Re: [PATCH 6/5] NUL hack to create_file()
From: Daniel Barkalow <hidden>
Date: 2016-06-15 22:44:39
On Wed, 28 May 2008, Junio C Hamano wrote:
This is not meant for application to the mainline. It allows your git to refuse to create a blob whose name is "nul".
I assume this is so you can test git's response to a defective filesystem without actually having a defective filesystem?
quoted hunk ↗ jump to hunk
--- entry.c | 8 ++++++++ 1 files changed, 8 insertions(+), 0 deletions(-)diff --git a/entry.c b/entry.c index 222aaa3..d24b803 100644 --- a/entry.c +++ b/entry.c@@ -81,6 +81,14 @@ static void remove_subtree(const char *path) static int create_file(const char *path, unsigned int mode) { + if (1) { + size_t len = strlen(path); + if (3 <= len && !strcmp(path + len - 3, "nul") && + (3 == len || path[len - 4] == '/')) { + errno = EPERM;
Shouldn't this be EEXIST? I think the issue is that the first exists for the purpose of open() but not for anything else we've done up to this point.
+ return -1; + } + } mode = (mode & 0100) ? 0777 : 0666; return open(path, O_WRONLY | O_CREAT | O_EXCL, mode); } -- 1.5.6.rc0.43.g823ea -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html