{ "tokens": { "input": 135577, "cache_read": 3042840, "cache_write": 760710, "output": 19705 }, "reviews": [ { "key": [ "https://ftp.debian.org/debian/pool/main/a/aalib/aalib_1.4p5.orig.tar.gz", "main aalib-1.4.0/src/aadoskbd.c:745:9: style: Condition 'i' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "The code does `if (!i) return AA_NONE;` and then `if (i)`, so 'i' is indeed always true at line 745. The warning itself is unchanged: the main and your lines are identical, and the PR only adds note lines to the error path ('Assuming that condition '!i' is not redundant'). This is a true positive in both versions. The added notes point to the earlier return that makes the condition redundant, so the message is slightly more informative." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/a/alpine/alpine_2.26+dfsg.orig.tar.xz", "main alpine-2.26/pith/adrbklib.c:2145:43: style: Condition 'extra!=NULL' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "The warning is a true positive: line 2142 already requires `extra != NULL` to enter the block, so the repeated check at line 2145 is always true. The same warning is still reported by the PR at the same location with the same text. The 'main' line only appears because the attached notes/error path differ slightly. The PR version's notes are correct, so the change makes no practical difference to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/a/alpine/alpine_2.26+dfsg.orig.tar.xz", "main alpine-2.26/web/src/pubcookie/auth_gss_proxy.c:194:11: style: Condition 'len==-1' is always false [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "high", "explanation": "`len` is an `unsigned long`: it is printed with %lu and passed as `unsigned long*` to the c-client challenger. In `len == -1`, the -1 is converted to ULONG_MAX. When `read_full` returns -1, `len` holds ULONG_MAX and the comparison is true, so it is not 'always false'. The old warning was a false positive caused by mishandling the signed/unsigned conversion. The PR's common-type-aware comparison logic removes it." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/a/anope/anope_2.0.17.orig.tar.gz", "main anope-2.0.17/src/misc.cpp:398:197: style: Condition 'hours!=1' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "In the nested ternary, the third branch is only reached when both `hours==1 && minutes==1` and `hours==1 && minutes!=1` are false, so `hours` cannot be 1 there. `hours!=1` is therefore always true, and the warning is a true positive in both versions. The message text and location are unchanged. The only visible difference is that the PR version now carries an error-path note pointing to the earlier `hours==1` condition. That is a plausible explanation of where the fact comes from, so the change does not materially affect users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/b/binutils-riscv64-unknown-elf/binutils-riscv64-unknown-elf_2.32.2020.04+dfsg.orig.tar.gz", "main binutils-riscv64-unknown-elf-2.32.2020.04+dfsg/sim/arm/iwmmxt.c:2566:8: error: Shifting 64-bit value by 96 bits is undefined behaviour [shiftTooManyBits]\n binutils-riscv64-unknown-elf-2.32.2020.04+dfsg/sim/arm/iwmmxt.c:2564:16: note: Assignment 'i=0', assigned value is 0\n binutils-riscv64-unknown-elf-2.32.2020.04+dfsg/sim/arm/iwmmxt.c:2566:8: note: Shift" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "In the Hqual loop `i` runs from 0 to 7, and the shifts involved are by `(i & 3)` halfwords (at most 48 bits) or by `i * 8` (at most 56 bits). A shift of a 64-bit value by 96 bits cannot happen here. The bogus count seems to come from merging values from an infeasible branch or from wrapped/unsound range arithmetic. The PR adds ternary-branch feasibility filtering and sounder range handling, so dropping this shiftTooManyBits warning removes a false positive." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/b/botan/botan_2.19.5+dfsg.orig.tar.xz", "your botan-2.19.5+dfsg/src/lib/prov/commoncrypto/commoncrypto_mode.cpp:189:7: style: Condition 'm_cipher' is always true [knownConditionTrueFalse]\n botan-2.19.5+dfsg/src/lib/prov/commoncrypto/commoncrypto_mode.cpp:184:16: note: Assuming that condition 'm_cipher==nullptr' is not redundant\n botan-2.19.5+dfsg/src/lib/prov/commoncrypto/commoncrypto_mode.cpp:189:7: note: Condition 'm_cipher' is always true" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "In clear(), the function returns early when m_cipher == nullptr (line 184), so the later `if(m_cipher)` at line 189 is always true. The warning is a correct true positive both before and after the PR. It shows up as a main/your pair only because some part of the message text changed, while the location, message and notes appear identical. Nothing changes for users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/c/connman-gtk/connman-gtk_1.1.1+git20180626.b72c6ab.orig.tar.xz", "main connman-gtk-master/src/service.c:592:5: style: Condition 'prop' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "The warning itself is the same in both runs. The PR version adds the note \"Assuming that condition '!prop' is not redundant\", which points to the earlier early-return check at line 589. That makes the reason for the warning clearer. The warning is a true positive: after `if(!prop) return ...;` at line 589, `prop` is always non-null at line 592." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/c/curl/curl_8.23.0~rc3.orig.tar.xz", "main curl-8.23.0-rc3/lib/vdns/dnscache.c:323:13: style: Condition 'dns' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "high", "explanation": "The warning at dnscache.c:323 is reported by both versions. Only the extra note lines (the error path) differ, so the main line and the your line are the same diagnostic. At line 323, 'dns' is inside 'if(dns && ...)', and the earlier branches that set dns = NULL are in else-if arms that do not reach this one. So 'dns' is indeed always true there, and the warning is a true positive in both versions. The change in the error path does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/d/deutex/deutex_5.2.3.orig.tar.gz", "your deutex-5.2.3/src/ident.c:786:20: style: Condition 'in_level' is always true [knownConditionTrueFalse]\n deutex-5.2.3/src/ident.c:766:13: note: Assuming that condition '!in_level' is not redundant\n deutex-5.2.3/src/ident.c:786:20: note: Condition 'in_level' is always true" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "The code is `if (!in_level) {...} else if (in_level)`. In the else branch `in_level` is necessarily nonzero, so 'Condition in_level is always true' is a correct style finding. Main already reported it. The only change is that the new output adds error-path notes ('Assuming that condition !in_level is not redundant' at line 766), which make the warning easier to understand. The warning itself is unchanged and still correct, and the explanation got slightly better." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/d/dolphin-emu/dolphin-emu_2609.a+dfsg.orig.tar.xz", "your dolphin-emu-2609a/Source/Core/DolphinQt/QtUtils/ClearLayoutRecursively.cpp:14:15: style: Condition 'child==nullptr' is always false [knownConditionTrueFalse]\n dolphin-emu-2609a/Source/Core/DolphinQt/QtUtils/ClearLayoutRecursively.cpp:12:29: note: Assuming that condition 'child=layout->takeAt(0)' is not redundant\n dolphin-emu-2609a/Source/Core/DolphinQt/QtUtils/ClearLayoutRecursively.cpp:14:15: note: Condition 'child==nullptr' is always false" ], "verdict": { "verdict": "neutral", "confidence": "high", "explanation": "Both main and the PR report the same warning at the same location. The 'your' line differs from the 'main' line only by the attached note lines, which suggests the error path changed slightly. The warning itself is a true positive: inside `while (QLayoutItem* child = layout->takeAt(0))`, child is non-null, so `child == nullptr` is always false. Nothing meaningful changes for users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/d/dyssol/dyssol_1.5.0.orig.tar.xz", "main dyssol-1.5.0/ModelsAPI/TDArray.cpp:150:15: style: Condition 'index!=-1' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "high", "explanation": "`index` is a `size_t`, so in `index != -1` the `-1` is converted to SIZE_MAX. This is the usual 'not found' sentinel that `GetIndexByTime` returns. Main ignored that conversion and treated `index` as non-negative, so it concluded `index != -1` is always true. That is a false positive, and the same pattern was removed at TDArray.cpp:127 and the other sites listed. The PR models the usual arithmetic conversion correctly, so removing this warning is an improvement." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/e/ezurio-qcacld-2.0-dkms/ezurio-qcacld-2.0-dkms_0.0~git20240408.aa96a9f+dfsg.orig.tar.xz", "main boundarydevices-qcacld-2.0-aa96a9f/CORE/CLD_TXRX/TXRX/ol_rx.c:329:13: style: Condition 'pdev->osif_rx_mon_cb' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "Both main and the PR report the same knownConditionTrueFalse at line 329; only the attached notes or ordering differ. The warning is a true positive: line 320 returns early when pdev->osif_rx_mon_cb is NULL, so the check at 329 is always true. The intervening call only receives htt_pdev, not pdev, so it does not plausibly reset the callback. The message is unchanged, so this does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/g/gkermit/gkermit_2.01.orig.tar.gz", "your gcmdline.c:97:17: style: Condition 'flag' is always true [knownConditionTrueFalse]\n gcmdline.c:92:10: note: Assuming that condition '!flag' is not redundant\n gcmdline.c:97:17: note: Condition 'flag' is always true" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "On line 97, `flag` is checked in the else branch of `if (!flag)`, so it is always true there. The warning is a true positive both before and after the PR. The only change is that the PR's output adds a note pointing to the `!flag` condition on line 92, which shows the user why `flag` is always true. That makes the message more informative." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/g/gl4es/gl4es_1.1.6+ds.orig.tar.xz", "main gl4es-1.1.6/src/gl/decompress.c:255:48: warning: Either the condition 'alphaCodeIndex<=12' is redundant or there is signed integer overflow for expression 'alphaCodeIndex-16'. [integerOverflowCond]\n gl4es-1.1.6/src/gl/decompress.c:250:23: note: Assuming that condition 'alphaCodeIndex<=12' is not redundant\n gl4es-1.1.6/src/gl/decompress.c:255:48: note: Integer overflow" ], "verdict": { "verdict": "improvement", "confidence": "high", "explanation": "Line 255 is in the else branch, which is reached only when alphaCodeIndex > 12 and != 15. alphaCodeIndex is 3*(4*j+i) with j,i in 0..3, so here it is 18..45. Computing alphaCodeIndex - 16 on an int in that range cannot overflow, so the integerOverflowCond warning in main was a false positive, likely caused by mishandled range bounds. Removing it is an improvement. The separate shiftNegative warning that the PR adds is a different line and is not judged here." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/h/htmldoc/htmldoc_1.9.25.orig.tar.gz", "main htmldoc-1.9.25/htmldoc/mmd.c:1026:34: style: Condition 'stackptr->parent->type!=MMD_TYPE_CODE_BLOCK' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "The warning is still reported. Line 1026 repeats the check from line 1023, so 'stackptr->parent->type != MMD_TYPE_CODE_BLOCK' really is always true there. The only change is that the PR's output now includes the error path notes: 'Assuming that condition ... is not redundant' at line 1023, then the result at line 1026. That makes the same true positive easier to understand." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/i/ipmitool/ipmitool_1.8.19.orig.tar.gz", "your ipmitool-IPMITOOL_1_8_19/lib/ipmi_sdr.c:3393:8: style: Condition 'sdrr' is always true [knownConditionTrueFalse]\n ipmitool-IPMITOOL_1_8_19/lib/ipmi_sdr.c:3383:7: note: Assuming that condition '!sdrr' is not redundant\n ipmitool-IPMITOOL_1_8_19/lib/ipmi_sdr.c:3392:7: note: Assuming condition is true\n ipmitool-IPMITOOL_1_8_19/lib/ipmi_sdr.c:3393:8: note: Condition 'sdrr' is always true" ], "verdict": { "verdict": "neutral", "confidence": "high", "explanation": "The warning appears in both main and your. Only the extra error-path notes differ, and the message is identical. After `if (!sdrr) break;`, `sdrr` is non-null, so `if (sdrr)` at line 3393 really is always true. The warning was a true positive before and is still a true positive. The change does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/j/jclassinfo/jclassinfo_0.19.1.orig.tar.gz", "main jclassinfo-0.19.1/jclass/manifest.c:100:11: style: Condition '*p==' '' is always false [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "The warning on line 100 is still reported with the same text. After `while (*p == ' ') p++;` the check `if (*p == ' ')` is indeed always false, so it is a true positive either way. The main/your lines differ only in the extra notes the PR adds, which point to line 98 as the place where the condition is assumed. This is a cosmetic change to the error path, not a new or lost finding." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/j/jclassinfo/jclassinfo_0.19.1.orig.tar.gz", "your jclassinfo-0.19.1/jclass/manifest.c:100:11: style: Condition '*p==' '' is always false [knownConditionTrueFalse]\n jclassinfo-0.19.1/jclass/manifest.c:98:14: note: Assuming that condition '*p==' '' is not redundant\n jclassinfo-0.19.1/jclass/manifest.c:100:11: note: Condition '*p==' '' is always false" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "The same warning is reported at the same location by both versions. The `while (*p == ' ')` loop exits only when `*p != ' '`, so the following `if (*p == ' ')` is always false. This is a true positive in both. The diff shows up only because the 'your' entry now includes notes (the condition at line 98 and the conclusion at line 100). That is a cosmetic difference in how the message is emitted and makes no practical difference to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/l/ldb/ldb_2.2.3.orig.tar.gz", "your ldb-2.2.3/third_party/popt/popt.c:936:9: style: Condition '*t=='\\0'' is always false [knownConditionTrueFalse]\n ldb-2.2.3/third_party/popt/popt.c:930:32: note: Assuming that condition '*t' is not redundant\n ldb-2.2.3/third_party/popt/popt.c:936:9: note: Condition '*t=='\\0'' is always false" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "Both main and the PR report the same knownConditionTrueFalse warning at line 936 with identical message text; at most the attached notes differ. The warning is a false positive either way. If the token at `t` is ',' (a leading comma or \",,\"), the inner loop does not advance `te`, so `*te++ = '\\0'` writes through `t`, and `*t == '\\0'` becomes true. The PR neither removes nor introduces this false positive, so the change does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/libe/libexplain/libexplain_1.4.D001.orig.tar.gz", "main libexplain-1.4/libexplain/ftello_or_die.c:33:16: style: Condition 'result<0' is always true [knownConditionTrueFalse]\n libexplain-1.4/libexplain/ftello_or_die.c:50:15: note: Assignment 'result=-1', assigned value is -1\n libexplain-1.4/libexplain/ftello_or_die.c:32:37: note: Calling function 'explain_ftello_on_error' returns -1\n libexplain-1.4/libexplain/ftello_or_die.c:32:37: note: Assignment 'result=explain_ftello_on_error(fp)', assigned value is -1\n libexplain-1.4/libexplain/ftello_or_die.c:33:16: note: Condition 'result<0' is always true" ], "verdict": { "verdict": "acceptable regression", "confidence": "low", "explanation": "When HAVE_FTELLO is not defined, the code takes the #else branch and does `result = -1;`, so explain_ftello_on_error() always returns -1. In that configuration, `result < 0` at line 33 really is always true, so the removed warning was technically a true positive. The removal most likely comes from the PR's stricter value propagation and conversion handling: `off_t` is probably an unknown type here because system headers are missing, and values on such types are now dropped. That is a lost true positive, not a removed false positive. It is minor, because the finding is a configuration artifact: real builds define HAVE_FTELLO, and there ftello() can return non-negative values." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/libe/libexplain/libexplain_1.4.D001.orig.tar.gz", "your libexplain-1.4/libexplain/buffer/errno/setgid.c:80:36: style: Condition 'gid!=sgid' is always true [knownConditionTrueFalse]\n libexplain-1.4/libexplain/buffer/errno/setgid.c:80:21: note: Assuming that condition 'gid!=rgid' is not redundant\n libexplain-1.4/libexplain/buffer/errno/setgid.c:78:18: note: sgid is assigned 'rgid' here.\n libexplain-1.4/libexplain/buffer/errno/setgid.c:80:36: note: Condition 'gid!=sgid' is always true" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "Both main and the PR report the same warning text at the same location. Only the attached notes (error path) appear to differ. The warning is a true positive for the #else configuration: `sgid = rgid`, so once `gid != rgid` holds, `gid != sgid` is always true. The PR's notes correctly cite the assignment at line 78. Main's notes are not shown, so this cannot be judged a clear improvement; for users the result is effectively unchanged." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/libe/libextractor/libextractor_1.19.orig.tar.gz", "main libextractor-1.19/src/main/extractor.c:474:14: style: Condition '-1!=pos->seek_request' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "Both main and the PR report the same warning at the same location with the same text. Only the attached notes or error path differ. The warning is a true positive: the preceding `if (-1 == pos->seek_request) { ...; break; }` guarantees that `-1 != pos->seek_request` is always true at line 474. The PR neither removes nor adds a finding here." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/libl/libldac/libldac_2.0.2.3+git20200429+ed310a0.orig.tar.xz", "main libldac-2.0.2.3+git20200429+ed310a0/src/sigana_fixp_ldac.c:175:32: warning: Either the condition 'in<=0' is redundant or the array 'sa_sqrt_ldac[97]' is accessed at index -31, which is out of bounds. [negativeIndex]\n libldac-2.0.2.3+git20200429+ed310a0/src/sigana_fixp_ldac.c:166:12: note: Assuming that condition 'in<=0' is not redundant\n libldac-2.0.2.3+git20200429+ed310a0/src/sigana_fixp_ldac.c:170:7: note: i is assigned '(INT16)(in>>24)' here.\n libldac-2.0.2.3+git20200429+ed310a0/src/sigana_fixp_ldac.c:173:11: note: Assignment 'i=i-32', assigned value is -32\n libldac-2.0.2.3+git20200429+ed310a0/src/sigana_fixp_ldac.c:175:32: note: Negative array index" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "`calc_sqrt_ldac` is a static inline helper. Its only callers pass `tmp = (acc1 << e) >> 32`, where `e` is chosen by `calc_exp_ldac` to normalize the value. That puts `in` roughly in [2^29, 2^31), so `in >> 24` is in [32, 127]. Then `i - 32` is 0..95 and `i + 1` is at most 96, which fits the 97-entry `sa_sqrt_ldac` table. The removed warning reasoned only that `in > 0` means `in >= 1`, so `i - 32` could be -32 and `i + 1` could be -31. That ignores the normalization the callers guarantee, and the `in <= 0` guard has nothing to do with how such a small value could arise. In practice this is a false positive, so removing it helps users. The removal probably comes from the PR no longer carrying an unbounded lower-bound range through the `(INT16)` cast, rather than from better understanding of the code." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/libs/libssc/libssc_0.4.4.orig.tar.gz", "your libssc/src/libssc-sensor.c:297:16: style: Condition 'suid_msg!=NULL' is always true [knownConditionTrueFalse]\n libssc/src/libssc-sensor.c:290:16: note: Assuming that condition 'suid_msg==NULL' is not redundant\n libssc/src/libssc-sensor.c:297:16: note: Condition 'suid_msg!=NULL' is always true" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "The warning at line 297 is reported by both main and the PR. Only the message text or note formatting differs slightly, so it shows up as a removed/added pair. It is a true positive either way: suid_msg==NULL returns at line 290, so 'suid_msg != NULL' at line 297 is always true. Nothing changes for users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/m/m1n1/m1n1_2.0.0.orig.tar.gz", "your m1n1-2.0.0/src/display.c:339:9: style: Condition 's_w' is always true [knownConditionTrueFalse]\n m1n1-2.0.0/src/display.c:332:9: note: Assuming that condition '!config' is not redundant\n m1n1-2.0.0/src/display.c:335:21: note: s_w is assigned 'config' here.\n m1n1-2.0.0/src/display.c:339:9: note: Condition 's_w' is always true" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "Both main and the PR report this warning at the same location with the same message. Only the main/your line pair shows up in the diff, so the change is probably in the notes or error path rather than the finding itself. The warning is a true positive. `config` is checked for null and the function returns early, so `s_w = config` is always non-null at line 339. The finding and its message are unchanged, so this does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/m/mate-settings-daemon/mate-settings-daemon_1.26.1.orig.tar.xz", "your mate-settings-daemon-1.26.1/plugins/smartcard/msd-smartcard-manager.c:646:13: style: Condition '!module_explicitly_specified' is always false [knownConditionTrueFalse]\n mate-settings-daemon-1.26.1/plugins/smartcard/msd-smartcard-manager.c:620:13: note: Assuming that condition 'module_explicitly_specified' is not redundant\n mate-settings-daemon-1.26.1/plugins/smartcard/msd-smartcard-manager.c:646:13: note: Condition '!module_explicitly_specified' is always false" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "`module_explicitly_specified` is `module_path != NULL`, and the `if`/`else` at line 620 covers both cases. At line 646, `!module_explicitly_specified` is true whenever `module_path` is NULL, so 'always false' is a false positive. Main already reported the same false positive at the same location; only the attached notes changed (it now claims to assume the line-620 condition is not redundant). The wrong warning is neither removed nor made worse in a meaningful way, so the net effect for users is unchanged." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/m/mtink/mtink_1.0.16.orig.tar.gz", "main mtink-1.0.16/mainSrc/ttink.c:736:23: style: Condition '*retBuf=='\\0'' is always false [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "The code checks `*retBuf` at line 734, so `*retBuf == '\\0'` on line 736 is always false. The warning is a true positive and both main and the PR report it. Only the 'main'/'your' line pair changed, and the message text and notes look identical. The difference is most likely an internal detail such as error-path ordering or column info, which does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/o/osmocom-dahdi-linux/osmocom-dahdi-linux_0.0~git20260630.77f32ae.orig.tar.xz", "main osmocom-dahdi-linux-0.0~git20260630.77f32ae/drivers/dahdi/wcte13xp-base.c:2555:7: style: Comparing expression of type 'unsigned int' against value -1. Condition is always false. [compareValueOutOfTypeRangeError]" ], "verdict": { "verdict": "improvement", "confidence": "high", "explanation": "`index` is an `unsigned int` initialized to -1, i.e. UINT_MAX. In `-1 == index`, the usual arithmetic conversions turn -1 into UINT_MAX, so the comparison is true whenever no free slot was found. The old check said 'always false', which is a false positive. The PR's rewritten checkCompareValueOutOfTypeRange works out the common type (unsigned int) and finds that -1 converted to it lies inside the range of `unsigned int`, so it no longer reports. Removing this false positive is an improvement." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/p/pdfmixtool/pdfmixtool_1.2.1.orig.tar.xz", "your pdfmixtool-1.2.1/src/pdf_edit_lib/pdf_editor.cpp:726:27: style: Condition 'pi.rot%360==90' is always false [knownConditionTrueFalse]\n pdfmixtool-1.2.1/src/pdf_edit_lib/pdf_editor.cpp:717:22: note: Assuming that condition 'pi.rot%360==90' is not redundant\n pdfmixtool-1.2.1/src/pdf_edit_lib/pdf_editor.cpp:721:27: note: Assuming condition is false\n pdfmixtool-1.2.1/src/pdf_edit_lib/pdf_editor.cpp:726:27: note: Condition 'pi.rot%360==90' is always false" ], "verdict": { "verdict": "improvement", "confidence": "high", "explanation": "Line 726 repeats the condition from line 717: `pi.rot % 360 == 90`. Nothing changes `pi.rot` in between. The else-if chain only reaches line 726 when that same condition was already false at line 717, so it really is always false. This is a genuine bug in pdfmixtool: the last branch was almost certainly meant to test 270. The warning is a true positive that main missed, so the PR added a correct finding." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/p/php-pinba/php-pinba_1.1.2.orig.tar.gz", "your php-pinba-1.1.2/protobuf-c.c:1338:23: style: Condition '*pstr!=NULL' is always true [knownConditionTrueFalse]\n php-pinba-1.1.2/protobuf-c.c:1335:34: note: Assuming that condition '*pstr!=NULL' is not redundant\n php-pinba-1.1.2/protobuf-c.c:1338:23: note: Condition '*pstr!=NULL' is always true" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "The warning appears on both sides with the same location and message. Inside the block guarded by `maybe_clear && *pstr != NULL` (line 1335), the inner check `*pstr != NULL` at line 1338 is redundant, so the warning is a true positive. The only difference between the lines is probably the attached notes or ordering, which is not visible here. It makes no difference to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/p/pidgin/pidgin_2.14.14.orig.tar.bz2", "your pidgin-2.14.14/libpurple/protocols/novell/nmuser.c:1619:10: style: Condition 'field' is always true [knownConditionTrueFalse]\n pidgin-2.14.14/libpurple/protocols/novell/nmuser.c:1618:19: note: Assuming that condition 'field=nm_locate_field(\"NM_A_SZ_DN\",field)' is not redundant\n pidgin-2.14.14/libpurple/protocols/novell/nmuser.c:1619:10: note: Condition 'field' is always true" ], "verdict": { "verdict": "neutral", "confidence": "high", "explanation": "The warning text and location are the same in both versions. Only the attached note lines differ, which is why the diff lists it as a main/your pair. The finding itself is correct: inside `while ((field = nm_locate_field(...)))` the variable `field` is non-null, so the `if (field && ...)` check at line 1619 is redundant. The note now says 'Assuming that condition ... is not redundant', which is a cosmetic difference and does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/p/psicode/psicode_3.4.0.orig.tar.gz", "your psi3/src/lib/libdpd/buf4_sort.cc:1050:25: style: Condition 'rows_left' is always true [knownConditionTrueFalse]\n psi3/src/lib/libdpd/buf4_sort.cc:1046:12: note: Assuming that condition 'rows_left' is not redundant\n psi3/src/lib/libdpd/buf4_sort.cc:1050:25: note: Condition 'rows_left' is always true" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "Line 1050 is inside `if(rows_left)` (line 1046), and `rows_left` is not modified in between. So `rows_left ? nbuckets-1 : nbuckets` really is always true, and the warning is a true positive. Main reports the same warning at the same location with the same text. Only the attached notes may differ, so users see the same diagnostic before and after." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/p/python-pairix/python-pairix_0.3.8.orig.tar.gz", "your pairix-0.3.8/src/index.c:769:12: style: Condition 'fname' is always true [knownConditionTrueFalse]\n pairix-0.3.8/src/index.c:766:12: note: Assuming that condition 'fname==0' is not redundant\n pairix-0.3.8/src/index.c:769:12: note: Condition 'fname' is always true" ], "verdict": { "verdict": "neutral", "confidence": "high", "explanation": "The same warning, 'Condition fname is always true' at line 769, is reported by both main and your. It is a true positive: line 766 returns early when fname is null, so the later check is redundant. Only the attached notes differ: the new output shows an error path pointing at line 766. That is the same finding with equivalent or slightly better context, so nothing meaningful changes for users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/p/python-sabyenc/python-sabyenc_5.4.3.orig.tar.gz", "main sabyenc-5.4.3/src/sabyenc3.cc:527:12: style: Condition 'Py_output_buffer' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "The same knownConditionTrueFalse warning at line 527 is reported by both main and the PR. Only the accompanying notes differ: the PR's version adds an error path (\"Assuming that condition 'Py_output_buffer==NULL' is not redundant\", \"Assuming condition is true\"). The warning is a true positive either way, because the earlier NULL check returns, so Py_output_buffer is non-null at line 527. The text and location are identical, so the change does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/r/r-cran-sass/r-cran-sass_0.4.10+dfsg.orig.tar.xz", "main sass/src/libsass/src/parser.cpp:757:21: style: Condition 'wrapped' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "high", "explanation": "The same warning still appears at the same location. The only difference is that the PR version adds the supporting notes ('Assuming that condition ... is not redundant'). Inside `if (SelectorListObj wrapped = parseSelectorList(true))`, the inner check of `wrapped` is indeed always true, so the warning is a true positive both before and after. The main/your pair comes from the extra note lines, not from any change in the reported finding." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/r/rjava/rjava_1.0-18.orig.tar.gz", "main rJava/src/jri/src/rjstring.c:45:10: style: Condition 'res==-1' is always false [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "high", "explanation": "`res` is a size_t holding the result of Riconv, and the code compares it with -1. In that comparison -1 is converted to SIZE_MAX, which is exactly the value Riconv/iconv returns on error. So `res == -1` can be true, and main's 'always false' warning was a false positive. The PR fixes how values are converted to the common integral type, so the false positive is gone." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/r/rjava/rjava_1.0-18.orig.tar.gz", "main rJava/src/rjstring.c:45:10: style: Condition 'res==-1' is always false [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "high", "explanation": "`res` is a `size_t` holding the result of Riconv. In `res == -1`, the -1 is converted to SIZE_MAX, which is the standard iconv error idiom, so the condition can be true. The old warning came from the -1 not being converted to unsigned before the comparison. The PR's implicit-conversion handling fixes this, so this false positive is gone." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/r/ruby-tioga/ruby-tioga_1.19.1.orig.tar.xz", "your ext/Tioga/FigureMaker/__shared_pdftext.c:278:17: style: Condition 'fontinfo==NULL' is always false [knownConditionTrueFalse]\n ext/Tioga/FigureMaker/__shared_pdftext.c:271:17: note: Assuming that condition 'fontinfo==NULL' is not redundant\n ext/Tioga/FigureMaker/__shared_pdftext.c:278:17: note: Condition 'fontinfo==NULL' is always false" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "Both main and the PR report the same warning, on the same line, with the same message. Only the notes/error path differ: the new output adds an 'Assuming that condition fontinfo==NULL is not redundant' note pointing at line 271. The warning itself is a true positive. After the early return at line 275, fontinfo cannot be NULL at line 278, so that part of the condition is always false. The extra note points at the earlier check that makes the later one redundant, which is fine. The warning's presence and accuracy are essentially unchanged, so this does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/s/sopwith/sopwith_2.9.0.orig.tar.gz", "main sdl-sopwith-2.9.0/src/sdl/video.c:834:14: style: Comparing expression of type 'char' against value 128. Condition is always false. [compareValueOutOfTypeRangeError]" ], "verdict": { "verdict": "acceptable regression", "confidence": "medium", "explanation": "`c` is a plain `char` and the code tests `c >= 0x80` to skip non-ASCII UTF-8 bytes. On signed-char platforms such as x86, which is what the analyzed configuration uses, `c` can never reach 128. So the condition is always false and non-ASCII bytes are not filtered: a real portability bug that should use `unsigned char`. The PR now evaluates plain char under both signed and unsigned interpretations. Because the result differs on unsigned-char platforms, it suppresses the warning. That avoids a platform-dependent claim, but it removes a warning that points at a genuine bug on the configured platform, i.e. a new false negative." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/s/spatialindex/spatialindex_2.1.0.orig.tar.gz", "your libspatialindex-2.1.0/src/capi/sidx_api.cc:2071:11: style: Condition 'it!=0' is always true [knownConditionTrueFalse]\n libspatialindex-2.1.0/src/capi/sidx_api.cc:2069:18: note: Assuming that condition 'results[i]!=NULL' is not redundant\n libspatialindex-2.1.0/src/capi/sidx_api.cc:2070:47: note: Assignment 'it=reinterpret_cast(results[i])', assigned value is never 0\n libspatialindex-2.1.0/src/capi/sidx_api.cc:2071:11: note: Condition 'it!=0' is always true" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "Both main and the PR report the same warning, `Condition 'it!=0' is always true`, at the same location. Only the attached notes seem to differ, and main's notes are not shown. The warning is a true positive. `it` is a `reinterpret_cast` of `results[i]`, which was just checked to be non-NULL, so `it != 0` is redundant. The PR's error path explains this correctly, but without main's notes there is no evidence that the message became more or less accurate, so the change does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/s/subnetcalc/subnetcalc_2.7.6.orig.tar.xz", "main subnetcalc-2.7.6/src/subnetcalc.cc:1007:18: warning: Undefined behaviour: float (1.84467e+19) to integer conversion overflow. [floatConversionOverflow]\n subnetcalc-2.7.6/src/subnetcalc.cc:1006:16: note: Assuming that condition 'hostBits<=64' is not redundant\n subnetcalc-2.7.6/src/subnetcalc.cc:1007:18: note: float to integer conversion" ], "verdict": { "verdict": "acceptable regression", "confidence": "medium", "explanation": "Line 1007 runs under `hostBits <= 64`, and `hostBits == 64` is realistic: an IPv6 /64 prefix gives `128 - 64`. Then `pow(2.0, 64.0)` is 2^64, which exceeds ULLONG_MAX (2^64-1), so the cast to `unsigned long long` is undefined behaviour. The removed warning was therefore a true positive. It most likely disappeared because the float value comes from the upper bound of the condition, and the PR now ignores Upper-bound float values as positive overflow evidence. The result is a new false negative, not a removed false positive." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/t/totem-pl-parser/totem-pl-parser_3.26.6.orig.tar.xz", "your totem-pl-parser-3.26.6/plparse/totem-pl-parser.c:2230:10: style: Condition 'ret!=TOTEM_PL_PARSER_RESULT_SUCCESS' is always true [knownConditionTrueFalse]\n totem-pl-parser-3.26.6/plparse/totem-pl-parser.c:2224:10: note: Assuming that condition 'ret==TOTEM_PL_PARSER_RESULT_SUCCESS' is not redundant\n totem-pl-parser-3.26.6/plparse/totem-pl-parser.c:2227:62: note: Assuming condition is false\n totem-pl-parser-3.26.6/plparse/totem-pl-parser.c:2230:10: note: Condition 'ret!=TOTEM_PL_PARSER_RESULT_SUCCESS' is always true" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "The warning is the same true positive in both versions. Line 2224 returns when `ret == TOTEM_PL_PARSER_RESULT_SUCCESS`, so at line 2230 `ret != TOTEM_PL_PARSER_RESULT_SUCCESS` is always true. The only change is that the PR now attaches an error path. The note at 2224 correctly explains why the condition is always true. The note at 2227 ('Assuming condition is false') is accurate, since line 2230 is only reached when that check is false, though it is not needed for the conclusion. The message is more informative and no less accurate." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/t/tupi/tupi_0.2+git08.orig.tar.gz", "main tupi-0.2-git08/src/libbase/tuppaintareabase.cpp:348:25: style: Condition 'frame' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "neutral", "confidence": "medium", "explanation": "The same knownConditionTrueFalse warning at line 348 appears in both runs. `frame` is the variable declared in the enclosing `if (TupFrame *frame = ...)`, so the inner `if (frame)` is always true. The warning is a true positive, and only the main line appears in this result. The 'your' line shows the warning with notes, so it probably only differs in its error path or the diff format. The warning itself is unchanged, so this does not matter to users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/u/ultrastardx/ultrastardx_2026.9.0+dfsg.orig.tar.xz", "main UltraStar-Deluxe-USDX-8b5324f/src/lib/openCV3/ApiWrapper.cpp:188:15: style: Condition '!src2' is always false [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "The code `if (!src2 || !src2 || !dst)` checks `!src2` twice, and the second check is always false. That is a true positive and likely a typo for `src1`. The warning is still reported in both versions. The only change is that the PR's version adds notes explaining the reasoning: 'Assuming that condition '!src2' is not redundant' points back to the first check. The main line disappears only because the message gained these notes. A warning with a clearer explanation path is slightly more accurate and helpful." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/u/ultrastardx/ultrastardx_2026.9.0+dfsg.orig.tar.xz", "your UltraStar-Deluxe-USDX-8b5324f/src/lib/openCV3/ApiWrapper.cpp:188:15: style: Condition '!src2' is always false [knownConditionTrueFalse]\n UltraStar-Deluxe-USDX-8b5324f/src/lib/openCV3/ApiWrapper.cpp:188:6: note: Assuming that condition '!src2' is not redundant\n UltraStar-Deluxe-USDX-8b5324f/src/lib/openCV3/ApiWrapper.cpp:188:15: note: Condition '!src2' is always false" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "Line 188 has the typo `!src2 || !src2 || !dst`. The second `!src2` can only be evaluated when the first one was false, so it really is always false. This is a true positive, and both versions report it with the same text. The only change is that the new version adds error-path notes pointing at the first `!src2` (188:6) as the reason. That explains why the duplicate check is always false and makes the warning easier to act on. Nothing was lost, so this is a small improvement." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/v/vlfeat/vlfeat_0.9.21+full.orig.tar.gz", "your vlfeat-0.9.21/vl/scalespace.c:373:18: warning: Expression '-o' can have a negative value. That is converted to an unsigned value and used in an unsigned calculation. [signConversion]\n vlfeat-0.9.21/vl/scalespace.c:721:14: note: Assignment 'op=-1', assigned value is -1\n vlfeat-0.9.21/vl/scalespace.c:724:43: note: Calling function 'vl_scalespace_get_level', 2nd argument 'op' value is -1\n vlfeat-0.9.21/vl/scalespace.c:392:77: note: Calling function 'vl_scalespace_get_octave_geometry', 2nd argument 'o' value is -1\n vlfeat-0.9.21/vl/scalespace.c:373:18: note: Negative value is converted to an unsigned value" ], "verdict": { "verdict": "regression", "confidence": "high", "explanation": "VL_SHIFT_LEFT(x,n) in vlfeat expands to `((n)>=0 ? (x)<<(n) : (x)>>-(n))`. The value path gives o = -1, so `-o` is 1, not negative. In either branch of the macro the shift count is nonnegative, and a shift count is not converted to the unsigned type of the left operand anyway. Reporting '-o' as possibly negative is a new false positive. It likely comes from the PR's changed unary-minus/conversion value handling." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/w/wmfrog/wmfrog_0.3.1+git20211013+ds.orig.tar.xz", "main wmfrog-0.3.1+git20211013/wmfrog/Src/wmfrog.c:177:112: style: Condition 'strcmp(weather.precip,\"IC\")==0' is always false [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "Line 177 repeats `strcmp(weather.precip, \"IC\") == 0` four times joined by `||`. If the first check is false, the later identical checks are also false, so the knownConditionTrueFalse warning is a true positive both before and after the PR. The warning text is unchanged. The only difference is that the new output adds an error path: a note saying the first condition at col 38 is assumed not redundant, and a note pointing at the repeated check. This explains why the condition is always false, which makes the message more informative for users." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/w/wmtv/wmtv_0.6.7.orig.tar.gz", "your wmtv-0.6.7/src/wmtv.c:512:20: style: Condition 'btime' is always true [knownConditionTrueFalse]\n wmtv-0.6.7/src/wmtv.c:505:15: note: Assuming that condition '!btime' is not redundant\n wmtv-0.6.7/src/wmtv.c:512:20: note: Condition 'btime' is always true" ], "verdict": { "verdict": "improvement", "confidence": "medium", "explanation": "Line 512 is `else if (btime)` and is only reached when `!btime` at line 505 was false, so `btime` is always true there. The warning is a correct, redundant-condition style finding. Main and the PR report the same warning at the same location. The PR version adds notes linking it to the `!btime` condition at line 505, which makes the reason clearer. The text change is at least neutral and slightly more informative." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/w/wtype/wtype_0.4.orig.tar.gz", "main wtype-0.4/main.c:385:11: style: Condition 'ret==-1' is always false [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "high", "explanation": "`ret` is a size_t and is compared with -1. The usual arithmetic conversions turn -1 into SIZE_MAX, which is exactly what mbstowcs returns on error, so `ret == -1` can be true. Main wrongly reported the condition as always false, likely because it treated -1 as an impossible value for an unsigned variable without converting it to the common unsigned type. The PR's conversion-aware inference removes this false positive." } }, { "key": [ "https://ftp.debian.org/debian/pool/main/z/zaz/zaz_1.0.1.orig.tar.xz", "main zaz-1.0.1/src/textureloader.cpp:114:10: style: Condition 'surface' is always true [knownConditionTrueFalse]" ], "verdict": { "verdict": "improvement", "confidence": "low", "explanation": "The same knownConditionTrueFalse warning at line 114 is reported by both versions; only its error path changed. The PR version adds notes pointing to the earlier `if (surface)` check at line 58, which shows why `surface` is non-null at line 114 (presumably the else branch's ERR(...) does not return). Whether the warning is a true positive depends on ERR, which is not shown and is unchanged by the PR. The new notes make the diagnostic easier to follow, so this is a slight improvement in message quality." } } ], "started": "2026-11-04 13:52:02 UTC", "runner": "server" }