From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from firstgate.proxmox.com (firstgate.proxmox.com [212.224.123.68]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits)) (No client certificate requested) by lists.proxmox.com (Postfix) with ESMTPS id E598CA193E for ; Thu, 15 Jun 2023 14:43:15 +0200 (CEST) Received: from firstgate.proxmox.com (localhost [127.0.0.1]) by firstgate.proxmox.com (Proxmox) with ESMTP id C5902270CC for ; Thu, 15 Jun 2023 14:43:15 +0200 (CEST) Received: from proxmox-new.maurer-it.com (proxmox-new.maurer-it.com [94.136.29.106]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits)) (No client certificate requested) by firstgate.proxmox.com (Proxmox) with ESMTPS for ; Thu, 15 Jun 2023 14:43:15 +0200 (CEST) Received: from proxmox-new.maurer-it.com (localhost.localdomain [127.0.0.1]) by proxmox-new.maurer-it.com (Proxmox) with ESMTP id F27BD457AE for ; Thu, 15 Jun 2023 14:43:14 +0200 (CEST) Date: Thu, 15 Jun 2023 14:43:11 +0200 From: Wolfgang Bumiller To: Thomas Lamprecht Cc: Proxmox VE development discussion , Dominik Csapak Message-ID: References: <20230615093215.1755347-1-d.csapak@proxmox.com> <320da32f-cb75-3f4c-4535-c52736325915@proxmox.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: X-SPAM-LEVEL: Spam detection results: 0 AWL 0.125 Adjusted score from AWL reputation of From: address BAYES_00 -1.9 Bayes spam probability is 0 to 1% DMARC_MISSING 0.1 Missing DMARC policy KAM_DMARC_STATUS 0.01 Test Rule for DKIM or SPF Failure with Strict Alignment SPF_HELO_NONE 0.001 SPF: HELO does not publish an SPF Record SPF_PASS -0.001 SPF: sender matches SPF record T_SCC_BODY_TEXT_LINE -0.01 - URIBL_BLOCKED 0.001 ADMINISTRATOR NOTICE: The query to URIBL was blocked. See http://wiki.apache.org/spamassassin/DnsBlocklists#dnsbl-block for more information. [jsonschema.pm] Subject: Re: [pve-devel] [PATCH RFC common] fix #4778: fix boolean type check for json parameters over the api X-BeenThere: pve-devel@lists.proxmox.com X-Mailman-Version: 2.1.29 Precedence: list List-Id: Proxmox VE development discussion List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 15 Jun 2023 12:43:15 -0000 On Thu, Jun 15, 2023 at 02:28:02PM +0200, Thomas Lamprecht wrote: > Am 15/06/2023 um 13:12 schrieb Dominik Csapak: > > On 6/15/23 11:51, Wolfgang Bumiller wrote: > >> On Thu, Jun 15, 2023 at 11:32:15AM +0200, Dominik Csapak wrote: > >>> if a real json boolean is sent via the api, $value is a > >>> JSON::PP::Boolean here instead of a string/scalar > >>> > >>> so we should validate that too > >>> > >>> the $value itself can be used normally in conditions like > >>> ---- > >>> if ($value) { > >>> ---- > >>> > >>> This worked for most api calls by accident before commit: > >>> f398a3d ("proxy request: forward json content type and parameters") > >>> > >>> since when the call was proxied to pvedaemon or another node, it would > >>> get translated to a www-form-urlencoded parameter instead of json > >>> and most (if not all) api calls that accept boolean parameters in the > >>> body (POST/PUT) are forwarded to pvedaemon > >>> > >>> Signed-off-by: Dominik Csapak > >>> --- > >>> i tested this with a few api calls (e.g. in the user creation/edit) > >>> and it worked, but maybe the safer option would be to convert those > >>> values to '1'/'0' ? we could reuse the 'normalize_legacy_param_formats' > >>> function in RESTHandler for this, but this only checks the top level > >>> parameters (which would be enough for now) > >>> > >>>   src/PVE/JSONSchema.pm | 5 ++++- > >>>   1 file changed, 4 insertions(+), 1 deletion(-) > >>> > >>> diff --git a/src/PVE/JSONSchema.pm b/src/PVE/JSONSchema.pm > >>> index 85d47f2..ebe443f 100644 > >>> --- a/src/PVE/JSONSchema.pm > >>> +++ b/src/PVE/JSONSchema.pm > >>> @@ -10,6 +10,7 @@ use Devel::Cycle -quiet; # todo: remove? > >>>   use PVE::Tools qw(split_list $IPV6RE $IPV4RE); > >>>   use PVE::Exception qw(raise); > >>>   use HTTP::Status qw(:constants); > >>> +use JSON; > >>>   use Net::IP qw(:PROC); > >>>   use Data::Dumper; > >>>   @@ -1039,7 +1040,9 @@ sub check_type { > >>>           # qr// regexes can be used as strings and make sense for format=regex > >>>           return 1; > >>>       } else { > >>> -        if ($vt) { > >> > >> ^ I think we should keep this > >> > >>> +        if ($type eq 'boolean' && JSON::is_bool($value)) { > >> > >> ^ and just have this _inside_ the `if ($vt)` case - since that should be > >> set in this case? > >> > > > > sure makes sense, i'll wait with the next version if someone has any additional thing > > to say about the general approach (@thomas? @fabian? @fiona?) > > > > doing it inside seams reasonable, but yeah, as this is the type check code, > not the actual use sites, it feels like we maybe could miss something.. Only if we have code that actually expects its booleans to be "1" or "0" via 'eq'/==, which is IMO always wrong and I'd rather run into *that* and fix it *there* than antifixing API clients to turn their booleans into integerstringthings... The API can already get JSON::Booleans - AFAIU we just happened to not have any where *we* send them anywhere where we didn't previously have a 'protected' flag as well? > > But, fixing that then, e.g., with normalizing to 1 and 0, respectively, wouldn't > be a change to public ABI or the like, so we can always go for that if reports pop > up.