You see, I thought the same, but I was wrong. You load your object using the ID in the URL and then throw it to the form.
You then bind your request parameters to the form, which are usually grouped into an array. Because the ID in this array is separate to the URL one, it can be different. When the validators are run, they will use the ID in the tainted values array (the core symfony ones do anyway), and when the object is saved, the ID property will be updated on the Object making it different, and when the object is actually saved, the result SQL will update the wrong row. Give it a go in a propel generated admin. You shouldn't be able to do this! On 10 May 2010 22:48, Russ <[email protected]> wrote: > Maybe I haven't read thouroughly enough, or maybe this is just to do > with the methods involved here - but I normally look up the object > using the id from $request->getParameter("my_form") or equivalent, > then check the current user is allowed to modify it, then pass it to > the form constructor. > > In this case, there is no way the "id" that I have used to retrieve > the object, and the "id" that is in the bind() array can be any > different, unless I do something weird in my own code. If someone > changes the id in the hidden field (or form action url depending on > the method) then that object will be looked up instead and credentials > checked, again with the same ID that will eventually be passed to the > bind array. > > Russ. > > > On May 10, 9:08 pm, Stephen Melrose <[email protected]> wrote: > > It's not that simple Russ. > > > > Where do you perform the check that they are allowed to access/save an > > object of a certain ID? > > > > I personally have always checked the object after I've taken it from the > > route and before I've passed it to the form. After I've done that, I > don't > > expect the object to magically be transformed into another record, a.k.a. > > something they're not allowed to access, and I bet the vast majority of > > symfony developers don't either. > > > > Your point is valid and one I've agreed with throughout this thread, if > > you're restricting what a user can edit, you need to make sure you safe > > guard your code properly, and that's something I didn't do purely because > I > > didn't expect the scenario I've detailed to occur, and to be honest, nor > > should it. > > > > The ID is passed as a hidden field (for whatever reason), but I don't > expect > > the PK to be changed. How often would you actually edit a PK? > > > > I'm simply arguing the PK should be read only be default. > > > > On 10 May 2010 20:01, Russ <[email protected]> wrote: > > > > > > > > > > > > > Personally I always check if the user has credentials to edit the > > > object anyway and I couldn't give a monkeys if they change the id > > > using Firebug or whatever as long as it's to one they have access to. > > > If not, they'll get a nice 403 response either way. > > > > > The way I see it, editing the ID using Firebug or some other method > > > would be just the same as if they opened that object up for editing in > > > the first place... As long as they are allowed to, then so be it. > > > > > On May 10, 12:16 pm, Stephen Melrose <[email protected]> wrote: > > > > Hi, > > > > > > We have discovered what could be a potential flaw in the form > > > > framework. The reason I'm discussing this here is because I'm in > mixed > > > > feelings as to whether this is actually bug or not, or rather poor > > > > implementation on our part. Either way, I'm also saying this flaw > > > > should be safe guarded against. > > > > > > We discovered that a malicious user can use the forms generated by > the > > > > form framework to edit content they shouldn't be able to. > > > > > > They do this by replacing the primary ID in the hidden form field > with > > > > that of the record they want to edit. When they hit save, the > > > > validation is run, and the Object is updated with the new ID, so when > > > > the save() is called, the other row is updated. > > > > > > Now, if we (as in developers) want to restrict editing of content for > > > > certain users, then it is our responsibility to make sure we put safe > > > > guards in place. I'm not arguing this fact. > > > > > > The reason I believe this to be a problem is how users will actually > > > > guard their code. Most people (including myself) run all the safe > > > > guard checks before the Object is passed into the Form on > > > > construction. I don't then expect the POST data to override the > > > > primary key of the Object on save. Infact, I can't think of an > > > > instance I would ever want this to happen. > > > > > > I therefore propose that some sort of restriction/block is put in > > > > place by default that stops the PK of an Object being altered on > > > > bind(). > > > > > > Thoughts? > > > > > > Stephen Melrose > > > > > > -- > > > > If you want to report a vulnerability issue on symfony, please send > it to > > > security at symfony-project.com > > > > > > You received this message because you are subscribed to the Google > > > > Groups "symfony developers" group. > > > > To post to this group, send email to [email protected] > > > > To unsubscribe from this group, send email to > > > > [email protected]<symfony-devs%[email protected]> > <symfony-devs%2bunsubscr...@googlegroups.com> > > > > For more options, visit this group athttp:// > > > groups.google.com/group/symfony-devs?hl=en > > > > > -- > > > If you want to report a vulnerability issue on symfony, please send it > to > > > security at symfony-project.com > > > > > You received this message because you are subscribed to the Google > > > Groups "symfony developers" group. > > > To post to this group, send email to [email protected] > > > To unsubscribe from this group, send email to > > > [email protected]<symfony-devs%[email protected]> > <symfony-devs%2bunsubscr...@googlegroups.com> > > > For more options, visit this group at > > >http://groups.google.com/group/symfony-devs?hl=en > > > > -- > > If you want to report a vulnerability issue on symfony, please send it to > security at symfony-project.com > > > > You received this message because you are subscribed to the Google > > Groups "symfony developers" group. > > To post to this group, send email to [email protected] > > To unsubscribe from this group, send email to > > [email protected]<symfony-devs%[email protected]> > > For more options, visit this group athttp:// > groups.google.com/group/symfony-devs?hl=en- Hide quoted text - > > > > - Show quoted text - > > -- > If you want to report a vulnerability issue on symfony, please send it to > security at symfony-project.com > > You received this message because you are subscribed to the Google > Groups "symfony developers" group. > To post to this group, send email to [email protected] > To unsubscribe from this group, send email to > [email protected]<symfony-devs%[email protected]> > For more options, visit this group at > http://groups.google.com/group/symfony-devs?hl=en > -- If you want to report a vulnerability issue on symfony, please send it to security at symfony-project.com You received this message because you are subscribed to the Google Groups "symfony developers" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/symfony-devs?hl=en
