.org/mailman/listinfo/qemu-devel > ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] win32 build broken 2005-01-04 20:55 ` René Korthaus @ 2005-01-06 22:35 ` Renzo Davoli 2005-01-06 22:52 ` Magnus Damm 2005-01-08 21:47 ` [Qemu-devel] Re: win32 build broken Ronald 0 siblings, 2 replies; 14+ messages in thread From: Renzo Davoli @ 2005-01-06 22:35 UTC (permalink / raw) To: qemu-devel; +Cc: daimon55 I have the same problem on linuxppc (also using tonight cvs updates). (the line is 1265 instead of 1266). renzo On Tue, Jan 04, 2005 at 09:55:53PM +0100, Ren? Korthaus wrote: > >Hi, > > > >build of cvs for win32 abort with: > > > >/home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c: In function > >`op_goto_tb0': > >/home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c:1266: > >warning: implicit declaration of function `GOTO_TB' > >../dyngen.exe -o op.h op.o > >dyngen: ret or jmp expected at the end of op_movslq_T0_T0 > >make[1]: *** [op.h] Erreur 1 > > > >GOTO_TB is not defined in exec-all.h because of undefined > >USE_DIRECT_JUMP. > > Same issue on Mac OS X: > > /Users/cordney/cocoa/qemu-cvs/target-i386/op.c: In function > `op_goto_tb0': > /Users/cordney/cocoa/qemu-cvs/target-i386/op.c:1266: warning: implicit > declaration of function `GOTO_TB' > /Users/cordney/cocoa/qemu-cvs/target-i386/op.c: At top level: > /Users/cordney/cocoa/qemu-cvs/dyngen-exec.h:201: warning: `__op_param3' > defined but not used > ../dyngen -o op.h op.o > dyngen: empty code for op_movslq_T0_T0 > make[1]: *** [op.h] Error 1 > make: *** [all] Error 1 > > May somebody provide a fix? ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] win32 build broken 2005-01-06 22:35 ` Renzo Davoli @ 2005-01-06 22:52 ` Magnus Damm 2005-01-07 7:37 ` [Qemu-devel] softmmu ppc Laurent Amon 2005-01-08 21:47 ` [Qemu-devel] Re: win32 build broken Ronald 1 sibling, 1 reply; 14+ messages in thread From: Magnus Damm @ 2005-01-06 22:52 UTC (permalink / raw) To: qemu-devel On Thu, 6 Jan 2005 23:35:07 +0100, Renzo Davoli <renzo@cs.unibo.it> wrote: > I have the same problem on linuxppc (also using tonight cvs updates). Speaking of linuxppc... Have you tried the improved softmmu patches on a ppc host? / magnus ^ permalink raw reply [flat|nested] 14+ messages in thread
* [Qemu-devel] softmmu ppc 2005-01-06 22:52 ` Magnus Damm @ 2005-01-07 7:37 ` Laurent Amon 2005-01-07 9:43 ` Daniel Egger 0 siblings, 1 reply; 14+ messages in thread From: Laurent Amon @ 2005-01-07 7:37 UTC (permalink / raw) To: Magnus Damm, qemu-devel On 6 janv. 05, at 23:52, Magnus Damm wrote: > Speaking of linuxppc... > Have you tried the improved softmmu patches on a ppc host? > I don't know about linuxppc, but on osx 1.3.7, as reported by Daniel Egger on Jan 2nd, it bombs out. Lga. ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] softmmu ppc 2005-01-07 7:37 ` [Qemu-devel] softmmu ppc Laurent Amon @ 2005-01-07 9:43 ` Daniel Egger 2005-01-07 12:06 ` Magnus Damm 0 siblings, 1 reply; 14+ messages in thread From: Daniel Egger @ 2005-01-07 9:43 UTC (permalink / raw) To: qemu-devel [-- Attachment #1: Type: text/plain, Size: 294 bytes --] On 07.01.2005, at 08:37, Laurent Amon wrote: > I don't know about linuxppc, but on osx 1.3.7, as reported by Daniel > Egger on Jan 2nd, it bombs out. I would try it on linuxppc but at the moment it clashes badly with Fabrices' changes. Is anyone working on an update? Servus, Daniel [-- Attachment #2: This is a digitally signed message part --] [-- Type: application/pgp-signature, Size: 478 bytes --] ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] softmmu ppc 2005-01-07 9:43 ` Daniel Egger @ 2005-01-07 12:06 ` Magnus Damm 2005-01-07 14:05 ` Piotras 2005-01-07 16:05 ` Daniel Egger 0 siblings, 2 replies; 14+ messages in thread From: Magnus Damm @ 2005-01-07 12:06 UTC (permalink / raw) To: Daniel Egger; +Cc: qemu-devel On Fri, 7 Jan 2005 10:43:15 +0100, Daniel Egger <de@axiros.com> wrote: > On 07.01.2005, at 08:37, Laurent Amon wrote: > > > I don't know about linuxppc, but on osx 1.3.7, as reported by Daniel > > Egger on Jan 2nd, it bombs out. > > I would try it on linuxppc but at the moment it clashes > badly with Fabrices' changes. Is anyone working on an > update? Well, I could update the patches soon unless Piotrek has some code locally that he wants to include in a new patch. What do you suggest Piotrek? Regarding ppc host support then updating that patch is probably a nobrainer and I would like to add as much OSX support as possible. So if someone could send me a patch with the changes that makes the code compile on OSX, then I would be happy to include them in the next patch. Thanks. / magnus ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] softmmu ppc 2005-01-07 12:06 ` Magnus Damm @ 2005-01-07 14:05 ` Piotras 2005-01-07 16:05 ` Daniel Egger 1 sibling, 0 replies; 14+ messages in thread From: Piotras @ 2005-01-07 14:05 UTC (permalink / raw) To: Magnus Damm; +Cc: qemu-devel Hi! I will not have time in the near future to update my patches. There is no point with my local patches to stop the update as they are still experimental and changes to published version are minimal anyway. Regards, Piotrek On Fri, 7 Jan 2005 13:06:38 +0100, Magnus Damm <magnus.damm@gmail.com> wrote: > On Fri, 7 Jan 2005 10:43:15 +0100, Daniel Egger <de@axiros.com> wrote: > > On 07.01.2005, at 08:37, Laurent Amon wrote: > > > > > I don't know about linuxppc, but on osx 1.3.7, as reported by Daniel > > > Egger on Jan 2nd, it bombs out. > > > > I would try it on linuxppc but at the moment it clashes > > badly with Fabrices' changes. Is anyone working on an > > update? > > Well, I could update the patches soon unless Piotrek has some code > locally that he wants to include in a new patch. What do you suggest > Piotrek? > > Regarding ppc host support then updating that patch is probably a > nobrainer and I would like to add as much OSX support as possible. So > if someone could send me a patch with the changes that makes the code > compile on OSX, then I would be happy to include them in the next > patch. > > Thanks. > > / magnus ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] softmmu ppc 2005-01-07 12:06 ` Magnus Damm 2005-01-07 14:05 ` Piotras @ 2005-01-07 16:05 ` Daniel Egger 2005-01-08 9:57 ` Magnus Damm 1 sibling, 1 reply; 14+ messages in thread From: Daniel Egger @ 2005-01-07 16:05 UTC (permalink / raw) To: Magnus Damm; +Cc: qemu-devel [-- Attachment #1.1: Type: text/plain, Size: 679 bytes --] On 07.01.2005, at 13:06, Magnus Damm wrote: > Regarding ppc host support then updating that patch is probably a > nobrainer and I would like to add as much OSX support as possible. So > if someone could send me a patch with the changes that makes the code > compile on OSX, then I would be happy to include them in the next > patch. It's mostly a matter of changing the assembler syntax of the inline assembler statements. I've attached my modified version which compiled with the Apple cctools; I cannot test whether they'll still work on Linux because I currently cannot compile qemu anymore with the patches. If in doubt we'll need to ifdef the assembly sections for OS X. [-- Attachment #1.2: mmu_map_templ.h --] [-- Type: text/plain, Size: 6539 bytes --] /* * MMU implementation using indirection table (map) * * Copyright (c) 2004 Piotr Krysiuk * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; either * version 2 of the License, or (at your option) any later version. * * This library is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this library; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ #if DATA_SIZE == 4 #define SUFFIX l #define USUFFIX l #define DATA_TYPE uint32_t #elif DATA_SIZE == 2 #define SUFFIX w #define USUFFIX uw #define DATA_TYPE uint16_t #define DATA_STYPE int16_t #elif DATA_SIZE == 1 #define SUFFIX b #define USUFFIX ub #define DATA_TYPE uint8_t #define DATA_STYPE int8_t #else #error unsupported data size #endif #if ACCESS_TYPE == 0 #define CPU_MEM_INDEX 0 #elif ACCESS_TYPE == 1 #define CPU_MEM_INDEX 1 #else #error invalid ACCESS_TYPE #endif #define RES_TYPE int #if defined(__i386__) static inline RES_TYPE glue(glue(ld, USUFFIX), MEMSUFFIX)(void *ptr) { RES_TYPE val; asm volatile ( "mov %1, %%eax\n" "shr %2, %%eax\n" "mov %4 (%3, %%eax, 4), %%eax\n" #if DATA_SIZE == 1 "movzbl (%1, %%eax, 1), %0\n" #elif DATA_SIZE == 2 "movzwl (%1, %%eax, 1), %0\n" #elif DATA_SIZE == 4 "movl (%1, %%eax, 1), %0\n" #else #error unsupported size #endif : "=r" (val) : "r" (ptr), "I" (TARGET_PAGE_BITS), "r" (env), "m" (*(char *)offsetof(CPUState, mmu_map[CPU_MEM_INDEX].add_read)) : "%eax", "memory", "cc"); return val; } #if DATA_SIZE <= 2 static inline int glue(glue(lds, SUFFIX), MEMSUFFIX)(void *ptr) { int val; asm volatile ( "mov %1, %%eax\n" "shr %2, %%eax\n" "mov %4 (%3, %%eax, 4), %%eax\n" #if DATA_SIZE == 1 "movsbl (%1, %%eax, 1), %0\n" #elif DATA_SIZE == 2 "movswl (%1, %%eax, 1), %0\n" #else #error unsupported size #endif : "=r" (val) : "r" (ptr), "I" (TARGET_PAGE_BITS), "r" (env), "m" (*(char *)offsetof(CPUState, mmu_map[CPU_MEM_INDEX].add_read)) : "%eax", "memory", "cc"); return val; } #endif static inline void glue(glue(st, SUFFIX), MEMSUFFIX)(void *ptr, RES_TYPE val) { asm volatile ( "mov %0, %%eax\n" "shr %2, %%eax\n" "mov %4 (%3, %%eax, 4), %%eax\n" #if DATA_SIZE == 1 "movb %b1, (%0, %%eax, 1)\n" #elif DATA_SIZE == 2 "movw %w1, (%0, %%eax, 1)\n" #elif DATA_SIZE == 4 "movl %1, (%0, %%eax, 1)\n" #else #error unsupported size #endif : : "r" (ptr), "r" (val), "I" (TARGET_PAGE_BITS), "r" (env), "m" (*(char *)offsetof(CPUState, mmu_map[CPU_MEM_INDEX].add_write)) : "%eax", "memory", "cc"); } #else #if defined(__powerpc__) /* this code assumes that the host ppc is running in big endian mode */ #ifdef TARGET_I386 #define ENDIAN_SWAP #endif #if 0 /* slow lookup, first shift right, then left */ "rlwinm 3, %1, 32 - %2, %2, 31\n" "addis 4, %3, %4@ha\n" "rlwinm 3, 3, 2, 0, 31 - 2\n" "addi 4, 4, %4@l\n" "lwzx 3, 4, 3\n" /* current lookup, one shift and mask operation */ "addis 4, %3, %4@ha\n" "rlwinm 3, %1, 32 - (%2 - 2), (%2 - 2), 31 - 2\n" "addi 4, 4, %4@l\n" "lwzx 3, 4, 3\n" /* better lookup, one shift and mask operation, mmu_map in register */ "rlwinm 3, %1, 32 - (%2 - 2), (%2 - 2), 31 - 2\n" "lwzx 3, %3, 3\n" #endif static inline RES_TYPE glue(glue(ld, USUFFIX), MEMSUFFIX)(void *ptr) { RES_TYPE val; asm volatile ( "addis r4, %3, ha16(%4)\n" "rlwinm r3, %1, 32 - (%2 - 2), (%2 - 2), 31 - 2\n" "addi r4, r4, lo16(%4)\n" "lwzx r3, r4, r3\n" #if DATA_SIZE == 1 "lbzx %0, %1, r3\n" #elif DATA_SIZE == 2 #ifdef ENDIAN_SWAP "lhbrx %0, %1, r3\n" #else "lhzx %0, %1, r3\n" #endif #elif DATA_SIZE == 4 #ifdef ENDIAN_SWAP "lwbrx %0, %1, r3\n" #else "lwzx %0, %1, r3\n" #endif #else #error unsupported size #endif : "=r" (val) : "r" (ptr), "I" (TARGET_PAGE_BITS), "r" (env), "i" (offsetof(CPUState, mmu_map[CPU_MEM_INDEX].add_read)) : "%r3", "%r4", "memory"); return val; } #if DATA_SIZE <= 2 static inline int glue(glue(lds, SUFFIX), MEMSUFFIX)(void *ptr) { int val; asm volatile ( "addis r4, %3, ha16(%4)\n" "rlwinm r3, %1, 32 - (%2 - 2), (%2 - 2), 31 - 2\n" "addi r4, r4, lo16(%4)\n" "lwzx r3, r4, r3\n" #if DATA_SIZE == 1 "lbzx r4, %1, r3\n" "extsb %0, r4\n" #elif DATA_SIZE == 2 #ifdef ENDIAN_SWAP "lhbrx r4, %1, r3\n" "extsh %0, r4\n" #else "lhax %0, %1, r3\n" #endif #else #error unsupported size #endif : "=r" (val) : "r" (ptr), "I" (TARGET_PAGE_BITS), "r" (env), "i" (offsetof(CPUState, mmu_map[CPU_MEM_INDEX].add_read)) : "%r3", "%r4", "memory"); return val; } #endif static inline void glue(glue(st, SUFFIX), MEMSUFFIX)(void *ptr, RES_TYPE val) { asm volatile ( "addis r4, %3, ha16(%4)\n" "rlwinm r3, %1, 32 - (%2 - 2), (%2 - 2), 31 - 2\n" "addi r4, r4, lo16(%4)\n" "lwzx r3, r4, r3\n" #if DATA_SIZE == 1 "stbx %0, %1, r3\n" #elif DATA_SIZE == 2 #ifdef ENDIAN_SWAP "sthbrx %0, %1, r3\n" #else "sthx %0, %1, r3\n" #endif #elif DATA_SIZE == 4 #ifdef ENDIAN_SWAP "stwbrx %0, %1, r3\n" #else "stwx %0, %1, r3\n" #endif #else #error unsupported size #endif : : "r" (val), "r" (ptr), "I" (TARGET_PAGE_BITS), "r" (env), "i" (offsetof(CPUState, mmu_map[CPU_MEM_INDEX].add_write)) : "%r3", "%r4", "memory"); } #else #error unsupported host CPU #endif #endif #undef RES_TYPE #undef DATA_TYPE #undef DATA_STYPE #undef SUFFIX #undef USUFFIX #undef DATA_SIZE #undef CPU_MEM_INDEX [-- Attachment #1.3: Type: text/plain, Size: 252 bytes --] It's starting to become really inconvenient to stay up-to-date with all the patches in my local tree. I'd be really grateful if some of mine and other patches which have been submitted for inclusion could be integrated in CVS. Servus, Daniel [-- Attachment #2: This is a digitally signed message part --] [-- Type: application/pgp-signature, Size: 478 bytes --] ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] softmmu ppc 2005-01-07 16:05 ` Daniel Egger @ 2005-01-08 9:57 ` Magnus Damm 2005-01-08 12:49 ` Daniel Egger 0 siblings, 1 reply; 14+ messages in thread From: Magnus Damm @ 2005-01-08 9:57 UTC (permalink / raw) To: Daniel Egger; +Cc: qemu-devel On Fri, 7 Jan 2005 17:05:21 +0100, Daniel Egger <de@axiros.com> wrote: > On 07.01.2005, at 13:06, Magnus Damm wrote: > > > Regarding ppc host support then updating that patch is probably a > > nobrainer and I would like to add as much OSX support as possible. So > > if someone could send me a patch with the changes that makes the code > > compile on OSX, then I would be happy to include them in the next > > patch. > > It's mostly a matter of changing the assembler syntax of > the inline assembler statements. I've attached my modified > version which compiled with the Apple cctools; I cannot test > whether they'll still work on Linux because I currently cannot > compile qemu anymore with the patches. If in doubt we'll need > to ifdef the assembly sections for OS X. I will try to use the syntax from the file you attached. And signal handling include files under OS X, should I include "sys/signal.h" ? / magnus ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] softmmu ppc 2005-01-08 9:57 ` Magnus Damm @ 2005-01-08 12:49 ` Daniel Egger 0 siblings, 0 replies; 14+ messages in thread From: Daniel Egger @ 2005-01-08 12:49 UTC (permalink / raw) To: Magnus Damm; +Cc: qemu-devel [-- Attachment #1: Type: text/plain, Size: 248 bytes --] On 08.01.2005, at 10:57, Magnus Damm wrote: > And signal handling include files under OS X, should I include > "sys/signal.h" ? Yes, either that or use _t declarations consequently trough-out all signal manipulation code. Servus, Daniel [-- Attachment #2: This is a digitally signed message part --] [-- Type: application/pgp-signature, Size: 478 bytes --] ^ permalink raw reply [flat|nested] 14+ messages in thread
* [Qemu-devel] Re: win32 build broken 2005-01-06 22:35 ` Renzo Davoli 2005-01-06 22:52 ` Magnus Damm @ 2005-01-08 21:47 ` Ronald 2005-01-08 22:12 ` Filip Navara 1 sibling, 1 reply; 14+ messages in thread From: Ronald @ 2005-01-08 21:47 UTC (permalink / raw) To: qemu-devel Le Thu, 06 Jan 2005 23:35:07 +0100, Renzo Davoli a écrit : > I have the same problem on linuxppc (also using tonight cvs updates). > (the line is 1265 instead of 1266). > > renzo > > On Tue, Jan 04, 2005 at 09:55:53PM +0100, Ren? Korthaus wrote: >> >Hi, >> > >> >build of cvs for win32 abort with: >> > >> >/home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c: In function >> >`op_goto_tb0': >> >/home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c:1266: >> >warning: implicit declaration of function `GOTO_TB' ../dyngen.exe -o >> >op.h op.o >> >dyngen: ret or jmp expected at the end of op_movslq_T0_T0 make[1]: *** >> >[op.h] Erreur 1 >> > >> >GOTO_TB is not defined in exec-all.h because of undefined >> >USE_DIRECT_JUMP. >> >> Same issue on Mac OS X: >> >> /Users/cordney/cocoa/qemu-cvs/target-i386/op.c: In function >> `op_goto_tb0': >> /Users/cordney/cocoa/qemu-cvs/target-i386/op.c:1266: warning: implicit >> declaration of function `GOTO_TB' >> /Users/cordney/cocoa/qemu-cvs/target-i386/op.c: At top level: >> /Users/cordney/cocoa/qemu-cvs/dyngen-exec.h:201: warning: `__op_param3' >> defined but not used >> ../dyngen -o op.h op.o >> dyngen: empty code for op_movslq_T0_T0 make[1]: *** [op.h] Error 1 >> make: *** [all] Error 1 >> >> May somebody provide a fix? After tonight's update here is what I get: /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c: In function `op_goto_tb0': /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/op.c:1265: warning: implicit declaration of function `GOTO_TB' ../dyngen.exe -o op.h op.o ../dyngen.exe -c -o opc.h op.o i386-mingw32-gcc -Wall -O2 -g -fno-strict-aliasing -fomit-frame-pointer -I. -I/home/ronald/Prog/Win32/combo/qemu-win32/target-i386 -I/home/ronald/Prog/Win32/combo/qemu-win32 -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -I/home/ronald/Prog/Win32/combo/qemu-win32/slirp -c -o translate-all.o /home/ronald/Prog/Win32/combo/qemu-win32/translate-all.c In file included from /home/ronald/Prog/Win32/combo/qemu-win32/translate-all.c:41: op.h: In function `dyngen_code': op.h:5241: syntax error before '.' token op.h:5244: syntax error before '.' token op.h:5252: syntax error before '.' token op.h:5255: syntax error before '.' token op.h:8049: warning: built-in function `sqrtl' declared as non-function op.h:8069: warning: built-in function `sqrtl' declared as non-function op.h:8083: warning: built-in function `sqrtl' declared as non-function op.h:8099: warning: built-in function `sqrtl' declared as non-function make: *** [translate-all.o] Erreur 1 I get the 4 warnings about sqrtl for linux build too but at different line number . The following are the lines 5239 to 5259 of the generated op.h _______________________________________________________________________ case INDEX_op_goto_tb0: { extern void op_goto_tb0(); extern char .text; extern char GOTO_TB; memcpy(gen_code_ptr, (void *)((char *)&op_goto_tb0+0), 14); *(uint32_t *)(gen_code_ptr + 3) = (long)(&.text) + 35472; *(uint32_t *)(gen_code_ptr + 8) = (long)(&GOTO_TB) - (long)(gen_code_ptr + 8) + 0 -4; gen_code_ptr += 14; } break; case INDEX_op_goto_tb1: { extern void op_goto_tb1(); extern char .text; extern char GOTO_TB; memcpy(gen_code_ptr, (void *)((char *)&op_goto_tb1+0), 14); *(uint32_t *)(gen_code_ptr + 3) = (long)(&.text) + 35488; *(uint32_t *)(gen_code_ptr + 8) = (long)(&GOTO_TB) - (long)(gen_code_ptr + 8) + 0 -4; gen_code_ptr += 14; } break; _______________________________________________________________________ qemu-system-ppc and qemu-system-sparc can still be built. ^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [Qemu-devel] Re: win32 build broken 2005-01-08 21:47 ` [Qemu-devel] Re: win32 build broken Ronald @ 2005-01-08 22:12 ` Filip Navara 2005-01-08 22:37 ` [Qemu-devel] " Ronald 0 siblings, 1 reply; 14+ messages in thread From: Filip Navara @ 2005-01-08 22:12 UTC (permalink / raw) To: daimon55, qemu-devel [-- Attachment #1: Type: text/plain, Size: 108 bytes --] Ronald wrote: >After tonight's update here is what I get: > [snip] Does the attached patch help? - Filip [-- Attachment #2: qemu.diff --] [-- Type: text/x-patch, Size: 1586 bytes --] --- exec-all.h Tue Jan 4 00:35:10 2005 +++ exec-all.h Fri Jan 7 17:44:27 2005 @@ -343,7 +341,7 @@ #elif defined(__i386__) && defined(USE_DIRECT_JUMP) /* we patch the jump instruction directly */ -#define GOTO_TB(opname, n)\ +#define GOTO_TB(opname, tbparam, n)\ do {\ asm volatile (".section .data\n"\ ASM_NAME(__op_label) #n "." ASM_NAME(opname) ":\n"\ @@ -372,6 +370,15 @@ } while (0) #else + +/* we patch the jump instruction directly */ +#define GOTO_TB(opname, tbparam, n)\ +do {\ + static void __attribute__((unused)) *dummy ## n = &&dummy_label ## n;\ + goto *(void *)(((TranslationBlock *)tbparam)->tb_next[n]);\ +dummy_label ## n:\ + EXIT_TB(); \ +} while (0) /* jump to next block operations (more portable code, does not need cache flushing, but slower because of indirect jump) */ --- slirp/slirp.h Sun Nov 14 19:58:04 2004 +++ slirp/slirp.h Fri Jan 7 16:08:04 2005 @@ -20,6 +20,7 @@ typedef uint64_t u_int64_t; typedef char *caddr_t; +# include <windows.h> # include <winsock2.h> # include <sys/timeb.h> # include <iphlpapi.h> Only in .: sparc-softmmu diff -ru ../qemu-snapshot-2005-01-06_23/target-i386/op.c ./target-i386/op.c --- ../qemu-snapshot-2005-01-06_23/target-i386/op.c Thu Jan 6 21:46:58 2005 +++ ./target-i386/op.c Fri Jan 7 17:39:47 2005 @@ -1262,12 +1262,12 @@ void OPPROTO op_goto_tb0(void) { - GOTO_TB(op_goto_tb0, 0); + GOTO_TB(op_goto_tb0, PARAM1, 0); } void OPPROTO op_goto_tb1(void) { - GOTO_TB(op_goto_tb1, 1); + GOTO_TB(op_goto_tb1, PARAM1, 1); } void OPPROTO op_jmp_label(void) ^ permalink raw reply [flat|nested] 14+ messages in thread
* [Qemu-devel] Re: Re: win32 build broken 2005-01-08 22:12 ` Filip Navara @ 2005-01-08 22:37 ` Ronald 0 siblings, 0 replies; 14+ messages in thread From: Ronald @ 2005-01-08 22:37 UTC (permalink / raw) To: qemu-devel Le Sat, 08 Jan 2005 23:12:13 +0100, Filip Navara a écrit : > Ronald wrote: > >>After tonight's update here is what I get: >> > [snip] > > Does the attached patch help? > The build abort later: i386-mingw32-gcc -Wall -O2 -g -fno-strict-aliasing -fomit-frame-pointer -I. -I/home/ronald/Prog/Win32/combo/qemu-win32/target-i386 -I/home/ronald/Prog/Win32/combo/qemu-win32 -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -I/home/ronald/Prog/Win32/combo/qemu-win32/slirp -c -o translate.o /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:868: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:869: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:874: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:875: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:887: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:888: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:889: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:893: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:894: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:895: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1168: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1169: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1170: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1171: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1173: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1174: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1178: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1179: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1180: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1181: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1183: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1184: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1188: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1189: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1190: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1191: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1193: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1194: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1211: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1212: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1213: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1216: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1217: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1218: warning: initialization from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c: In function `gen_jcc': /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1772: warning: assignment from incompatible pointer type /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1811: too many arguments to function `gen_op_jnz_T0_label' /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:1813: too many arguments to function `gen_op_jmp_label' /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c: In function `disas_insn': /home/ronald/Prog/Win32/combo/qemu-win32/target-i386/translate.c:5264: too many arguments to function `gen_op_jmp_label' make: *** [translate.o] Erreur 1 > - Filip--- exec-all.h Tue Jan 4 00:35:10 2005 +++ exec-all.h Fri Jan 7 17:44:27 > 2005 @@ -343,7 +341,7 @@ > #elif defined(__i386__) && defined(USE_DIRECT_JUMP) > > /* we patch the jump instruction directly */ > -#define GOTO_TB(opname, n)\ > +#define GOTO_TB(opname, tbparam, n)\ > do {\ > asm volatile (".section .data\n"\ > ASM_NAME(__op_label) #n "." ASM_NAME(opname) ":\n"\ > @@ -372,6 +370,15 @@ > } while (0) > > #else > + > +/* we patch the jump instruction directly */ +#define GOTO_TB(opname, > tbparam, n)\ +do {\ > + static void __attribute__((unused)) *dummy ## n = &&dummy_label ## > n;\ + goto *(void *)(((TranslationBlock *)tbparam)->tb_next[n]);\ > +dummy_label ## n:\ > + EXIT_TB(); \ > +} while (0) > > /* jump to next block operations (more portable code, does not need > cache flushing, but slower because of indirect jump) */ > --- slirp/slirp.h Sun Nov 14 19:58:04 2004 +++ slirp/slirp.h Fri Jan 7 > 16:08:04 2005 @@ -20,6 +20,7 @@ > typedef uint64_t u_int64_t; > typedef char *caddr_t; > > +# include <windows.h> > # include <winsock2.h> > # include <sys/timeb.h> > # include <iphlpapi.h> > Only in .: sparc-softmmu > diff -ru ../qemu-snapshot-2005-01-06_23/target-i386/op.c > ./target-i386/op.c --- ../qemu-snapshot-2005-01-06_23/target-i386/op.c Thu > Jan 6 21:46:58 2005 +++ ./target-i386/op.c Fri Jan 7 17:39:47 2005 @@ > -1262,12 +1262,12 @@ > > void OPPROTO op_goto_tb0(void) > { > - GOTO_TB(op_goto_tb0, 0); > + GOTO_TB(op_goto_tb0, PARAM1, 0); > } > > void OPPROTO op_goto_tb1(void) > { > - GOTO_TB(op_goto_tb1, 1); > + GOTO_TB(op_goto_tb1, PARAM1, 1); > } > > void OPPROTO op_jmp_label(void)_______________________________________________ Qemu-devel mailing list > Qemu-devel@nongnu.org > http://lists.nongnu.org/mailman/listinfo/qemu-devel ^ permalink raw reply [flat|nested] 14+ messages in thread
end of thread, other threads:[~2005-01-08 22:50 UTC | newest] Thread overview: 14+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2005-01-04 13:03 [Qemu-devel] win32 build broken Ronald 2005-01-04 20:55 ` René Korthaus 2005-01-06 22:35 ` Renzo Davoli 2005-01-06 22:52 ` Magnus Damm 2005-01-07 7:37 ` [Qemu-devel] softmmu ppc Laurent Amon 2005-01-07 9:43 ` Daniel Egger 2005-01-07 12:06 ` Magnus Damm 2005-01-07 14:05 ` Piotras 2005-01-07 16:05 ` Daniel Egger 2005-01-08 9:57 ` Magnus Damm 2005-01-08 12:49 ` Daniel Egger 2005-01-08 21:47 ` [Qemu-devel] Re: win32 build broken Ronald 2005-01-08 22:12 ` Filip Navara 2005-01-08 22:37 ` [Qemu-devel] " Ronald
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).