From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from deimos.uberspace.de (deimos.uberspace.de [185.26.156.152]) by sourceware.org (Postfix) with ESMTPS id C4B3F3858D3C for ; Thu, 14 Apr 2022 08:59:24 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.1 sourceware.org C4B3F3858D3C Authentication-Results: sourceware.org; dmarc=none (p=none dis=none) header.from=jdoubleu.de Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=jdoubleu.de Received: (qmail 5923 invoked by uid 989); 14 Apr 2022 08:59:23 -0000 Authentication-Results: deimos.uberspace.de; auth=pass (plain) Message-ID: <25cfc7a2-2c66-f9fe-581b-8d3cec5d3bd9@jdoubleu.de> Date: Thu, 14 Apr 2022 10:59:22 +0200 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101 Thunderbird/91.8.0 Subject: Re: [PATCH] add tests for tzset(3) Content-Language: en-US To: newlib@sourceware.org References: <822e81a0-ed9f-200e-3318-0495456ad67e@SystematicSw.ab.ca> <20220407233425.2012-1-Brian.Inglis@SystematicSW.ab.ca> <9603e3aa-bd7d-6740-c710-27ace1d80397@SystematicSw.ab.ca> From: jdoubleu In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit X-Rspamd-Bar: / X-Rspamd-Report: BAYES_HAM(-0.322799) MIME_GOOD(-0.1) X-Rspamd-Score: -0.422799 Received: from unknown (HELO unkown) (::1) by deimos.uberspace.de (Haraka/2.8.28) with ESMTPSA; Thu, 14 Apr 2022 10:59:23 +0200 X-Spam-Status: No, score=-4.7 required=5.0 tests=BAYES_00, BODY_8BITS, KAM_DMARC_STATUS, MSGID_FROM_MTA_HEADER, NICE_REPLY_A, RCVD_IN_DNSWL_NONE, SPF_HELO_NONE, SPF_PASS, TXREP, T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.4 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on server2.sourceware.org X-BeenThere: newlib@sourceware.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Newlib mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 14 Apr 2022 08:59:26 -0000 On 2022-04-13 14:33, Jeff Johnston wrote: > Looking at the glibc tzset code I have locally (not latest/greatest, but > does support angle brackets): I can confirm the behavior with glibc[1]. As it turns out, glibc does not directly impose a character limit on the timezone name, but requires at least 3 characters. From the man page[2]: > The std string specifies an abbreviation for the timezone and must be > three or more alphabetic characters. To my misunderstanding, they don't even ignore remaining characters, but keep all of them, as you can see in the output[1] and Jeff Johnston explained. > but you imply that glibc in fact uses the equivalent of the scanf "%m[...]" (malloc) > modifier, and I think using that would be against the newlib philosophy to keep things > limited and under control to support small targets. I agree, newlib SHOULD impose a limit. Especially, since the POSIX standard[3] already introduces an upper limit, though unspecified. The current limit is 11 characters, if I'm not mistaken. The longest name from the tzdb[4] is "<+1030>" i.e. 5 chars (see all extracted names[5]). All others usually are 3 or 4 chars long. That said, I think 11 is reasonably large enough. However, it could be helpful to get the limit from user-code, because there is no error reporting mechanism used. Right now, the limit is only defined in tzset_r.c[6]. So maybe move it to limits.h? One thing to not forget here is to keep limit in sync with the sscanf format's maximum field width[7]. To summarize, the following cases are errors: 1. name is too short (less than 3 chars) 2. name is too long (more than TZNAME_MAX) 3. name includes arbitrary chars (not <>+-ALPHANUM) In all of these error cases, the time should be set back to UTC, right? I'm going to prepare some test cases for the test suite to check for the errors as well. [1]: https://godbolt.org/z/o93zo3qxv [2]: https://www.man7.org/linux/man-pages/man3/tzset.3.html [3]: https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap08.html#tag_08_03 [4]: https://github.com/eggert/tz [5]: https://raw.githubusercontent.com/nayarsystems/posix_tz_db/master/zones.csv [6]: https://sourceware.org/git/?p=newlib-cygwin.git;a=blob;f=newlib/libc/time/tzset_r.c;h=9cb30b188f989f65ec9eb6417f5d74020f8c72e9;hb=HEAD#l13 [7]: https://sourceware.org/git/?p=newlib-cygwin.git;a=blob;f=newlib/libc/time/tzset_r.c;h=9cb30b188f989f65ec9eb6417f5d74020f8c72e9;hb=HEAD#l57 Cheers --- 🙎🏻‍♂️ jdoubleu On 4/14/2022 12:19 AM, Brian Inglis wrote: > On 2022-04-13 14:33, Jeff Johnston wrote: >> Looking at the glibc tzset code I have locally (not latest/greatest, but >> does support angle brackets): >> >> If there any parse failures, UTC is defaulted. > > We currently leave the time zone info unchanged. > >> Extraneous characters inside brackets or less than 3 characters is a >> parse failure. > ✔ Check    ✔ Check > >> Glibc parses the tz name string char by char and allocates space for >> the name strings so there is no max size. > > The suggestion was that glibc ignores the remaining characters, but you > imply that glibc in fact uses the equivalent of the scanf "%m[...]" > (malloc) modifier, and I think using that would be against the newlib > philosophy to keep things limited and under control to support small > targets.  Larger targets like Cygwin (do our own thing including > zoneinfo), and perhaps RTEMS, can supply their own enhancements. > >> the name strings so there is no max size.  I am fine if you want to >> mandate a maximum, but if you do, then too many chars should be >> treated as a failure.  If you aren't certain of the limit, make the >> limit higher than you expect. > Current limits are 3-10 allowing for e.g. which is the most > ever likely to be used. It might be reasonable to bump it up to say 15. > >> If people run into max limit with reasonable timezone format strings, >> then >> we can up the limit. > > The conditions are more or less what is implemented, but we could do > with a couple more tweaks to improve things, like check for more or > extraneous chars within the bracket quotes, and that no characters > remain unconsumed at the end of the parse.