any 35mm without swirly bokeh?

[mtargz]

The swirl has nothing to do at all with mechanical vignetting. Mechanical vignetting makes the corners go dark and that's it. For an overview over various kinds of vignetting and their causes and results, see this page.

What you mean is probably the cat's eye effect, which is one cause for swirl that is not lens-specific. Lens-specific swirliness results from undercorrected aberrations, in particular coma, which manifests as little diagonal trails around light sources that are arranged in circles around the image center. It is thus particularly well-known among astrophotographers because it makes stars in the corners look like this:

I stand corrected, it is Optical Vignetting I was referring to.

Cat's Eye isn't lens-specific, but certain lenses combat it better, others don't. This effect is far more common with rangefinders, because the flange-distance is so much closer - and thus, this vignetting (which is, in a way, affected by the construction - including optical - of the lens) is harder to avoid. With SLRs, notable "cat's-eyed" lenses are the Helios/Biotar formula lenses, for example, and certain ultra-fast lenses (Canon's f/1.2 duo, for example). In rangefinders, it's found in anything from the infamous f/1 Noctilux, through the ancient 135mm f/4.5 Hektor to the "plain-jane" 50mm f/2 M-Hexanon.

Come doesn't have much (I hesitate to use a word as strong as "anything" until I actually study it academically) to do with this form of bokeh, simply because what you give as an example of coma-blur refers to the in-focus parts of the image, while cat's-eye ("swirly") bokeh is by definition out-of-focus. In the previously-discussed examples on this thread, as far as I understand it, the discussion revolved around "swirly" backgrounds.

 

[rxmd]

Cat's Eye isn't lens-specific, but certain lenses combat it better, others don't. [...] With SLRs, notable "cat's-eyed" lenses are the Helios/Biotar formula lenses, for example, and certain ultra-fast lenses (Canon's f/1.2 duo, for example). In rangefinders, it's found in anything from the infamous f/1 Noctilux, through the ancient 135mm f/4.5 Hektor to the "plain-jane" 50mm f/2 M-Hexanon.

OK, you are right in that it's lens-specific in the way that it depends on the placement of the aperture vis-a-vis the lens register and the position and size of the front element.

Come doesn't have much (I hesitate to use a word as strong as "anything" until I actually study it academically) to do with this form of bokeh, simply because what you give as an example of coma-blur refers to the in-focus parts of the image, while cat's-eye ("swirly") bokeh is by definition out-of-focus.

I'm not an expert on optical aberrations and ready to stand corrected - but since we're talking about the distortion of discs into concentric shapes, I don't think it makes much of a difference whether the light rays such affected belong to in-focus or out-of-focus parts of an image - at least that's what my layman's understanding says if I look at examples such as this (Canon 50/f0.95, from here):

Or this (modified Canon EF 50/f1.8, from here):

 

[mtargz]

Coma causes "swirly" photos indeed, but not of the same type: If you notice, the coma on the f/0.95 Canon looks "seagull-shaped" (I recall a description of it as 2nd/3rd-order coma). The swirly backgrounds, however, take a more oval shape. I have lenses with awful coma and lenses that combat it well, and both kinds may or may not swirl under certain conditions.

I'm not considering the modified Canon f/1.8 here, as it's nowhere near the intended construction: The thread mentions reversing half the optical cell. It's an impressive picture, though.