Let's imagine a roughly (or perfectly?) cylindrical celestial body. Other than it's shape, everything about it astronomically (distance from sun, orbit shape, etc.) is the same as Earth. Is it feasible for it to be habitable by humans, or would something about the shape make that impossible?
A 'planet' is generally understood to be a natural body. The IAU distinguishes planets from planetoids and planetesimals by their round shape and that they have swept their orbit clean. This definition excludes Pluto, Ceres and Eris.
A cylindrical planet would not form naturally. If you had one, any irregularity would cause it to collapse into a string of beads under its own gravity. If it is not going to collapse, then that pretty much means it has to be artificial. This is 'Worldbuilding', so we allow artificial planets, even if the IAU does not recognise them yet.
To avoid confusion, let's get the O'Neill cylinder out of the way. That is a practical world where you live on the inside of the cylinder and centripetal force replaces gravity. You can make these and join them into a line of arbitrary length. That gets around the construction site problem, where you have to pay interest on the loan to build the thing for years before you can sell the first apartments.
It might be possible to make a ring (torus) that was really thin. You could make it in empty space, or in orbit around a planet. It would have some gravitational pull. It could even rotate (the flexing as the inside became the outside would be tiny). However, if you are want to retain an atmosphere, it will have roughly an Earth diameter. If you could magically make something like that, it might exist for a while, but it would have to be very rigid to avoid collapsing into beads. It would be very vulnerable to asteroids, both from collisions and tidal effects. You are probably nailed by the construction site problem: it would collapse before you finished building it.
As mentioned, this is not actually a planet, but let's pretend.
One scenario would be a perfectly cylindrical object, created by some ancient race, billions of years ago. Let's presume that the ancients included a mechanism to keep its rotation polar-oriented. Maybe it even has a way to keep the poles pointed towards the north and south of the nearest star.
It gets encrusted with ejecta from supernova over the aeons. It gets trapped by a proto-star and winds up orbiting in the goldilocks zone, with just the right amount of material splattered across it to make a crust across the surface. We can be extra nice and say that water also gets applied.
This scenario leaves us with a layer of basalt rock crowded along the equator. A spherical layer of water would mostly cover the basalt, and the atmosphere would pool on top of that. It would look a lot like a normal planet that someone stuck a pole through.
The experience on the planet would depend on the rock-to-object ratio. If the rock was thick and the cylinder narrow, you wouldn't notice the object until you went far enough north or south. At that point, you would run into a wall on either side that stretched up out of the atmosphere. I'm sure there would be wall-climbing expeditions.
If the rock was thin, then the rock would pile up against an ever-increasing slope until it hit the "maximum angle of repose", which would start around 30 degrees. That zone would undoubtedly be above the gravitationally constrained atmosphere. This geography was explored by Larry Niven, with the planet "Jinx," referred to as "God's own Easter Egg."
Could people live there? Sure. We're tough and clever. Could we evolve there? No. Explaining why not would be a book, but there are a lot of chemicals involved in evolution that are exclusively related to volcanism, and such an object would have no volcanism.
The lack of volcanism would also result in a very flat crust and shallow oceans. Hydrology would wear down rocks, and there would be no tectonic plates to uplift mountains. There would be no granite, because that's chemically created through subduction.