Hi Greg,
On Tue, Mar 25, 2025 at 08:20:13AM -0400, Greg Kroah-Hartman wrote:
6.1-stable review patch. If anyone has any objections, please let me know.
From: Jan Beulich jbeulich@suse.com
[ Upstream commit 75ad02318af2e4ae669e26a79f001bd5e1f97472 ]
It's sole user (pci_xen_swiotlb_init()) is __init, too.
This is not true in 6.1 though... which results in:
WARNING: modpost: vmlinux.o: section mismatch in reference: pci_xen_swiotlb_init_late (section: .text) -> xen_swiotlb_fixup (section: .init.text)
Perhaps commit f9a38ea5172a ("x86: always initialize xen-swiotlb when xen-pcifront is enabling") and its dependency 358cd9afd069 ("xen/pci: add flag for PCI passthrough being possible") should be added (I did not test if they applied cleanly though) but it seems like a revert would be more appropriate. I don't see this change as a dependency of another one and the reason it exists upstream does not apply in this tree so why should it be here?
Cheers, Nathan
Signed-off-by: Jan Beulich jbeulich@suse.com Reviewed-by: Stefano Stabellini sstabellini@kernel.org
Message-ID: e1198286-99ec-41c1-b5ad-e04e285836c9@suse.com Signed-off-by: Juergen Gross jgross@suse.com Signed-off-by: Sasha Levin sashal@kernel.org
drivers/xen/swiotlb-xen.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c index 0893c1012de62..fe52c8cbf1364 100644 --- a/drivers/xen/swiotlb-xen.c +++ b/drivers/xen/swiotlb-xen.c @@ -112,7 +112,7 @@ static int is_xen_swiotlb_buffer(struct device *dev, dma_addr_t dma_addr) } #ifdef CONFIG_X86 -int xen_swiotlb_fixup(void *buf, unsigned long nslabs) +int __init xen_swiotlb_fixup(void *buf, unsigned long nslabs) { int rc; unsigned int order = get_order(IO_TLB_SEGSIZE << IO_TLB_SHIFT); -- 2.39.5