)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"b7844a1ada2c5011173ab7c6833f3ed7745ef37a","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":2,"id":"069c52fb_208a7cea","updated":"2025-10-22 10:27:33.000000000","message":"Evgeniy, I rebased this patch to top of riscv sync series.","commit_id":"bb221506055c5101e3175b97d7449dac79d71d67"},{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"1badef2c3c4e78fc9b5fc8c779c7e3dd4cd757a4","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":4,"id":"be3f9ec9_f1102adc","updated":"2025-11-07 07:26:10.000000000","message":"Thanks!\n\nAntonio, as the patch could potentially break other targets with hwthread, please take a look. After running architecture independent part of riscv-test on 2 core Cortex-M33 and 2 core Cortex-A7 I\u0027m pretty confident there is no such regression. Unfortunately I didn\u0027t have time to run tests on aarch64 - let\u0027s hope there is no bad surprise.","commit_id":"3d5c42a6f9cacec1180bf4ad5bdb7bff9c0cfa47"}],"src/server/gdb_server.c":[{"author":{"_account_id":1002047,"name":"Evgeniy Naydanov","email":"eugnay@gmail.com","username":"en-sc"},"change_message_id":"4e79758aa450036a248c3cbe2d986a7062b5b761","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"016a9975_69dbfb9e","line":3133,"updated":"2025-10-23 14:07:37.000000000","message":"Anatoly, Tomas, I\u0027m sorry, I should have actually read what I was cherry-picking.\n\nIn the original patch `rtos_update_threads()` was commented-out (not removed) and this comment actually describes why.\n\nThe removal of this `rtos_update_threads()` is exactly what results in the difference in the behavior between RISC-V fork and the Sync that was observed in the failure of ProgramSwWatchpoint test.\n\nThe idea is: active GDB thread for hwthread RTOS is determined by `hwthread_update_threads()` function based solemnly on current debug reason.\nTo handle SW watchpoint GDB steps both threads, one after the other.\n\nLet\u0027s assume GDB starts from thread 1. Then:\n1. GDB steps thread 1.\n2. Debug reasons: thread 1 -- step, thread 2 -- some other reason with a lower priority (see `hwthread_update_threads()`).\n3. Active thread is 1.\n4. GDB steps thread 2.\n5. Debug reasons: thread 1 -- step, thread 2 -- also step.\nFrom this point onward the active thread no longer depends on which thread was stepped, but on the order the threads are traversed in `hwthread_update_threads()` and no longer changes.\n\nI don\u0027t think removing `rtos_update_threads(target)` from here is the proper fix though. Can the step be considered atomic? Meaning we can delay anything that happens on the target until after the step finishes.","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"4503c12784441df0ee66d440247c09d644bfbef0","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"703a936a_ed719778","line":3133,"in_reply_to":"016a9975_69dbfb9e","updated":"2025-10-23 14:22:15.000000000","message":"Uhh, very complicated for me.","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"fd0b29c554d548a8d16ab600cb8aceb7ed88b1df","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"f4a99824_dc33259a","line":3133,"in_reply_to":"19b296e4_7d5aee69","updated":"2025-10-24 20:26:10.000000000","message":"Thanks, I confirm vanilla gdb 16.2 compiled for --target\u003driscv64-unknown-elf really uses vCont;s and ProgramSwWatchpoint test often fails\n```\n../../gdb/infrun.c:6711: internal-error: finish_step_over: Assertion `ecs-\u003eevent_thread-\u003econtrol.trap_expected\u0027 failed.\nA problem internal to GDB has been detected,\n```\nwith the series up to \n9176: rtos: server: target: ask the RTOS which target to set swbp on. | https://review.openocd.org/c/openocd/+/9176\n\nWith the complete series including the topmost patch\n9183: Draft: fix missing thread ID in stop reply when smp-configured hart (but not hart 0) single-stepped | https://review.openocd.org/c/openocd/+/9183\nall tests to spike64-2-hwthread target passed (all tests ran 3 times, ProgramSwWatchpoint at least 10 times).\nLooks promising.\n\nBTW gdb-multiarch 16.3 from Debian Trixie workarounds vCont;s by setting a temp BP and vCont;c again as the older versions (13.1, 15.1) did. No idea why, and if there is some wrong patch applied by Debian. Or some problem with multiarch config?","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002047,"name":"Evgeniy Naydanov","email":"eugnay@gmail.com","username":"en-sc"},"change_message_id":"51629f42d5b80308c6f301fce44fa7ae5ad30920","unresolved":false,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"a803d704_28b9b795","line":3133,"in_reply_to":"36d1308d_c70208d9","updated":"2025-11-06 10:38:20.000000000","message":"Done","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002047,"name":"Evgeniy Naydanov","email":"eugnay@gmail.com","username":"en-sc"},"change_message_id":"6f03f1f905e937cad419b1dd25e60bbbff9393bd","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"56e9e389_032e29cd","line":3133,"in_reply_to":"4347c20d_1b56528e","updated":"2025-10-23 17:01:54.000000000","message":"One more thing: cherry-picking 9183 changes the behavior of GDB:\nwithout 9183 it sends `vCont;s;1;c` and `vCont;s;2;c`, explicitly specifying the thread. With 9183 it starts sending `vCont;s:0;c` (0 here means any thread [1]). I don\u0027t get the reason why.\n[1]: https://sourceware.org/gdb/current/onlinedocs/gdb.html/Packets.html#thread_002did-syntax","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002161,"name":"Anatoly P","email":"kupokupokupopo@gmail.com","username":"ecco_the_dolphin"},"change_message_id":"00c12fe89251e02379bd4f6ff3dd0146f3f77c94","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"687c55d3_d0470e82","line":3133,"in_reply_to":"56e9e389_032e29cd","updated":"2025-10-24 01:59:22.000000000","message":"\u003e BTW: I re-run ProgramSwWatchpoint test on spike64-2-hwthread.py target 5 times and always passed, as it did in all tests run this morning. Let\u0027s wait what results Anatoly gets..\n\nWell, I\u0027ve re-run these tests once again. 10 times. ProgramSwWatchpoint failed once :( .\n\nThe symptoms for the failure are the same:\n\n```\nw $vCont;s:1;c#c1\n...\nr $T05thread:2;#d8\n...\nw $vCont;s:2;c#c2\n...\nr $T05#b9\n...\nw $vCont;s:1#23\n...\nr $T05thread:2;#d8\n``` \n\nLooks like thread IDs on the respected reply to vCont packet are still messed up.\n\n\nWhen test passes the sequence looks as follows:\n\n```\nw $vCont;s:1;c#c1\n...\nr $T05thread:1;#d7\n...\nw $vCont;s:1;c#c1\n..\nr $T05thread:1;#d7\n...\nw $vCont;s:1;c#c1\n...\nr $T05thread:1;#d7\n```","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002047,"name":"Evgeniy Naydanov","email":"eugnay@gmail.com","username":"en-sc"},"change_message_id":"9a02f0122f1f6e7b9d28c8b450bfd6cc306a21e1","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"aa222b67_61749a41","line":3133,"in_reply_to":"687c55d3_d0470e82","updated":"2025-10-24 09:06:12.000000000","message":"Is it with 9183 or without?","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"27265df8449677f06ab478a5bbd2e6d1b27bdd22","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"41fab532_f02023a9","line":3133,"in_reply_to":"703a936a_ed719778","updated":"2025-10-23 16:23:56.000000000","message":"BTW: I re-run ProgramSwWatchpoint test on spike64-2-hwthread.py target 5 times and always passed, as it did in all tests run this morning. Let\u0027s wait what results Anatoly gets...","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002047,"name":"Evgeniy Naydanov","email":"eugnay@gmail.com","username":"en-sc"},"change_message_id":"24b93d8029d8d454b2c74dfc4e901f7e4ebbea09","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"4347c20d_1b56528e","line":3133,"in_reply_to":"703a936a_ed719778","updated":"2025-10-23 16:53:14.000000000","message":"I\u0027m not quite sure my analysis is correct any more(\n1. 9183 is also needed to fix the GDB assertion failure from in ProgramSwWatchpoint\n2. The assertion failure is reported in GDB Bugzilla:\nhttps://sourceware.org/bugzilla/show_bug.cgi?id\u003d22925\nhttps://sourceware.org/bugzilla/show_bug.cgi?id\u003d26819\n\nI\u0027ll investigate further tomorrow.","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"c98511450cf04fa38467f1dc17cb356a930d8569","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"991940ec_5c307e79","line":3133,"in_reply_to":"70ba8b1b_b1943904","updated":"2025-10-27 14:04:48.000000000","message":"The bug tracking is interesting but probably not relevant: in the meantime\nI recompiled 16.2 to include all architectures of my interest and some of them have OS ABI \"GNU/Linux\" and this version behaves exactly as riscv64-unknown-elf-gdb (built from the same source) did. Perhaps related to https://sourceware.org/bugzilla/show_bug.cgi?id\u003d28440","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002161,"name":"Anatoly P","email":"kupokupokupopo@gmail.com","username":"ecco_the_dolphin"},"change_message_id":"31db70c51417d89deed4633af90d9dbd0800546a","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"19b296e4_7d5aee69","line":3133,"in_reply_to":"71507d44_1440b885","updated":"2025-10-24 17:55:08.000000000","message":"\u003e  Here is the command I used to build gdb:\n\nForgot the actual `make` commands:\n\n```\nmake -j8\nmake install-gdb\n```","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"b92e9df490c12978c38961a27a6ccd809659bd93","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"c6da7c22_b9ccda75","line":3133,"in_reply_to":"9821fa5f_74264b74","updated":"2025-10-24 10:32:15.000000000","message":"\u003e BTW: I re-run ProgramSwWatchpoint test on spike64-2-hwthread.py target 5 times and always passed, as it did in all tests run this morning.\n\nI found that gdb-multiarch in Debian/Ubuntu (any version of 13.1, 15.0.50, 15.1) is broken in that it workarounds vCont;s by setting a BP and using vCont;c - so no wonder I was not able to reproduce the problem.\n\nUnfortunately the binary riscv64-unknown-elf-gdb from sc-dt pack (sc.sc-dt.2025.06-experimental-8dc903a9) 16.2 fails with an internal error\n```\n(gdb) watch counter \u003d\u003d 5\nwatch counter \u003d\u003d 5\n/home/ci_user/.conan2/conan_homes/2.0.9/conan_0/p/b/riscv0aad492f0cbc1/b/gdb/gmp-utils.c:168: internal-error: export_bits: Assertion `word_countp \u003d\u003d 1\u0027 failed.\nA problem internal to GDB has been detected,\nfurther debugging may prove unreliable.\n----- Backtrace -----\n0x5b0230aa7c76 gdb_internal_backtrace_1\n        /home/ci_user/.conan2/conan_homes/2.0.9/conan_0/p/b/riscv0aad492f0cbc1/b/gdb/bt-utils.c:121\n0x5b0230aa7c76 _Z22gdb_internal_backtracev\n        /home/ci_user/.conan2/conan_homes/2.0.9/conan_0/p/b/riscv0aad492f0cbc1/b/gdb/bt-utils.c:182\n...\n```\nWhat gdb version should I use to run riscv-tests?","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002161,"name":"Anatoly P","email":"kupokupokupopo@gmail.com","username":"ecco_the_dolphin"},"change_message_id":"c24d3f95eda8621715e82ebc27226110c474c7a2","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"ee719abc_ba75f601","line":3133,"in_reply_to":"991940ec_5c307e79","updated":"2025-10-28 06:23:17.000000000","message":"Tomas, could you please clarify a little bit on your latest comment?\n\nOn my end it looks like that interaction with vanilla gdb 16.2 compiled only with for riscv target support seems to be fixed by  https://review.openocd.org/c/openocd/+/9183 .\n\nAre you saying that the issue still present in you environment/your compilation options of choice? Or is it gone too?","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"fc79de912a8e5245a1a961a8c7cffa1bf270bf02","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"c379763b_018c52bb","line":3133,"in_reply_to":"9c0e587c_806ef6ff","updated":"2025-11-03 23:06:38.000000000","message":"@Evgeniy, we need this patch. Without it, some tests fails.\nCould you re-consider the score?","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002161,"name":"Anatoly P","email":"kupokupokupopo@gmail.com","username":"ecco_the_dolphin"},"change_message_id":"307b8e611d2bea14159c51b78785a5cae04bc4c4","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"9821fa5f_74264b74","line":3133,"in_reply_to":"aa222b67_61749a41","updated":"2025-10-24 09:15:51.000000000","message":"without. 9183 is marked as Draft, so I decided to ignore it for now.\n\nThe latest one I used was 9176:\n```\ngit fetch https://review.openocd.org/openocd refs/changes/76/9176/3 \u0026\u0026 git checkout -b change-9176 FETCH_HEAD\n```","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002047,"name":"Evgeniy Naydanov","email":"eugnay@gmail.com","username":"en-sc"},"change_message_id":"aed91d99b17202e985cb8d9f048bfc98f1aabc0d","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"36d1308d_c70208d9","line":3133,"in_reply_to":"c379763b_018c52bb","updated":"2025-11-06 10:31:39.000000000","message":"Sure.\nConsidering proposals around re-imagining SMP targets, this can be addressed later.\nI just need to adjust the patch a bit -- I\u0027ve removed the commented-out call to `update_threads()` and the comment no longer makes sense.","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002161,"name":"Anatoly P","email":"kupokupokupopo@gmail.com","username":"ecco_the_dolphin"},"change_message_id":"b8afca2be9a917036130ce3c93355d9f28369934","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"71507d44_1440b885","line":3133,"in_reply_to":"c6da7c22_b9ccda75","updated":"2025-10-24 15:09:49.000000000","message":"\u003e What gdb version should I use to run riscv-tests?\n\nIt\u0027s a shame that sc-dt does not work for you. I would contact you regarding the details of you environment in an email (if you don\u0027t mind) once the dust is settled.\n\nI\u0027ve double-checked and the issue is reproducible with vanilla gdb-16.2 . I used this archive.\n\n```\nhttps://sourceware.org/pub/gdb/releases/gdb-16.2.tar.xz\n```\n\nHere is the command I used to build gdb:\n\n```\n../os_gdb/gdb-16.2/configure \\\n  --target\u003driscv64-unknown-elf \\\n  --host\u003dx86_64-pc-linux-gnu \\\n  --build\u003dx86_64-pc-linux-gnu \\\n  --with-gmp\u003d${PROJECT}/os_gdb/gmp/ \\\n  --with-mpfr\u003d${PROJECT}/os_gdb/mpfr/ \\\n  --with-libexpat-prefix\u003d${PROJECT}/os_gdb/expat \\\n  --prefix\u003d${PROJECT}/os_gdb/install\n```\n\nThough, you\u0027ll need to either build gmp,mpfr,expat or install them in your system.\n\nIf needed,  I think I can provide you with step-by-step guide to build it in docker if you experience any difficulties.","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002047,"name":"Evgeniy Naydanov","email":"eugnay@gmail.com","username":"en-sc"},"change_message_id":"c7c3239332b2c28bae4f452106485a7cf85856ac","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"70ba8b1b_b1943904","line":3133,"in_reply_to":"e9704c42_43b52b82","updated":"2025-10-27 13:10:52.000000000","message":"\u003e BTW gdb-multiarch 16.3 from Debian Trixie workarounds vCont;s by setting a temp BP and vCont;c again as the older versions (13.1, 15.1) did. No idea why, and if there is some wrong patch applied by Debian. Or some problem with multiarch config?\n\nSeems similar to the issue described in this comment: https://sourceware.org/bugzilla/show_bug.cgi?id\u003d26819#c38\n\n\u003e The problem turns out to be that by default the all-targets GDB would default to osabi GNU/Linux, while the riscv-elf GDB does not include Linux support, so defaults to osabi \u0027none\u0027.  If I use the all-targets GDB and explicitly do \u0027set osabi none\u0027 then I can reproduce the failure with all-targets GDB.","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1000687,"name":"Tomas Vanek","display_name":"Tomas Vanek","email":"vanekt@fbl.cz","username":"vanekt"},"change_message_id":"ff418ab43af5f498269c7a567b3bd2082b64d23a","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"9c0e587c_806ef6ff","line":3133,"in_reply_to":"ee719abc_ba75f601","updated":"2025-10-29 14:37:15.000000000","message":"My previous comment was only about the relevance of the findings from  https://sourceware.org/bugzilla/show_bug.cgi?id\u003d26819#c38\nFor riscv-tests the vanilla gdb 16.2 works regardless it was build for the single target riscv64-unknown-elf-gdb or in a multiarch build with other targets bringing OS ABI \"GNU/Linux\"","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"},{"author":{"_account_id":1002161,"name":"Anatoly P","email":"kupokupokupopo@gmail.com","username":"ecco_the_dolphin"},"change_message_id":"93bf05293461e2087e04288cb74fede028e966be","unresolved":true,"context_lines":[{"line_number":3130,"context_line":"\t\t\t * qXfer:threads:read::0,fff$"},{"line_number":3131,"context_line":"\t\t\t * P8\u003dcc060607eb89ca7f$\t\t# write r8 on other thread"},{"line_number":3132,"context_line":"\t\t\t * g$"},{"line_number":3133,"context_line":"\t\t\t * */"},{"line_number":3134,"context_line":""},{"line_number":3135,"context_line":"\t\t\ttarget-\u003ertos-\u003egdb_target_for_threadid(connection, thread_id, \u0026ct);"},{"line_number":3136,"context_line":""}],"source_content_type":"text/x-csrc","patch_set":3,"id":"e9704c42_43b52b82","line":3133,"in_reply_to":"f4a99824_dc33259a","updated":"2025-10-25 07:24:19.000000000","message":"\u003e Looks promising.\n\nI\u0027ve run all tests for all 6 platforms **40 times**. No failures are observed (aside from FreeRTOS). Looks promising indeed :).\n\n\u003e BTW gdb-multiarch 16.3 from Debian Trixie workarounds vCont;s by setting a temp BP and vCont;c again as the older versions (13.1, 15.1) did. No idea why, and if there is some wrong patch applied by Debian. Or some problem with multiarch config?\n\nAt the moment I can\u0027t provide a definitive answer, since internally we haven\u0027t switched to 16.3 and the respected transition is still in my TOOD list.","commit_id":"7533f365976c16b2b1dc917a0ddf633b85250a73"}]}