[valgrind] [Bug 468575] Add support for RISC-V
https://bugs.kde.org/show_bug.cgi?id=468575 --- Comment #19 from JojoR --- (In reply to Paul Floyd from comment #18) > On a machine that is StarFive JH7100 SoC (2x SiFive U74 at 1.20 GHz) > > I have a problem with make check. > > gcc -Winline -Wall -Wshadow -Wno-long-long -g -fno-stack-protector -o > compressed compressed.o > compressed.o: in function `test_compressed_10': > /home/paulf/valgrind-riscv64/none/tests/riscv64/compressed.c:347:(. > text+0x56d16): dangerous relocation: The addend isn't allowed for > R_RISCV_GOT_HI20 > collect2: error: ld returned 1 exit status > make[5]: *** [Makefile:741: compressed] Error 1 > make[5]: Leaving directory '/home/paulf/valgrind-riscv64/none/tests/riscv64' > make[4]: *** [Makefile:899: check-am] Error 2 > make[4]: Leaving directory '/home/paulf/valgrind-riscv64/none/tests/riscv64' > make[3]: *** [Makefile:2270: check-recursive] Error 1 > make[3]: Leaving directory '/home/paulf/valgrind-riscv64/none/tests' > make[2]: *** [Makefile:1101: check-recursive] Error 1 > make[2]: Leaving directory '/home/paulf/valgrind-riscv64/none' > make[1]: *** [Makefile:926: check-recursive] Error 1 > make[1]: Leaving directory '/home/paulf/valgrind-riscv64' > make: *** [Makefile:1225: check] Error 2 > > Maybe a GCC problem? > > gcc --version > gcc (Debian 13.2.0-13) 13.2.0 Yes, looks ld met some bugs :( and I suggest you can get the latest toolchain from here: https://github.com/riscv-collab/riscv-gnu-toolchain/releases/ -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 468575] Add support for RISC-V
https://bugs.kde.org/show_bug.cgi?id=468575 --- Comment #15 from JojoR --- (In reply to Paul Floyd from comment #14) > I built and ran regtest on my github clone of > https://github.com/petrpavlu/valgrind-riscv64 on FreeBSD amd64. > > There were a couple of unexpected failures in memcheck/tests/freebsd (one of > the scalars and eventfd2). I'll look into them and also try some more > platforms. Good news, thanks for your efforts, hope that will been fixed soon :) -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 468575] Add support for RISC-V
https://bugs.kde.org/show_bug.cgi?id=468575 --- Comment #13 from JojoR --- @Paul Floyd Has anyone reviewed these patches ? It has taken long time and there is not any feedback :) Could we review & push these basic ISAs (rv64gc) implementation first ? other of extensions like Vector of RISCV depend on this, that will be the next steps :) -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 474280] Add Ijk_ExitBB IR for potential critical state
https://bugs.kde.org/show_bug.cgi?id=474280 --- Comment #2 from JojoR --- authors: - zhaoming...@linux.alibaba.com - rjie...@linux.alibaba.com -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 474280] Add Ijk_ExitBB IR for potential critical state
https://bugs.kde.org/show_bug.cgi?id=474280 JojoR changed: What|Removed |Added Attachment #161502|0 |1 is obsolete|| --- Comment #1 from JojoR --- Created attachment 161503 --> https://bugs.kde.org/attachment.cgi?id=161503=edit v1-0001-VEX-Add-Ijk_ExitBB-IR-for-potential-critical-state-c.patch -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 474280] New: Add Ijk_ExitBB IR for potential critical state
https://bugs.kde.org/show_bug.cgi?id=474280 Bug ID: 474280 Summary: Add Ijk_ExitBB IR for potential critical state Classification: Developer tools Product: valgrind Version: unspecified Platform: Other OS: Other Status: REPORTED Severity: normal Priority: NOR Component: vex Assignee: jsew...@acm.org Reporter: rjie...@gmail.com Target Milestone: --- Created attachment 161502 --> https://bugs.kde.org/attachment.cgi?id=161502=edit v1-0001-VEX-Add-Ijk_ExitBB-IR-for-potential-critical-state-c.patch Generally we use Ijk_Boring in most jump instructions, and some backends use this IR to make BB optimization 'chaining' in instruction selection stage. For some special ISAs like RISC-V vector [1], vsetvl/vsetvli instruction will set VTYPE/VL CSRs which are depended by next BB, so we add this new Ijk_ExitBB IR to exit BB without any optimization. [1] https://github.com/riscv/riscv-v-spec -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 473978] Add support for checking of cpu state in code translation
https://bugs.kde.org/show_bug.cgi?id=473978 --- Comment #7 from JojoR --- Created attachment 161305 --> https://bugs.kde.org/attachment.cgi?id=161305=edit v1-0003-scheduler-Add-code-basic-block-flag-support.patch -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 473978] Add support for checking of cpu state in code translation
https://bugs.kde.org/show_bug.cgi?id=473978 --- Comment #6 from JojoR --- Created attachment 161304 --> https://bugs.kde.org/attachment.cgi?id=161304=edit v1-0002-transtab-Add-code-basic-block-flag-support.patch -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 473978] Add support for checking of cpu state in code translation
https://bugs.kde.org/show_bug.cgi?id=473978 JojoR changed: What|Removed |Added Attachment #161300|0 |1 is obsolete|| Attachment #161301|0 |1 is obsolete|| Attachment #161302|0 |1 is obsolete|| --- Comment #5 from JojoR --- Created attachment 161303 --> https://bugs.kde.org/attachment.cgi?id=161303=edit v1-0001-VEX-Add-code-basic-block-flag-support.patch -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 473978] Add support for checking of cpu state in code translation
https://bugs.kde.org/show_bug.cgi?id=473978 --- Comment #4 from JojoR --- authors: - zhaoming...@linux.alibaba.com - rjie...@linux.alibaba.com -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 473978] Add support for checking of cpu state in code translation
https://bugs.kde.org/show_bug.cgi?id=473978 --- Comment #3 from JojoR --- Created attachment 161302 --> https://bugs.kde.org/attachment.cgi?id=161302=edit v1-0003-scheduler-Add-code-basic-block-flag-support.patch -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 473978] Add support for checking of cpu state in code translation
https://bugs.kde.org/show_bug.cgi?id=473978 --- Comment #2 from JojoR --- Created attachment 161301 --> https://bugs.kde.org/attachment.cgi?id=161301=edit v1-0002-transtab-Add-code-basic-block-flag-support.patch -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 473978] Add support for checking of cpu state in code translation
https://bugs.kde.org/show_bug.cgi?id=473978 --- Comment #1 from JojoR --- Created attachment 161300 --> https://bugs.kde.org/attachment.cgi?id=161300=edit v1-0001-VEX-Add-code-basic-block-flag-support.patch -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 473978] New: Add support for checking of cpu state in code translation
https://bugs.kde.org/show_bug.cgi?id=473978 Bug ID: 473978 Summary: Add support for checking of cpu state in code translation Classification: Developer tools Product: valgrind Version: unspecified Platform: Other OS: Other Status: REPORTED Severity: normal Priority: NOR Component: general Assignee: jsew...@acm.org Reporter: rjie...@gmail.com Target Milestone: --- valgrind buffered translated codes in "transtab", and reuse these codes directly if next PC is same as these code. there is a limit for this design, every code block must have enough instruction information in current code block. In other ISAs like RISCV Vector [1], current instruction codes depend other state of CSRs like VTYPE/VL, only PC information is not enough to judge that valgrind should reuse directly buffered code or not. we need to add cpu state in code translation, and save state into "transtab". [1] https://github.com/riscv/riscv-v-spec -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 468979] Add support for RISC-V vector instructions
https://bugs.kde.org/show_bug.cgi?id=468979 --- Comment #1 from JojoR --- Hi, We are glad to open source RVV implementation here: https://github.com/rjiejie/valgrind-riscv64 4 kinds extra ISAs were added in this repo: RV64Zfh : Half-precision floating-point RV64Xthead [1] : T-HEAD vendor extension for RV64G RV64V0p7 [2] : Vector 0.7.1 RV64V [3] : Vector 1.0 [1] https://github.com/T-head-Semi/thead-extension-spec [2] https://github.com/riscv/riscv-v-spec/releases/tag/0.7.1 [3] https://github.com/riscv/riscv-v-spec/releases/tag/v1.0 -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 468575] Add support for RISC-V
https://bugs.kde.org/show_bug.cgi?id=468575 --- Comment #11 from JojoR --- @Paul Floyd Has anyone reviewed these patches ? It has taken two months and there is not any feedback :) -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 383010] Add support for AVX-512 instructions
https://bugs.kde.org/show_bug.cgi?id=383010 JojoR changed: What|Removed |Added CC||rjie...@gmail.com --- Comment #88 from JojoR --- Excellent work, it's a long time :) Is there any schedule for upstreaming about this feature ? these patches looks like a prototype implementation ? I opened another thread about RISC-V vector, some issues should be common in VEX or plugin Memcheck, even more vector ISA generator :) Any suggestions for these ? @Tanya @Julian Seward BTW, is there anyone interested in ARM's SVE or RISC-V vector ? See more details from https://bugs.kde.org/show_bug.cgi?id=468979 -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 468979] New: Add support for RISC-V vector instructions
https://bugs.kde.org/show_bug.cgi?id=468979 Bug ID: 468979 Summary: Add support for RISC-V vector instructions Classification: Developer tools Product: valgrind Version: unspecified Platform: Other OS: Linux Status: REPORTED Severity: normal Priority: NOR Component: vex Assignee: jsew...@acm.org Reporter: rjie...@gmail.com Target Milestone: --- We consider to add RVV/Vector [1] feature in valgrind, there are some challenges. RVV like ARM's SVE [2] programming model, it's scalable/VLA, that means the vector length is agnostic. ARM's SVE is not supported in valgrind :( See more details from RFC [3] [1] https://github.com/riscv/riscv-v-spec [2] https://community.arm.com/arm-research/b/articles/posts/the-arm-scalable-vector-extension-sve [3] https://sourceforge.net/p/valgrind/mailman/valgrind-developers/thread/84b7a55c-1868-ca14-2626-ffb88925741a%40linux.alibaba.com/#msg37834524 -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 468575] Add support for RISC-V
https://bugs.kde.org/show_bug.cgi?id=468575 JojoR changed: What|Removed |Added CC||rjie...@gmail.com --- Comment #10 from JojoR --- (In reply to Petr Pavlu from comment #9) > (In reply to Paul Floyd from comment #8) > > hginfo fails on several systems > > TLS is, well, tricky > > > > Are the other failures "reaonable"? > > > memcheck/tests/sh-mem-random (stdout) > > memcheck/tests/sh-mem-random (stderr) > > Another unexpected mmap() failure. >From our investigation, high address mmap is not supported by some riscv board/os if it support feature "Sv39" only, so it is reasonable also :) -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 467839] New: Gdbserver: Improve compatibility of library directory name
https://bugs.kde.org/show_bug.cgi?id=467839 Bug ID: 467839 Summary: Gdbserver: Improve compatibility of library directory name Classification: Developer tools Product: valgrind Version: unspecified Platform: Other OS: Other Status: REPORTED Severity: normal Priority: NOR Component: general Assignee: jsew...@acm.org Reporter: rjie...@gmail.com Target Milestone: --- Created attachment 157620 --> https://bugs.kde.org/attachment.cgi?id=157620=edit 0001-Gdbserver-Improve-compatibility-of-library-directory.patch Some linux os make softlink from customized directory like lib64xxx into standard system lib64 directory, we could improve compatibility for these cases, see Attachment :) -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 467036] Add time cost statistics for Regtest
https://bugs.kde.org/show_bug.cgi?id=467036 JojoR changed: What|Removed |Added CC||rjie...@gmail.com -- You are receiving this mail because: You are watching all bug changes.
[valgrind] [Bug 467036] New: Add time cost statistics for Regtest
https://bugs.kde.org/show_bug.cgi?id=467036 Bug ID: 467036 Summary: Add time cost statistics for Regtest Classification: Developer tools Product: valgrind Version: unspecified Platform: Other OS: Other Status: REPORTED Severity: normal Priority: NOR Component: general Assignee: jsew...@acm.org Reporter: rjie...@gmail.com Target Milestone: --- Created attachment 157099 --> https://bugs.kde.org/attachment.cgi?id=157099=edit 0001-Regtest-add-time-cost-statistics.patch This is a small improved patch for regtest, some testers hope that valgrind could collect every test modules/dirs time costs for helping resource evaluation :) -- You are receiving this mail because: You are watching all bug changes.