{"thread":{"id":"41527","subject":"What's cooking in git.git (Feb 2016, #07; Thu, 25)","startedAt":"2016-02-25T23:14:21Z","lastAt":"2016-02-29T23:37:40Z","messageCount":14,"participants":["Junio C Hamano","Torsten Bögershausen","Jeff King","tboegi@web.de","Eric Sunshine"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"279433","messageId":"xmqqfuwgmlgy.fsf@gitster.mtv.corp.google.com","threadId":"41527","inReplyTo":null,"subject":"What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-02-25T23:14:21Z","receivedAt":"2016-02-25T23:14:21Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Here are the topics that have been cooking.  Commits prefixed with\n'-' are only in 'pu' (proposed updates) while commits prefixed with\n'+' are in 'next'.  The ones marked with '.' do not appear in any of\nthe integration branches, but I am still holding onto them.\n\nYou can find the changes described here in the integration branches of the\nrepositories listed at\n\n    http://git-blame.blogspot.com/p/git-public-repositories.html\n\n--------------------------------------------------\n[Graduated to \"master\"]\n\n* bc/http-empty-auth (2016-02-15) 1 commit\n  (merged to 'next' on 2016-02-18 at 6d2fc88)\n + http: add option to try authentication without username\n\n Some authentication methods do not need username or password, but\n libcurl needs some hint that it needs to perform authentication.\n Supplying an empty username and password string is a valid way to\n do so, but you can set the http.[<url>.]emptyAuth configuration\n variable to achieve the same, if you find it cleaner.\n\n\n* ce/https-public-key-pinning (2016-02-15) 1 commit\n  (merged to 'next' on 2016-02-18 at f382cf0)\n + http: implement public key pinning\n\n You can now set http.[<url>.]pinnedpubkey to specify the pinned\n public key when building with recent enough versions of libcURL.\n\n\n* ew/force-ipv4 (2016-02-12) 1 commit\n  (merged to 'next' on 2016-02-18 at d0211b0)\n + connect & http: support -4 and -6 switches for remote operations\n\n \"git fetch\" and friends that make network connections can now be\n told to only use ipv4 (or ipv6).\n\n\n* jc/am-i-v-fix (2016-02-17) 3 commits\n  (merged to 'next' on 2016-02-18 at 331e518)\n + am -i: fix \"v\"iew\n + pager: factor out a helper to prepare a child process to run the pager\n + pager: lose a separate argv[]\n\n The \"v(iew)\" subcommand of the interactive \"git am -i\" command was\n broken in 2.6.0 timeframe when the command was rewritten in C.\n\n\n* jk/lose-name-path (2016-02-12) 5 commits\n  (merged to 'next' on 2016-02-18 at 85831a6)\n + list-objects: pass full pathname to callbacks\n + list-objects: drop name_path entirely\n + list-objects: convert name_path to a strbuf\n + show_object_with_name: simplify by using path_name()\n + http-push: stop using name_path\n\n The \"name_path\" API was an attempt to reduce the need to construct\n the full path out of a series of path components while walking a\n tree hierarchy, but over time made less efficient because the path\n needs to be flattened, e.g. to be compared with another path that\n is already flat.  The API has been removed and its users have been\n rewritten to simplify the overall code complexity.\n\n\n* nd/dwim-wildcards-as-pathspecs (2016-02-10) 3 commits\n  (merged to 'next' on 2016-02-18 at fc7caf4)\n + get_sha1: don't die() on bogus search strings\n + check_filename: tighten dwim-wildcard ambiguity\n + checkout: reorder check_filename conditional\n\n \"git show 'HEAD:Foo[BAR]Baz'\" did not interpret the argument as a\n rev, i.e. the object named by the the pathname with wildcard\n characters in a tree object.\n\n\n* nd/exclusion-regression-fix (2016-02-15) 4 commits\n  (merged to 'next' on 2016-02-18 at 85412ab)\n + dir.c: don't exclude whole dir prematurely\n + dir.c: support marking some patterns already matched\n + dir.c: support tracing exclude\n + dir.c: fix match_pathname()\n\n Another try to add support to the ignore mechanism that lets you\n say \"this is excluded\" and then later say \"oh, no, this part (that\n is a subset of the previous part) is not excluded\".\n\n\n* nd/git-common-dir-fix (2016-02-12) 1 commit\n  (merged to 'next' on 2016-02-18 at 957afc5)\n + rev-parse: take prefix into account in --git-common-dir\n\n \"git rev-parse --git-common-dir\" used in the worktree feature\n misbehaved when run from a subdirectory.\n\n\n* nd/worktree-add-B (2016-02-15) 2 commits\n  (merged to 'next' on 2016-02-18 at bf77016)\n + worktree add -B: do the checkout test before update branch\n + worktree: fix \"add -B\"\n\n \"git worktree add -B <branchname>\" did not work.\n\n\n* sp/remote-curl-ssl-strerror (2016-02-15) 1 commit\n  (merged to 'next' on 2016-02-18 at 1339fe4)\n + remote-curl: include curl_errorstr on SSL setup failures\n\n Help those who debug http(s) part of the system.\n\n--------------------------------------------------\n[New Topics]\n\n* jg/credential-cache-chdir-to-sockdir (2016-02-23) 3 commits\n  (merged to 'next' on 2016-02-25 at ad35811)\n + credential-cache--daemon: change to the socket dir on startup\n + credential-cache--daemon: disallow relative socket path\n + credential-cache--daemon: refactor check_socket_directory\n\n The \"credential-cache\" daemon process used to run in whatever\n directory it happened to start in, but this made umount(2)ing the\n filesystem that houses the repository harder; now the process\n chdir()s to the directory that house its own socket on startup.\n\n Will merge to 'master'.\n\n\n* jk/no-diff-emit-common (2016-02-22) 3 commits\n  (merged to 'next' on 2016-02-24 at 72626ea)\n + xdiff: drop XDL_EMIT_COMMON\n + merge-tree: drop generate_common strategy\n + merge-one-file: use empty blob for add/add base\n\n \"git merge-tree\" used to mishandle \"both sides added\" conflict with\n its own \"create a fake ancestor file that has the common parts of\n what both sides have added and do a 3-way merge\" logic; this has\n been updated to use the usual \"3-way merge with an empty blob as\n the fake common ancestore file\" approach used in the rest of the\n system.\n\n Will merge to 'master'.\n\n\n* mm/push-default-warning (2016-02-25) 1 commit\n  (merged to 'next' on 2016-02-25 at 753352f)\n + push: remove \"push.default is unset\" warning message\n\n Across the transition at around Git version 2.0, the user used to\n get a pretty loud warning when running \"git push\" without setting\n push.default configuration variable.  We started to tone down the\n warning message, given that the transition is over long time ago.\n\n Will merge to 'master'.\n\n\n* mm/push-simple-doc (2016-02-23) 1 commit\n  (merged to 'next' on 2016-02-25 at dc4f57c)\n + Documentation/git-push: document that 'simple' is the default\n\n The documentation did not clearly state that the 'simple' mode is\n now the default for \"git push\" when push.default configuration is\n not set.\n\n Will merge to 'master'.\n\n\n* nd/shallow-deepen (2016-02-23) 25 commits\n - fetch, upload-pack: --deepen=N extends shallow boundary by N commits\n - upload-pack: add get_reachable_list()\n - upload-pack: split check_unreachable() in two, prep for get_reachable_list()\n - t5500, t5539: tests for shallow depth excluding a ref\n - clone: define shallow clone boundary with --shallow-exclude\n - fetch: define shallow boundary with --shallow-exclude\n - upload-pack: support define shallow boundary by excluding revisions\n - refs: add expand_ref()\n - t5500, t5539: tests for shallow depth since a specific date\n - clone: define shallow clone boundary based on time with --shallow-since\n - fetch: define shallow boundary with --shallow-since\n - upload-pack: add deepen-since to cut shallow repos based on time\n - shallow.c: implement a generic shallow boundary finder based on rev-list\n - fetch-pack: use a separate flag for fetch in deepening mode\n - fetch-pack: use a common function for verbose printing\n - fetch-pack: use skip_prefix() instead of starts_with()\n - upload-pack: move rev-list code out of check_non_tip()\n - upload-pack: tighten number parsing at \"deepen\" lines\n - upload-pack: use skip_prefix() instead of starts_with()\n - upload-pack: move \"unshallow\" sending code out of deepen()\n - upload-pack: remove unused variable \"backup\"\n - upload-pack: move \"shallow\" sending code out of deepen()\n - upload-pack: move shallow deepen code out of receive_needs()\n - transport-helper.c: refactor set_helper_option()\n - remote-curl.c: convert fetch_git() to use argv_array\n\n The existing \"git fetch --depth=<n>\" option was hard to use\n correctly when making the history of an existing shallow clone\n deeper.  A new option, \"--deepen=<n>\", has been added to make this\n easier to use.  \"git clone\" also learned \"--shallow-since=<date>\"\n and \"--shallow-exclude=<tag>\" options to make it easier to specify\n \"I am interested only in the recent N months worth of history\" and\n \"Give me only the history since that version\".\n\n Needs review.\n\n\n* ps/plug-xdl-merge-leak (2016-02-23) 1 commit\n  (merged to 'next' on 2016-02-25 at 523609e)\n + xdiff/xmerge: fix memory leak in xdl_merge\n\n Will merge to 'master'.\n\n\n* jk/epipe-in-async (2016-02-25) 4 commits\n  (merged to 'next' on 2016-02-25 at 4482ddf)\n + t5504: handle expected output from SIGPIPE death\n + test_must_fail: report number of unexpected signal\n + fetch-pack: ignore SIGPIPE in sideband demuxer\n + write_or_die: handle EPIPE in async threads\n\n Handling of errors while writing into our internal asynchronous\n process has been made more robust, which reduces flakiness in our\n tests.\n\n Will merge to 'master'.\n\n\n* js/config-set-in-non-repository (2016-02-25) 1 commit\n  (merged to 'next' on 2016-02-25 at a675267)\n + git config: report when trying to modify a non-existing repo config\n\n \"git config section.var value\" to set a value in per-repository\n configuration file should have failed when it was run outside any\n repository, but didn't.\n\n Will merge to 'master'.\n\n\n* sb/submodule-module-list-fix (2016-02-24) 1 commit\n  (merged to 'next' on 2016-02-25 at 21be607)\n + submodule helper list: respect correct path prefix\n\n A helper function \"git submodule\" uses since v2.7.0 to list the\n modules that match the pathspec argument given to its subcommands\n (e.g. \"submodule add <repo> <path>\") has been fixed.\n\n Will merge to 'master'.\n\n\n* sb/submodule-fetch-nontip (2016-02-24) 1 commit\n  (merged to 'next' on 2016-02-25 at 1e412f0)\n + submodule: try harder to fetch needed sha1 by direct fetching sha1\n\n When \"git submodule update\" did not result in fetching the commit\n object in the submodule that is referenced by the superproject, the\n command learned to retry another fetch, specifically asking for\n that commit that may not be connected to the refs it usually\n fetches.\n\n Will merge to 'master'.\n\n\n* dt/initial-ref-xn-commit-doc (2016-02-25) 1 commit\n  (merged to 'next' on 2016-02-25 at e32d0f6)\n + refs: document transaction semantics\n\n Will merge to 'master'.\n\n\n* jk/pack-idx-corruption-safety (2016-02-25) 3 commits\n - use_pack: handle signed off_t overflow\n - nth_packed_object_offset: bounds-check extended offset\n - t5313: test bounds-checks of corrupted/malicious pack/idx files\n\n The code to read the pack data using the offsets stored in the pack\n idx file has been made more carefully check the validity of the\n data in the idx.\n\n Will merge to 'next'.\n\n\n* ls/makefile-cflags-developer-tweak (2016-02-25) 1 commit\n  (merged to 'next' on 2016-02-25 at 14bb065)\n + add DEVELOPER makefile knob to check for acknowledged warnings\n\n There is a new DEVELOPER knob that enables many compiler warning\n options in the Makefile.\n\n Will merge to 'master'.\n\n\n* ma/update-hooks-sample-typofix (2016-02-25) 1 commit\n  (merged to 'next' on 2016-02-25 at be49c79)\n + templates/hooks: fix minor typo in the sample update-hook\n\n Will merge to 'master'.\n\n\n* mm/diff-renames-default (2016-02-25) 5 commits\n  (merged to 'next' on 2016-02-25 at 947c399)\n + diff: activate diff.renames by default\n + log: introduce init_log_defaults()\n + t: add tests for diff.renames (true/false/unset)\n + t4001-diff-rename: wrap file creations in a test\n + Documentation/diff-config: fix description of diff.renames\n\n The end-user facing Porcelain level commands like \"diff\" and \"log\"\n now enables the rename detection by default.\n\n Will merge to 'master'.\n\n\n* mm/readme-markdown (2016-02-25) 5 commits\n  (merged to 'next' on 2016-02-25 at b8bda29)\n + README.md: move down historical explanation about the name\n + README.md: don't call git stupid in the title\n + README.md: move the link to git-scm.com up\n + README.md: add hyperlinks on filenames\n + README: use markdown syntax\n\n Move README to README.md and tweak its contents slightly to make it\n easier on the eyes.\n\n Will merge to 'master'.\n\n\n* mp/upload-pack-use-embedded-args (2016-02-25) 1 commit\n - upload-pack: use argv_array for pack_objects\n\n The embedded args argv-array in the child process is used to build\n the command line to run pack-objects instead of using a separate\n array of strings.\n\n Will merge to 'next'.\n\n--------------------------------------------------\n[Stalled]\n\n* ec/annotate-deleted (2015-11-20) 1 commit\n - annotate: skip checking working tree if a revision is provided\n\n Usability fix for annotate-specific \"<file> <rev>\" syntax with deleted\n files.\n\n Waiting for review.\n\n\n* dg/subtree-rebase-test (2016-01-19) 1 commit\n - contrib/subtree: Add a test for subtree rebase that loses commits\n\n Reviewed up to v5.\n Will be rerolled.\n ($gmane/284426)\n\n\n* js/am-3-merge-recursive-direct (2015-10-12) 2 commits\n - am: make a direct call to merge_recursive\n - merge_recursive_options: introduce the \"gently\" flag\n\n The merge_recursive_generic() function has been made a bit safer to\n call from inside a process.  \"git am -3\" was taught to make a direct\n call to the function when falling back to three-way merge.\n\n Being able to make a direct call would be good in general, but as a\n performance thing, the change needs to be backed up by numbers.\n\n Needs review.\n\n I haven't gone through the \"gently\" change with fine toothed comb;\n I can see that the change avoids calling die(), but I haven't made\n sure that the program states (e.g. what's in the in-core index) are\n adjusted sensibly when it returns to the caller instead of dying,\n or the codepaths that used to die() are free of resource leaks.\n The original code certainly did not care the program states at the\n point of dying exactly because it knew it is going to exit, but now\n they have to care, and they need to be audited.\n\n\n* dk/gc-more-wo-pack (2016-01-13) 4 commits\n - gc: clean garbage .bitmap files from pack dir\n - t5304: ensure non-garbage files are not deleted\n - t5304: test .bitmap garbage files\n - prepare_packed_git(): find more garbage\n\n Follow-on to dk/gc-idx-wo-pack topic, to clean up stale\n .bitmap and .keep files.\n\n Waiting for a reroll.\n ($gmane/284368).\n\n\n* mg/httpd-tests-update-for-apache-2.4 (2015-04-08) 2 commits\n - t/lib-git-svn: check same httpd module dirs as lib-httpd\n - t/lib-httpd: load mod_unixd\n\n This is the first two commits in a three-patch series $gmane/266962\n\n Becoming tired of waiting for a reroll.\n with updated log message ($gmane/268061).\n Will discard.\n\n\n* jc/diff-b-m (2015-02-23) 5 commits\n . WIPWIP\n . WIP: diff-b-m\n - diffcore-rename: allow easier debugging\n - diffcore-rename.c: add locate_rename_src()\n - diffcore-break: allow debugging\n\n \"git diff -B -M\" produced incorrect patch when the postimage of a\n completely rewritten file is similar to the preimage of a removed\n file; such a resulting file must not be expressed as a rename from\n other place.\n\n The fix in this patch is broken, unfortunately.\n Will discard.\n\n--------------------------------------------------\n[Cooking]\n\n* sb/submodule-init (2016-02-25) 2 commits\n - submodule: port init from shell to C\n - submodule: port resolve_relative_url from shell to C\n (this branch uses sb/submodule-parallel-update; is tangled with dt/refs-backend-lmdb.)\n\n Update of \"git submodule\" to move pieces of logic to C continues.\n\n Comments?\n ($gmane/286092)\n\n\n* ak/extract-argv0-last-dir-sep (2016-02-19) 1 commit\n  (merged to 'next' on 2016-02-24 at f590eb0)\n + exec_cmd.c: use find_last_dir_sep() for code simplification\n\n Code simplification.\n\n Will merge to 'master'.\n\n\n* az/p4-bare-no-rebase (2016-02-19) 1 commit\n - git-p4.py: Don't try to rebase on submit from bare repository\n\n \"git p4 submit\" attempts to do a rebase, which would fail if done\n in a bare repository.  Not doing this rebase would paper over the\n failure, which is what this patch does, but it is unclear what the\n side effect of not rebasing is.\n\n Needs a better explanation.\n\n\n* ak/git-strip-extension-from-dashed-command (2016-02-21) 1 commit\n  (merged to 'next' on 2016-02-24 at 8a84f91)\n + git.c: simplify stripping extension of a file in handle_builtin()\n\n Code simplification.\n\n Will merge to 'master'.\n\n\n* jk/grep-binary-workaround-in-test (2016-02-23) 2 commits\n  (merged to 'next' on 2016-02-25 at 0e96ae8)\n + t9200: avoid grep on non-ASCII data\n + t8005: avoid grep on non-ASCII data\n\n Recent versions of GNU grep are pickier when their input contains\n arbitrary binary data, which some of our tests uses.  Rewrite the\n tests to sidestep the problem.\n\n Will merge to 'master'.\n\n\n* mg/work-tree-tests (2016-02-20) 1 commit\n  (merged to 'next' on 2016-02-25 at ca8cbb0)\n + tests: rename work-tree tests to *work-tree*\n\n Traditionally, the tests that try commands that work on the\n contents in the working tree were named with \"worktree\" in their\n filenames, but with the recent addition of \"git worktree\"\n subcommand, whose tests are also named similarly, it has become\n harder to tell them apart.  The traditional tests have been renamed\n to use \"work-tree\" instead in an attempt to differentiate them.\n\n Will merge to 'master'.\n\n\n* jk/more-comments-on-textconv (2016-02-22) 1 commit\n  (merged to 'next' on 2016-02-24 at 4ccfdc5)\n + diff: clarify textconv interface\n\n The memory ownership rule of fill_textconv() API, which was a bit\n tricky, has been documented a bit better.\n\n Will merge to 'master'.\n\n\n* cn/deprecate-ssh-git-url (2016-02-15) 1 commit\n - Disown ssh+git and git+ssh\n\n Needs toning down by replacing the condescendence with an\n explanation and enlightenment.\n ($gmane/286222)\n\n\n* fa/merge-recursive-no-rename (2016-02-24) 7 commits\n  (merged to 'next' on 2016-02-25 at 8f18cc1)\n + t3034: test deprecated interface\n + t3034: test option to disable renames\n + t3034: add rename threshold tests\n + merge-recursive: find-renames resets threshold\n + merge-strategies.txt: fix typo\n  (merged to 'next' on 2016-02-18 at dd6125b)\n + merge-recursive: more consistent interface\n + merge-recursive: option to disable renames\n\n \"git merge-recursive\" learned \"--no-renames\" option to disable its\n rename detection logic.\n\n Will merge to 'master'.\n\n\n* jk/tighten-alloc (2016-02-22) 22 commits\n  (merged to 'next' on 2016-02-24 at 78b3251)\n + ewah: convert to REALLOC_ARRAY, etc\n + convert ewah/bitmap code to use xmalloc\n + diff_populate_gitlink: use a strbuf\n + transport_anonymize_url: use xstrfmt\n + git-compat-util: drop mempcpy compat code\n + sequencer: simplify memory allocation of get_message\n + test-path-utils: fix normalize_path_copy output buffer size\n + fetch-pack: simplify add_sought_entry\n + fast-import: simplify allocation in start_packfile\n + write_untracked_extension: use FLEX_ALLOC helper\n + prepare_{git,shell}_cmd: use argv_array\n + use st_add and st_mult for allocation size computation\n + convert trivial cases to FLEX_ARRAY macros\n + use xmallocz to avoid size arithmetic\n + convert trivial cases to ALLOC_ARRAY\n + convert manual allocations to argv_array\n + argv-array: add detach function\n + add helpers for allocating flex-array structs\n + harden REALLOC_ARRAY and xcalloc against size_t overflow\n + tree-diff: catch integer overflow in combine_diff_path allocation\n + add helpers for detecting size_t overflow\n + reflog_expire_cfg: NUL-terminate pattern field\n\n Update various codepaths to avoid manually-counted malloc().\n\n Will merge to 'master'.\n\n\n* ps/config-error (2016-02-22) 15 commits\n  (merged to 'next' on 2016-02-25 at b378629)\n + config: rename git_config_set_or_die to git_config_set\n + config: rename git_config_set to git_config_set_gently\n + compat: die when unable to set core.precomposeunicode\n + sequencer: die on config error when saving replay opts\n + init-db: die on config errors when initializing empty repo\n + clone: die on config error in cmd_clone\n + remote: die on config error when manipulating remotes\n + remote: die on config error when setting/adding branches\n + remote: die on config error when setting URL\n + submodule--helper: die on config error when cloning module\n + submodule: die on config error when linking modules\n + branch: die on config error when editing branch description\n + branch: die on config error when unsetting upstream\n + branch: report errors in tracking branch setup\n + config: introduce set_or_die wrappers\n\n Many codepaths forget to check return value from git_config_set();\n the function is made to die() to make sure we do not proceed when\n setting a configuration variable failed.\n\n Will merge to 'master'.\n\n\n* ss/commit-dry-run-resolve-merge-to-no-op (2016-02-17) 1 commit\n - wt-status.c: set commitable bit if there is a meaningful merge.\n\n \"git commit --dry-run\" reported \"No, no, you cannot commit.\" in one\n case where \"git commit\" would have allowed you to commit, and this\n improves it a little bit (\"git commit --dry-run --short\" still does\n not give you the correct answer, for example).\n\n\n* tg/git-remote (2016-02-16) 4 commits\n  (merged to 'next' on 2016-02-18 at 091d43e)\n + remote: use remote_is_configured() for add and rename\n + remote: actually check if remote exits\n + remote: simplify remote_is_configured()\n + remote: use parse_config_key\n\n The internal API to interact with \"remote.*\" configuration\n variables has been streamlined.\n\n Will merge to 'master'.\n\n\n* ls/config-origin (2016-02-22) 4 commits\n  (merged to 'next' on 2016-02-24 at 239fcc2)\n + config: add '--show-origin' option to print the origin of a config value\n + config: add 'origin_type' to config_source struct\n + rename git_config_from_buf to git_config_from_mem\n + t: do not hide Git's exit code in tests using 'nul_to_q'\n\n The configuration system has been taught to phrase where it found a\n bad configuration variable in a better way in its error messages.\n \"git config\" learnt a new \"--show-origin\" option to indicate where\n the values come from.\n\n Will merge to 'master'.\n\n\n* nd/icase (2016-02-15) 12 commits\n - grep.c: reuse \"icase\" variable\n - diffcore-pickaxe: support case insensitive match on non-ascii\n - diffcore-pickaxe: \"share\" regex error handling code\n - grep/pcre: support utf-8\n - gettext: add is_utf8_locale()\n - grep/pcre: prepare locale-dependent tables for icase matching\n - grep/icase: avoid kwsset when -F is specified\n - grep/icase: avoid kwsset on literal non-ascii strings\n - test-regex: expose full regcomp() to the command line\n - test-regex: isolate the bug test code\n - grep: break down an \"if\" stmt in preparation for next changes\n - grep: allow -F -i combination\n\n \"git grep -i\" has been taught to fold case in non-ascii locales.\n\n Needs review.\n ($gmane/286137)\n\n\n* tb/conversion (2016-02-23) 7 commits\n  (merged to 'next' on 2016-02-25 at 1bb466d)\n + convert.c: correct attr_action()\n  (merged to 'next' on 2016-02-12 at 6faf27b)\n + convert.c: simplify text_stat\n + convert.c: refactor crlf_action\n + convert.c: use text_eol_is_crlf()\n + convert.c: remove input_crlf_action()\n + convert.c: remove unused parameter 'path'\n + t0027: add tests for get_stream_filter()\n\n Code simplification.\n\n Will merge to 'master'.\n\n\n* sb/submodule-parallel-update (2016-02-25) 11 commits\n - clone: allow an explicit argument for parallel submodule clones\n - submodule update: expose parallelism to the user\n - git submodule update: have a dedicated helper for cloning\n - run_processes_parallel: correctly terminate callbacks with an LF\n - run_processes_parallel: rename parameters for the callbacks\n - run-command: expose default_{start_failure, task_finished}\n - run_processes_parallel: treat output of children as byte array\n - submodule update: direct error message to stderr\n - fetching submodules: respect `submodule.fetchJobs` config option\n - submodule-config: drop check against NULL\n - submodule-config: keep update strategy around\n (this branch is used by dt/refs-backend-lmdb and sb/submodule-init.)\n\n A major part of \"git submodule update\" has been ported to C to take\n advantage of the recently added framework to run download tasks in\n parallel.\n\n Seems to break tests when merged to the tip of 'pu'.\n\n\n* dt/refs-backend-lmdb (2016-02-25) 34 commits\n - SQUASH??? Minimum compilation band-aid\n - tests: add ref-storage argument\n - refs: tests for lmdb backend\n - refs: add LMDB refs storage backend\n - refs: break out resolve_ref_unsafe_submodule\n - config: read ref storage config on startup\n - refs: register ref storage backends\n - svn: learn ref-storage argument\n - clone: allow ref storage backend to be set for clone\n - refs: check submodules' ref storage config\n - init: allow alternate ref strorage to be set for new repos\n - refs: always handle non-normal refs in files backend\n - refs: resolve symbolic refs first\n - refs: on symref reflog expire, lock symref not referrent\n - refs: don't dereference on rename\n - refs: allow log-only updates\n - refs: move duplicate check to common code\n - refs: make lock generic\n - refs: handle non-normal ref renames\n - refs: add method to rename refs\n - refs: add methods to init refs db\n - refs: add method for delete_refs\n - refs: add method for initial ref transaction commit\n - refs: add methods for reflog\n - refs: add do_for_each_per_worktree_ref\n - refs: reduce the visibility of do_for_each_ref()\n - refs: add method for do_for_each_ref\n - refs: add methods for misc ref operations\n - refs: add a backend method structure with transaction functions\n - refs: move resolve_ref_unsafe into common code\n - files-backend: break out ref reading\n - refs: move for_each_*ref* functions into common code\n - refs: move head_ref{,_submodule} to the common code\n - Merge branch 'sb/submodule-parallel-update' into dt/refs-backend-lmdb\n (this branch uses sb/submodule-parallel-update; is tangled with sb/submodule-init.)\n\n Building on top of a few refs-backend preparatory series, LMDB\n based refs backend has been plugged into the system.\n\n ($gmane/286572)\n\n\n* kn/ref-filter-atom-parsing (2016-02-17) 11 commits\n  (merged to 'next' on 2016-02-18 at 9606218)\n + ref-filter: introduce objectname_atom_parser()\n + ref-filter: introduce contents_atom_parser()\n + ref-filter: introduce remote_ref_atom_parser()\n + ref-filter: align: introduce long-form syntax\n + ref-filter: introduce align_atom_parser()\n + ref-filter: introduce parse_align_position()\n + ref-filter: introduce color_atom_parser()\n + ref-filter: introduce parsing functions for each valid atom\n + ref-filter: introduce struct used_atom\n + ref-filter: bump 'used_atom' and related code to the top\n + ref-filter: use string_list_split over strbuf_split\n\n The ref-filter's format-parsing code has been refactored, in\n preparation for \"branch --format\" and friends.\n\n Will merge to 'master'.\n\n\n* ad/cygwin-wants-rename (2015-08-07) 1 commit\n - config.mak.uname: Cygwin needs OBJECT_CREATION_USES_RENAMES\n\n Will hold.\n ($gmane/275680).\n\n\n* jc/rerere-multi (2016-02-08) 7 commits\n - rerere: do use multiple variants\n - t4200: rerere a merge with two identical conflicts\n - rerere: allow multiple variants to exist\n - rerere: delay the recording of preimage\n - rerere: handle leftover rr-cache/$ID directory and postimage files\n - rerere: scan $GIT_DIR/rr-cache/$ID when instantiating a rerere_id\n - rerere: split conflict ID further\n\n \"git rerere\" can encounter two or more files with the same conflict\n signature that have to be resolved in different ways, but there was\n no way to record these separate resolutions.\n\n Needs further work on forget/gc.\n\n\n* jc/merge-drop-old-syntax (2015-04-29) 1 commit\n - merge: drop 'git merge <message> HEAD <commit>' syntax\n\n Stop supporting \"git merge <message> HEAD <commit>\" syntax that has\n been deprecated since October 2007.  It has been reported that\n git-gui still uses the deprecated syntax, which needs to be fixed\n before this final step can proceed.\n ($gmane/282594)\n\n--------------------------------------------------\n[Discarded]\n\n* dv/http-try-negotiate-at-least-once (2016-02-02) 1 commit\n . remote-curl: don't fall back to Basic auth if we haven't tried Negotiate\n\n bc/http-empty-auth satisfies the original need that triggered this\n topic better.\n\n\n* ep/format-printf (2016-02-11) 22 commits\n . wt-status.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . utf8.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . transport-helper.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . trace.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . strbuf.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . remote.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . pkt-line.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . merge-recursive.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . imap-send.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . http-backend.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . fsck.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . daemon.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . config.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . color.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . cache.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . builtin/upload-archive.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . builtin/update-index.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . builtin/receive-pack.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . builtin/index-pack.c: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . argv-array.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . advice.h: use the FORMAT_PRINTF macro to declare the gcc function attribute 'format printf'\n . git-compat-util.h: add the FORMAT_PRINTF macro\n\n Replace __attribute__((format(printf, a, b))) with\n FORMAT_PRINTF(a,b) everywhere in the code.\n\n It would be a different story if another compiler had the same\n feature implemented using a different syntax, in which case we\n could use conditionally #define what FORMAT_PRINTF(a,b) expands to,\n but because we do not have such a compiler supported, this change\n is a net negative, churning the code without buying us anything\n (yet).\n\n\n* jk/merge-tree-merge-blobs (2016-02-16) 1 commit\n . merge_blobs: use strbuf instead of manually-sized mmfile_t\n\n \"git merge-tree\" (a throw-away demonstration) did not work very\n well when merging \"both sides added a new file at the same path\"\n case.\n\n Solved differently by jk/no-diff-emit-common topic.\n"},{"id":"279526","messageId":"56D078F3.9070905@web.de","threadId":"41527","inReplyTo":"xmqqfuwgmlgy.fsf@gitster.mtv.corp.google.com","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Torsten Bögershausen","fromEmail":"tboegi@web.de","sentAt":"2016-02-26T16:10:27Z","receivedAt":"2016-02-26T16:10:27Z","isPatch":false,"sender":{"key":"tboegi@web.de","avatar":"https://avatars.githubusercontent.com/u/7138363?v=4"},"body":"\n> * jk/tighten-alloc (2016-02-22) 22 commits\n>   (merged to 'next' on 2016-02-24 at 78b3251)\n>  + ewah: convert to REALLOC_ARRAY, etc\n>  + convert ewah/bitmap code to use xmalloc\n>  + diff_populate_gitlink: use a strbuf\n>  + transport_anonymize_url: use xstrfmt\n>  + git-compat-util: drop mempcpy compat code\n>  + sequencer: simplify memory allocation of get_message\n>  + test-path-utils: fix normalize_path_copy output buffer size\n>  + fetch-pack: simplify add_sought_entry\n>  + fast-import: simplify allocation in start_packfile\n>  + write_untracked_extension: use FLEX_ALLOC helper\n>  + prepare_{git,shell}_cmd: use argv_array\n>  + use st_add and st_mult for allocation size computation\n>  + convert trivial cases to FLEX_ARRAY macros\n>  + use xmallocz to avoid size arithmetic\n>  + convert trivial cases to ALLOC_ARRAY\n>  + convert manual allocations to argv_array\n>  + argv-array: add detach function\n>  + add helpers for allocating flex-array structs\n>  + harden REALLOC_ARRAY and xcalloc against size_t overflow\n>  + tree-diff: catch integer overflow in combine_diff_path allocation\n>  + add helpers for detecting size_t overflow\n>  + reflog_expire_cfg: NUL-terminate pattern field\n> \n>  Update various codepaths to avoid manually-counted malloc().\n> \n>  Will merge to 'master'.\n> \n\nSorry being late,\nbut this is what my older test system says:\n\nCC combine-diff.o\ncombine-diff.c: In function ‘diff_tree_combined’:\ncombine-diff.c:1391: internal compiler error: Segmentation fault\nPlease submit a full bug report,\nwith preprocessed source if appropriate.\nSee <URL:http://developer.apple.com/bugreporter> for instructions.\nmake: *** [combine-diff.o] Error 1\n\nRevert \"use st_add and st_mult for allocation size computation\" didn't fix it.\nI haven't digged deeper yet.\n"},{"id":"279547","messageId":"xmqqsi0fjpez.fsf@gitster.mtv.corp.google.com","threadId":"41527","inReplyTo":"56D078F3.9070905@web.de","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-02-26T18:29:40Z","receivedAt":"2016-02-26T18:29:40Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Torsten Bögershausen <tboegi@web.de> writes:\n\n> CC combine-diff.o\n> combine-diff.c: In function ‘diff_tree_combined’:\n> combine-diff.c:1391: internal compiler error: Segmentation fault\n> Please submit a full bug report,\n> with preprocessed source if appropriate.\n> See <URL:http://developer.apple.com/bugreporter> for instructions.\n> make: *** [combine-diff.o] Error 1\n>\n> Revert \"use st_add and st_mult for allocation size computation\" didn't fix it.\n> I haven't digged deeper yet.\n\nHmph, I am not quite sure what you meant by \"I haven't digged\ndeeper\"; I do not think you should be debugging Apple's compiler\n(unless you have the source to it and that is your job to do so,\nthat is).\n\nIn any case, merging the topic to 'master', and reverting that\ncommit will leave this only this change between 'master' and the\nresult.  I do not immediately spot anything trickier in the\npostimage compared to the preimage that may trickle a compiler bug,\nbut what would I know without the source ;-)\n\nThanks for a report, anyway.\n\ndiff --git a/combine-diff.c b/combine-diff.c\nindex 5571304..be09a2b 100644\n--- a/combine-diff.c\n+++ b/combine-diff.c\n@@ -319,7 +319,7 @@ static void append_lost(struct sline *sline, int n, const char *line, int len)\n \tif (line[len-1] == '\\n')\n \t\tlen--;\n \n-\tlline = xmalloc(sizeof(*lline) + len + 1);\n+\tFLEX_ALLOC_MEM(lline, line, line, len);\n \tlline->len = len;\n \tlline->next = NULL;\n \tlline->prev = sline->plost.lost_tail;\n@@ -330,8 +330,6 @@ static void append_lost(struct sline *sline, int n, const char *line, int len)\n \tsline->plost.lost_tail = lline;\n \tsline->plost.len++;\n \tlline->parent_map = this_mask;\n-\tmemcpy(lline->line, line, len);\n-\tlline->line[len] = 0;\n }\n \n struct combine_diff_state {\n@@ -1043,7 +1041,7 @@ static void show_patch_diff(struct combine_diff_path *elem, int num_parent,\n \t\t\t\telem->mode = canon_mode(S_IFLNK);\n \n \t\t\tresult_size = len;\n-\t\t\tresult = xmalloc(len + 1);\n+\t\t\tresult = xmallocz(len);\n \n \t\t\tdone = read_in_full(fd, result, len);\n \t\t\tif (done < 0)\n@@ -1051,8 +1049,6 @@ static void show_patch_diff(struct combine_diff_path *elem, int num_parent,\n \t\t\telse if (done < len)\n \t\t\t\tdie(\"early EOF '%s'\", elem->path);\n \n-\t\t\tresult[len] = 0;\n-\n \t\t\t/* If not a fake symlink, apply filters, e.g. autocrlf */\n \t\t\tif (is_file) {\n \t\t\t\tstruct strbuf buf = STRBUF_INIT;\n@@ -1372,7 +1368,7 @@ static struct combine_diff_path *find_paths_multitree(\n \tstruct combine_diff_path paths_head;\n \tstruct strbuf base;\n \n-\tparents_sha1 = xmalloc(nparent * sizeof(parents_sha1[0]));\n+\tALLOC_ARRAY(parents_sha1, nparent);\n \tfor (i = 0; i < nparent; i++)\n \t\tparents_sha1[i] = parents->sha1[i];\n \n@@ -1483,7 +1479,7 @@ void diff_tree_combined(const unsigned char *sha1,\n \tif (opt->orderfile && num_paths) {\n \t\tstruct obj_order *o;\n \n-\t\to = xmalloc(sizeof(*o) * num_paths);\n+\t\tALLOC_ARRAY(o, num_paths);\n \t\tfor (i = 0, p = paths; p; p = p->next, i++)\n \t\t\to[i].obj = p;\n \t\torder_objects(opt->orderfile, path_path, o, num_paths);\n"},{"id":"279574","messageId":"56D0D2DF.1040807@web.de","threadId":"41527","inReplyTo":"xmqqsi0fjpez.fsf@gitster.mtv.corp.google.com","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Torsten Bögershausen","fromEmail":"tboegi@web.de","sentAt":"2016-02-26T22:34:07Z","receivedAt":"2016-02-26T22:34:07Z","isPatch":false,"sender":{"key":"tboegi@web.de","avatar":"https://avatars.githubusercontent.com/u/7138363?v=4"},"body":"On 2016-02-26 19.29, Junio C Hamano wrote:\n> Torsten Bögershausen <tboegi@web.de> writes:\n> \n>> CC combine-diff.o\n>> combine-diff.c: In function ‘diff_tree_combined’:\n>> combine-diff.c:1391: internal compiler error: Segmentation fault\n>> Please submit a full bug report,\n>> with preprocessed source if appropriate.\n>> See <URL:http://developer.apple.com/bugreporter> for instructions.\n>> make: *** [combine-diff.o] Error 1\n>>\n>> Revert \"use st_add and st_mult for allocation size computation\" didn't fix it.\n>> I haven't digged deeper yet.\n> \n> Hmph, I am not quite sure what you meant by \"I haven't digged\n> deeper\"; I do not think you should be debugging Apple's compiler\n> (unless you have the source to it and that is your job to do so,\n> that is).\n> \n> In any case, merging the topic to 'master', and reverting that\n> commit will leave this only this change between 'master' and the\n> result.  I do not immediately spot anything trickier in the\n> postimage compared to the preimage that may trickle a compiler bug,\n> but what would I know without the source ;-)\n> \nDigging means:\nrun git bisect and report the commit.\nAnd this makes the compiler happy:\n  Revert \"tree-diff: catch integer overflow in combine_diff_path allocation\"\n\n    This reverts commit 5b442c4f2723211ce0d862571e88ee206bfd51bf.\n\nTest suite passes,\n( except t9115, the 2 new test cases \"svn.pathnameencoding=cp932\".\n  More info about that later)\n"},{"id":"279584","messageId":"xmqqtwkvhwpd.fsf@gitster.mtv.corp.google.com","threadId":"41527","inReplyTo":"56D0D2DF.1040807@web.de","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-02-26T23:35:10Z","receivedAt":"2016-02-26T23:35:10Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Torsten Bögershausen <tboegi@web.de> writes:\n\n> On 2016-02-26 19.29, Junio C Hamano wrote:\n>> Torsten Bögershausen <tboegi@web.de> writes:\n>> \n>>> CC combine-diff.o\n>>> combine-diff.c: In function ‘diff_tree_combined’:\n>>> combine-diff.c:1391: internal compiler error: Segmentation fault\n>>> Please submit a full bug report,\n>>> with preprocessed source if appropriate.\n>>> See <URL:http://developer.apple.com/bugreporter> for instructions.\n>>> make: *** [combine-diff.o] Error 1\n>>>\n>>> Revert \"use st_add and st_mult for allocation size computation\" didn't fix it.\n>>> I haven't digged deeper yet.\n>> \n>> Hmph, I am not quite sure what you meant by \"I haven't digged\n>> deeper\"; I do not think you should be debugging Apple's compiler\n>> (unless you have the source to it and that is your job to do so,\n>> that is).\n>> \n>> In any case, merging the topic to 'master', and reverting that\n>> commit will leave this only this change between 'master' and the\n>> result.  I do not immediately spot anything trickier in the\n>> postimage compared to the preimage that may trickle a compiler bug,\n>> but what would I know without the source ;-)\n>> \n> Digging means:\n> run git bisect and report the commit.\n> And this makes the compiler happy:\n>   Revert \"tree-diff: catch integer overflow in combine_diff_path allocation\"\n\nSo?\n\nIdentifying two versions of the same C file, for one of which the\nApple compiler does not give you \"internal compiler error\" and for\nthe other it does, may help Apple engineers to debug their compiler.\n\nBut we cannot and will not revert any of our code only because some\nvendor compiler is broken.  It would be a different story if we were\nthrowing an invalid C at compilers and relying on a bug in GCC that\naccepts an invalid code, but I do not think that is the case here.\n\n\n\n>\n>     This reverts commit 5b442c4f2723211ce0d862571e88ee206bfd51bf.\n>\n> Test suite passes,\n> ( except t9115, the 2 new test cases \"svn.pathnameencoding=cp932\".\n>   More info about that later)\n"},{"id":"279599","messageId":"20160227032907.GB10313@sigill.intra.peff.net","threadId":"41527","inReplyTo":"xmqqtwkvhwpd.fsf@gitster.mtv.corp.google.com","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Jeff King","fromEmail":"peff@peff.net","sentAt":"2016-02-27T03:29:08Z","receivedAt":"2016-02-27T03:29:08Z","isPatch":false,"sender":{"key":"peff@peff.net","avatar":"https://avatars.githubusercontent.com/u/45925?v=4"},"body":"On Fri, Feb 26, 2016 at 03:35:10PM -0800, Junio C Hamano wrote:\n\n> > Digging means:\n> > run git bisect and report the commit.\n> > And this makes the compiler happy:\n> >   Revert \"tree-diff: catch integer overflow in combine_diff_path allocation\"\n> \n> So?\n> \n> Identifying two versions of the same C file, for one of which the\n> Apple compiler does not give you \"internal compiler error\" and for\n> the other it does, may help Apple engineers to debug their compiler.\n> \n> But we cannot and will not revert any of our code only because some\n> vendor compiler is broken.  It would be a different story if we were\n> throwing an invalid C at compilers and relying on a bug in GCC that\n> accepts an invalid code, but I do not think that is the case here.\n\nI think knowing the commit that causes the problem is interesting for\nus, because it lets us see whether we are throwing invalid C at the\ncompiler or not (the compiler should never segfault, obviously, but I\ncan believe that it is more likely to if you throw garbage at it).\n\nBut that commit is one of the tamest, I think; it just seems like a\nboring compiler bug.  I'd be curious to reduce it to a minimal change\nthat causes the bug, just to be sure.\n\nTorsten, what is the compiler version (I don't have Apple compilers, but\nit seems plausible that older clang might have the same problem).\n\n-Peff\n"},{"id":"279629","messageId":"56D14C56.6070306@web.de","threadId":"41527","inReplyTo":"20160227032907.GB10313@sigill.intra.peff.net","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Torsten Bögershausen","fromEmail":"tboegi@web.de","sentAt":"2016-02-27T07:12:22Z","receivedAt":"2016-02-27T07:12:22Z","isPatch":false,"sender":{"key":"tboegi@web.de","avatar":"https://avatars.githubusercontent.com/u/7138363?v=4"},"body":"\n\nOn 27.02.16 04:29, Jeff King wrote:\n> On Fri, Feb 26, 2016 at 03:35:10PM -0800, Junio C Hamano wrote:\n>\n>>> Digging means:\n>>> run git bisect and report the commit.\n>>> And this makes the compiler happy:\n>>>   Revert \"tree-diff: catch integer overflow in combine_diff_path allocation\"\n>> So?\n>>\n>> Identifying two versions of the same C file, for one of which the\n>> Apple compiler does not give you \"internal compiler error\" and for\n>> the other it does, may help Apple engineers to debug their compiler.\n>>\n>> But we cannot and will not revert any of our code only because some\n>> vendor compiler is broken.  It would be a different story if we were\n>> throwing an invalid C at compilers and relying on a bug in GCC that\n>> accepts an invalid code, but I do not think that is the case here.\n> I think knowing the commit that causes the problem is interesting for\n> us, because it lets us see whether we are throwing invalid C at the\n> compiler or not (the compiler should never segfault, obviously, but I\n> can believe that it is more likely to if you throw garbage at it).\n>\n> But that commit is one of the tamest, I think; it just seems like a\n> boring compiler bug.  I'd be curious to reduce it to a minimal change\n> that causes the bug, just to be sure.\n>\n> Torsten, what is the compiler version (I don't have Apple compilers, but\n> it seems plausible that older clang might have the same problem).\n>\n> -Peff\nThat's machine is running Mac OS X 10.6, which is no longer supported\nwith updates.\n\n gcc --version\ni686-apple-darwin10-gcc-4.2.1 (GCC) 4.2.1 (Apple Inc. build 5666) (dot 3)\nCopyright (C) 2007 Free Software Foundation, Inc.\nThis is free software; see the source for copying conditions.  There is NO\nwarranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\nThere seems to be a clang as well:\nclang --version\nApple clang version 1.7 (tags/Apple/clang-77) (based on LLVM 2.9svn)\nTarget: x86_64-apple-darwin10\nThread model: posix\n\n\n\nIf that information is helpful:\nOn later versions of Mac OS X (and/or Xcode) gcc is a wrapper to clang,\nand this is what I see under Mac OS 10.9:\ngcc --version\nConfigured with: --prefix=/Library/Developer/CommandLineTools/usr --with-gxx-include-dir=/usr/include/c++/4.2.1\nApple LLVM version 6.0 (clang-600.0.57) (based on LLVM 3.5svn)\nTarget: x86_64-apple-darwin13.4.0\nThread model: posix\n"},{"id":"279648","messageId":"56D1DF53.1010006@web.de","threadId":"41527","inReplyTo":"56D14C56.6070306@web.de","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Torsten Bögershausen","fromEmail":"tboegi@web.de","sentAt":"2016-02-27T17:39:31Z","receivedAt":"2016-02-27T17:39:31Z","isPatch":false,"sender":{"key":"tboegi@web.de","avatar":"https://avatars.githubusercontent.com/u/7138363?v=4"},"body":"How about something like this as a workaround ?\n(I can send a proper patch, if this is the way forward)\n\ncommit dcd7d5551d6931e47829c7febbee0877340eb17f\nAuthor: Torsten Bögershausen <tboegi@web.de>\nDate:   Sat Feb 27 15:18:28 2016 +0100\n\n    config.mak.uname: Darwin: Use clang for Mac OS X 10.6\n\n    Commit\n    \"tree-diff: catch integer overflow in combine_diff_path allocation\"\n    make gcc under Mac OX 10.6 crash like this:\n\n    CC combine-diff.o\n    combine-diff.c: In function ‘diff_tree_combined’:\n    combine-diff.c:1391: internal compiler error: Segmentation fault\n\n    Xcode for Mac OS X 10.6 has both gcc and clang.\n    Later versions of Mac OSX/Xcode only provide clang, and gcc is a wrapper\n    to it.\n    Make Git compile under Mac OS X 10.6 by using clang instead of gcc\n\ndiff --git a/config.mak.uname b/config.mak.uname\nindex d6f7980..211afec 100644\n--- a/config.mak.uname\n+++ b/config.mak.uname\n@@ -102,6 +102,9 @@ ifeq ($(uname_S),Darwin)\n        ifeq ($(shell expr \"$(uname_R)\" : '[15]\\.'),2)\n                NO_STRLCPY = YesPlease\n        endif\n+       ifeq ($(shell test \"`expr \"$(uname_R)\" : '\\([0-9][0-9]*\\)\\.'`\" -ge 10 &&\necho 1),1)\n+    CC = clang\n+       endif\n        ifeq ($(shell test \"`expr \"$(uname_R)\" : '\\([0-9][0-9]*\\)\\.'`\" -ge 11 &&\necho 1),1)\n                HAVE_GETDELIM = YesPlease\n        endif\n"},{"id":"279721","messageId":"xmqqfuwcg0g8.fsf@gitster.mtv.corp.google.com","threadId":"41527","inReplyTo":"56D1DF53.1010006@web.de","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-02-28T18:21:43Z","receivedAt":"2016-02-28T18:21:43Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Torsten Bögershausen <tboegi@web.de> writes:\n\n> How about something like this as a workaround ?\n> (I can send a proper patch, if this is the way forward)\n> ...\n>     Make Git compile under Mac OS X 10.6 by using clang instead of gcc\n\nGiven that it is understandable that Apple is no longer interesting\nin fixing compiler bugs for 10.6, I think that is the best we could\ndo to help its users.\n\n> ...\n> +       ifeq ($(shell test \"`expr \"$(uname_R)\" : '\\([0-9][0-9]*\\)\\.'`\" -ge 10 &&\n> echo 1),1)\n\nThis singles out 10.6-and-nothing-else, 10.6-or-older or\n10.6-or-newer?\n\nThat \"-ge\", which smells like a copy-and-pasta from the next one\nthat says \"we have getdelim on 10.7-and-newer\" bothers me.\n\n> +    CC = clang\n> +       endif\n>         ifeq ($(shell test \"`expr \"$(uname_R)\" : '\\([0-9][0-9]*\\)\\.'`\" -ge 11 &&\n> echo 1),1)\n>                 HAVE_GETDELIM = YesPlease\n>         endif\n"},{"id":"279730","messageId":"1456690184-25127-1-git-send-email-tboegi@web.de","threadId":"41527","inReplyTo":"xmqqfuwcg0g8.fsf@gitster.mtv.corp.google.com","subject":"[PATCH 1/1] config.mak.uname: Darwin: Use clang for Mac OS X 10.6","fromName":"","fromEmail":"tboegi@web.de","sentAt":"2016-02-28T20:09:44Z","receivedAt":"2016-02-28T20:09:44Z","isPatch":true,"sender":{"key":"tboegi@web.de","avatar":"https://avatars.githubusercontent.com/u/7138363?v=4"},"body":"From: Torsten Bögershausen <tboegi@web.de>\n\nCommit\n\"tree-diff: catch integer overflow in combine_diff_path allocation\"\nmake gcc under Mac OX 10.6 crash like this:\n\nCC combine-diff.o\ncombine-diff.c: In function ‘diff_tree_combined’:\ncombine-diff.c:1391: internal compiler error: Segmentation fault\n\nXcode for Mac OS X 10.6 has both gcc and clang.\nLater versions of Mac OSX/Xcode only provide clang, and gcc is a wrapper\nto it.\nMake Git compile under Mac OS X 10.6 by using clang instead of gcc\n\nSigned-off-by: Torsten Bögershausen <tboegi@web.de>\n---\n config.mak.uname | 3 +++\n 1 file changed, 3 insertions(+)\n\ndiff --git a/config.mak.uname b/config.mak.uname\nindex d6f7980..d0a3df3 100644\n--- a/config.mak.uname\n+++ b/config.mak.uname\n@@ -102,6 +102,9 @@ ifeq ($(uname_S),Darwin)\n \tifeq ($(shell expr \"$(uname_R)\" : '[15]\\.'),2)\n \t\tNO_STRLCPY = YesPlease\n \tendif\n+\tifeq ($(shell test \"`expr \"$(uname_R)\" : '\\([0-9][0-9]*\\)\\.'`\" -eq 10 && echo 1),1)\n+    CC = clang\n+\tendif\n \tifeq ($(shell test \"`expr \"$(uname_R)\" : '\\([0-9][0-9]*\\)\\.'`\" -ge 11 && echo 1),1)\n \t\tHAVE_GETDELIM = YesPlease\n \tendif\n-- \n2.7.0.303.g2c4f448.dirty\n"},{"id":"279761","messageId":"20160229101808.GD2950@sigill.intra.peff.net","threadId":"41527","inReplyTo":"56D14C56.6070306@web.de","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Jeff King","fromEmail":"peff@peff.net","sentAt":"2016-02-29T10:18:08Z","receivedAt":"2016-02-29T10:18:08Z","isPatch":false,"sender":{"key":"peff@peff.net","avatar":"https://avatars.githubusercontent.com/u/45925?v=4"},"body":"On Sat, Feb 27, 2016 at 08:12:22AM +0100, Torsten Bögershausen wrote:\n\n> > Torsten, what is the compiler version (I don't have Apple compilers, but\n> > it seems plausible that older clang might have the same problem).\n>\n> That's machine is running Mac OS X 10.6, which is no longer supported\n> with updates.\n> \n>  gcc --version\n> i686-apple-darwin10-gcc-4.2.1 (GCC) 4.2.1 (Apple Inc. build 5666) (dot 3)\n> Copyright (C) 2007 Free Software Foundation, Inc.\n> This is free software; see the source for copying conditions.  There is NO\n> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\nThanks. Out of curiosity, I tried to reproduce by with a build of gcc\n4.2.1, to see if I could bisect. But it seems the toolchain is quite\ncomplex. After much munging, I managed to build a broken compiler (which\nI think is due to a much too-new version of bison, but I stopped\ndigging).\n\nYour suggestion elsewhere in the thread to just use clang instead sounds\ngood to me. :)\n\n-Peff\n"},{"id":"279788","messageId":"CAPig+cQkP++rnE4L1QSr-sW8VkS8ce7CyY9KAscebLao2Hw73A@mail.gmail.com","threadId":"41527","inReplyTo":"20160229101808.GD2950@sigill.intra.peff.net","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Eric Sunshine","fromEmail":"sunshine@sunshineco.com","sentAt":"2016-02-29T16:28:41Z","receivedAt":"2016-02-29T16:28:41Z","isPatch":false,"sender":{"key":"sunshine@sunshineco.com","avatar":"https://avatars.githubusercontent.com/u/163641?v=4"},"body":"On Mon, Feb 29, 2016 at 5:18 AM, Jeff King <peff@peff.net> wrote:\n> On Sat, Feb 27, 2016 at 08:12:22AM +0100, Torsten Bögershausen wrote:\n>> > Torsten, what is the compiler version (I don't have Apple compilers, but\n>> > it seems plausible that older clang might have the same problem).\n>>\n>> That's machine is running Mac OS X 10.6, which is no longer supported\n>> with updates.\n>>\n>>  gcc --version\n>> i686-apple-darwin10-gcc-4.2.1 (GCC) 4.2.1 (Apple Inc. build 5666) (dot 3)\n>> Copyright (C) 2007 Free Software Foundation, Inc.\n>> This is free software; see the source for copying conditions.  There is NO\n>> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n>\n> Thanks. Out of curiosity, I tried to reproduce by with a build of gcc\n> 4.2.1, to see if I could bisect. But it seems the toolchain is quite\n> complex. After much munging, I managed to build a broken compiler (which\n> I think is due to a much too-new version of bison, but I stopped\n> digging).\n>\n> Your suggestion elsewhere in the thread to just use clang instead sounds\n> good to me. :)\n\nIf possible, for the moment, I'd prefer to hold off on that\nsledge-hammer approach of unconditionally making the build use clang.\nIt would be nice to have a more detailed understanding of what exactly\nis triggering the Apple compiler bug, and I've been trying to find\ntime to check it out on my old Mac.\n"},{"id":"279791","messageId":"xmqq37sbe8qr.fsf@gitster.mtv.corp.google.com","threadId":"41527","inReplyTo":"CAPig+cQkP++rnE4L1QSr-sW8VkS8ce7CyY9KAscebLao2Hw73A@mail.gmail.com","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-02-29T17:17:48Z","receivedAt":"2016-02-29T17:17:48Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Eric Sunshine <sunshine@sunshineco.com> writes:\n\n>> Your suggestion elsewhere in the thread to just use clang instead sounds\n>> good to me. :)\n>\n> If possible, for the moment, I'd prefer to hold off on that\n> sledge-hammer approach of unconditionally making the build use clang.\n> It would be nice to have a more detailed understanding of what exactly\n> is triggering the Apple compiler bug, and I've been trying to find\n> time to check it out on my old Mac.\n\nThe (fixed) patch posted was \"on OSX 10.6, use clang by default\"; I\nthink that for majority of the users, that is a reasonable thing to do.\n\nThe result of the patch still allows those who want to see how badly\nold Gcc breaks to do so with \"make CC=gcc\", doesn't it?\n"},{"id":"279879","messageId":"CAPig+cTAv8eNcTkN6K1kJ5zBcibmbRR59MLV6PCM=qa84HNdDg@mail.gmail.com","threadId":"41527","inReplyTo":"xmqq37sbe8qr.fsf@gitster.mtv.corp.google.com","subject":"Re: What's cooking in git.git (Feb 2016, #07; Thu, 25)","fromName":"Eric Sunshine","fromEmail":"sunshine@sunshineco.com","sentAt":"2016-02-29T23:37:40Z","receivedAt":"2016-02-29T23:37:40Z","isPatch":false,"sender":{"key":"sunshine@sunshineco.com","avatar":"https://avatars.githubusercontent.com/u/163641?v=4"},"body":"On Mon, Feb 29, 2016 at 12:17 PM, Junio C Hamano <gitster@pobox.com> wrote:\n> Eric Sunshine <sunshine@sunshineco.com> writes:\n>>> Your suggestion elsewhere in the thread to just use clang instead sounds\n>>> good to me. :)\n>>\n>> If possible, for the moment, I'd prefer to hold off on that\n>> sledge-hammer approach of unconditionally making the build use clang.\n>> It would be nice to have a more detailed understanding of what exactly\n>> is triggering the Apple compiler bug, and I've been trying to find\n>> time to check it out on my old Mac.\n>\n> The (fixed) patch posted was \"on OSX 10.6, use clang by default\"; I\n> think that for majority of the users, that is a reasonable thing to do.\n>\n> The result of the patch still allows those who want to see how badly\n> old Gcc breaks to do so with \"make CC=gcc\", doesn't it?\n\nI was not able to reproduce the problem on OSX 10.5.8.\n\nSince the patch applies specifically to 10.6 (via -eq) rather than\n\"10.6 and earlier\" (via -le) it happily does not break the build for\nolder OSX releases (such as 10.5.8) which lack clang. Thanks.\n"}]}