https://bugzilla.mindrot.org/show_bug.cgi?id=2687
--- Comment #29 from Darren Tucker <dtuc...@dtucker.net> --- Comment on attachment 2954 --> https://bugzilla.mindrot.org/attachment.cgi?id=2954 2nd part with lower priority >diff --git a/krl.c b/krl.c > /* Handled above, but still need to stay in synch */ >- sshbuf_reset(sect); >+ sshbuf_free(sect); This one is done. >@@ -1288,7 +1288,8 @@ ssh_krl_file_contains_key(const char *path, const struct >sshkey *key) > debug2("%s: checking KRL %s", __func__, path); > r = ssh_krl_check_key(krl, key); > out: >- close(fd); >+ if (fd != -1) >+ close(fd); This function doesn't do any descriptor handling any more. >diff --git a/readconf.c b/readconf.c >index acc1391..c4dff15 100644 >--- a/readconf.c >+++ b/readconf.c >@@ -1185,7 +1185,7 @@ parse_int: > value = cipher_number(arg); > if (value == -1) > fatal("%.200s line %d: Bad cipher '%s'.", >- filename, linenum, arg ? arg : "<NONE>"); >+ filename, linenum, arg); This code is gone (I think it was part of SSH1 handling). > if (*activep && *intptr == -1) > *intptr = value; > break; >@@ -1196,7 +1196,7 @@ parse_int: > fatal("%.200s line %d: Missing argument.", filename, > linenum); > if (*arg != '-' && !ciphers_valid(*arg == '+' ? arg + 1 : arg)) > fatal("%.200s line %d: Bad SSH2 cipher spec '%s'.", >- filename, linenum, arg ? arg : "<NONE>"); >+ filename, linenum, arg); I think this one is wrong: it'll potentially cause a NULL pointer deref platforms whose string libs don't guard against it (eg Solaris). Same for the following instances of the same thing. >diff --git a/ssh-pkcs11.c b/ssh-pkcs11.c >index aaf712d..62a76b3 100644 >--- a/ssh-pkcs11.c >+++ b/ssh-pkcs11.c >@@ -536,8 +536,8 @@ pkcs11_fetch_keys_filter(struct pkcs11_provider *p, >CK_ULONG slotidx, This function has been removed. >diff --git a/sshconnect1.c b/sshconnect1.c SSH1 support has been removed. >diff --git a/sshkey.c b/sshkey.c >index 58c1051..6afacb5 100644 >--- a/sshkey.c >+++ b/sshkey.c >@@ -1239,6 +1239,9 @@ sshkey_read(struct sshkey *ret, char **cpp) > u_long bits; > #endif /* WITH_SSH1 */ > >+ if (ret == NULL) >+ return SSH_ERR_INVALID_ARGUMENT; This has already been fixed. -- You are receiving this mail because: You are watching the assignee of the bug. You are watching someone on the CC list of the bug. _______________________________________________ openssh-bugs mailing list openssh-bugs@mindrot.org https://lists.mindrot.org/mailman/listinfo/openssh-bugs
