Similar to what we did in edd9ed6a, disconnect the relationship with our stack allocated error buffer from the curl handle. Just as an FTP connection might have some network chatter on teardown causing the progress callback to be triggered, we might also hit an error condition that causes curl to write to our (now out of scope) error buffer.
I'm unable to reproduce FS#26327, but I have a suspicion that this should fix it. Signed-off-by: Dave Reisner <[email protected]> --- lib/libalpm/dload.c | 5 ++++- 1 files changed, 4 insertions(+), 1 deletions(-) diff --git a/lib/libalpm/dload.c b/lib/libalpm/dload.c index 33824be..e848b30 100644 --- a/lib/libalpm/dload.c +++ b/lib/libalpm/dload.c @@ -377,8 +377,11 @@ static int curl_download_internal(struct dload_payload *payload, /* perform transfer */ payload->curlerr = curl_easy_perform(curl); - /* immediately unhook the progress callback */ + /* disconnect relationships from the curl handle for things that might go out + * of scope, but could still be touched on connection teardown. This really + * only applies to FTP transfers. See FS#26327 for an example. */ curl_easy_setopt(curl, CURLOPT_NOPROGRESS, 1L); + curl_easy_setopt(curl, CURLOPT_ERRORBUFFER, (char *)0); /* was it a success? */ switch(payload->curlerr) { -- 1.7.7
