)]}'
{"tools/mlock_report.py":[{"author":{"_account_id":7118,"name":"Ian Wienand","email":"iwienand@redhat.com","username":"iwienand"},"change_message_id":"3293198acaa1be69f649f86bbdf5e2704516e759","unresolved":false,"context_lines":[{"line_number":26,"context_line":"        # iterate over the /proc/%pid/status files manually"},{"line_number":27,"context_line":"        try:"},{"line_number":28,"context_line":"            s \u003d open(\"%s/%d/status\" % (psutil.PROCFS_PATH, proc.pid), \u0027r\u0027)"},{"line_number":29,"context_line":"        except EnvironmentError:"},{"line_number":30,"context_line":"            continue"},{"line_number":31,"context_line":"        with s:"},{"line_number":32,"context_line":"            for line in s:"}],"source_content_type":"text/x-python","patch_set":2,"id":"5f2577a7_b8d6024f","line":29,"updated":"2017-09-11 00:27:54.000000000","message":"I\u0027m fairly certain this needs to extend over the whole block below (\"with s:\") because the process can just disappear ... I\u0027m not really sure if holding a reference to a /proc file ensures it remains (and keeps the process as a zombie?) but i think we\u0027re better to just ignore failures.\n\nor, if this is what the pmap code does underneath, then i\u0027m ok with it, but someone will want to check that out :)","commit_id":"02f9e8bbdddd69603cbb0afc93279c4151afd5f5"},{"author":{"_account_id":7118,"name":"Ian Wienand","email":"iwienand@redhat.com","username":"iwienand"},"change_message_id":"b471c2f6f6068aafaf65cf0177f785da045214b3","unresolved":false,"context_lines":[{"line_number":26,"context_line":"        # iterate over the /proc/%pid/status files manually"},{"line_number":27,"context_line":"        try:"},{"line_number":28,"context_line":"            s \u003d open(\"%s/%d/status\" % (psutil.PROCFS_PATH, proc.pid), \u0027r\u0027)"},{"line_number":29,"context_line":"        except EnvironmentError:"},{"line_number":30,"context_line":"            continue"},{"line_number":31,"context_line":"        with s:"},{"line_number":32,"context_line":"            for line in s:"}],"source_content_type":"text/x-python","patch_set":2,"id":"5f2577a7_6b56662c","line":29,"in_reply_to":"5f2577a7_5df6e291","updated":"2017-09-12 08:40:40.000000000","message":"alright, yeah it doesn\u0027t disappear.  the procfs code base is ... interesting ... and this may be an improvement :)","commit_id":"02f9e8bbdddd69603cbb0afc93279c4151afd5f5"},{"author":{"_account_id":6593,"name":"Dirk Mueller","email":"dirk@dmllr.de","username":"dmllr"},"change_message_id":"2daed09446b1032ad10b2681c7666f8bd7c4f3da","unresolved":false,"context_lines":[{"line_number":26,"context_line":"        # iterate over the /proc/%pid/status files manually"},{"line_number":27,"context_line":"        try:"},{"line_number":28,"context_line":"            s \u003d open(\"%s/%d/status\" % (psutil.PROCFS_PATH, proc.pid), \u0027r\u0027)"},{"line_number":29,"context_line":"        except EnvironmentError:"},{"line_number":30,"context_line":"            continue"},{"line_number":31,"context_line":"        with s:"},{"line_number":32,"context_line":"            for line in s:"}],"source_content_type":"text/x-python","patch_set":2,"id":"5f2577a7_5df6e291","line":29,"in_reply_to":"5f2577a7_b8d6024f","updated":"2017-09-11 18:42:17.000000000","message":"I don\u0027t think the kernel can take away the reference to the open fd from userspace, the worst thing that can happen is that we\u0027re reading a file that is now empty, and that would work with the code below. \n\nI\"m fine with extending the try:/except: block over all of it, but in general that is being frowned upon to catch exceptions unnecessarily wide.","commit_id":"02f9e8bbdddd69603cbb0afc93279c4151afd5f5"}]}