From mboxrd@z Thu Jan 1 00:00:00 1970 From: Gavin Lambert Date: Fri, 19 Jul 2019 12:40:10 +1200 Subject: [Intel-wired-lan] [e1000e] Linux 4.9: unable to send packets after link recovery with patched driver In-Reply-To: References: <3acf459ddbbd30687cda0a79523afe04@mirality.co.nz> <000661bda5687541e895a949c76712fb@mirality.co.nz> Message-ID: <3a63201c552a9cb6a9737fec92bc1264@mirality.co.nz> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: intel-wired-lan@osuosl.org List-ID: On 2019-07-18 20:24, Neftin, Sasha wrote: > Please, refer to the commit def4ec6dce393e2136b62a05712f35a7fa5f5e56 > on the Jeff Kirsher's next-queue: > https://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue.git/commit/drivers/net/ethernet/intel/e1000e?id=def4ec6dce393e2136b62a05712f35a7fa5f5e56 > > We are working to push this patch to upstream. Thanks, that does sound identical to my symptoms. However I tried applying this patch to my driver in 4.9 and it does not resolve the problem. Are some additional patches required as well? FWIW, I added some extra logging around the new code. I can confirm that it does execute on link regain but doesn't actually enter the loop in my problem case. The pcim_state is 0x00080083 at the time. So the e1000_phy_hw_reset is never actually called. If I try changing it to call that unconditionally, then it can't successfully establish a link in the first place.