On Sun, Apr 06, 2025 at 12:01:01PM +0200, Mark Kettenis wrote: > > From: Heinrich Schuchardt <heinrich.schucha...@canonical.com> > > Date: Sun, 6 Apr 2025 07:14:49 +0200 > > > > Simon Glass <s...@chromium.org> schrieb am Sa., 5. Apr. 2025, 22:46: > > > > > Hi, > > > > > > On Sat, 5 Apr 2025 at 19:58, Heinrich Schuchardt > > > <heinrich.schucha...@canonical.com> wrote: > > > > > > > > It has been reported that memory corruption can occurred because network > > > > packages where received after EXIT_BOOT_SERVICES. See the thread > > > > starting at [1]. > > > > > > > > We try to remove all drivers when EXIT_BOOT_SERVICES is called. But > > > > > > > > * Some network drivers don't call their own stop method when removed. > > > > * Some network drivers don't have a remove method. > > > > * Some devices have CONFIG_DM_DEVICE_REMOVE=n. > > > > > > A few points: > > > 1. Only three boards disable DM_DEVICE_REMOVE and enable EFI_LOADER; > > > EFI_LOADER must depend on it or select it > > > > > > > Why does that option exist at all? > > Presumably it shaves a few bytes off the final binaries? And if you > don't have any drivers that do uncontrolled DMA you don't really need > to do any driver tear down.
The symbol exists, for SPL, so we can save space. The symbol also exists for full U-Boot and in turn, a number of Xilinx configurations disable it. Michal, can you elaborate a bit on this please? Thanks. -- Tom
signature.asc
Description: PGP signature
