Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
acs_override_patch fails to compile in 6.1.19
View unanswered posts
View posts from last 24 hours

 
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware
View previous topic :: View next topic  
Author Message
genericusername56
n00b
n00b


Joined: 20 Jan 2023
Posts: 8

PostPosted: Fri Mar 17, 2023 11:07 pm    Post subject: acs_override_patch fails to compile in 6.1.19 Reply with quote

Hello,

I grabbed kernel 6.1.19 today and for some reason building the kernel with this patch does not work. I'm currently on 6.1.12 where it works fine, but applying it to 6.1.19 and merging gentoo-kernel again fails as soon as the patch is applied.

Code:
 * Messages for package sys-kernel/gentoo-kernel-6.1.19:

 * ERROR: sys-kernel/gentoo-kernel-6.1.19::gentoo failed (prepare phase):
 *   patch -p1  failed with /etc/portage/patches/sys-kernel/gentoo-kernel-6.1.19/override_for_missing_acs_capabilities.patch
 *
 * Call stack:
 *               ebuild.sh, line  136:  Called src_prepare
 *             environment, line 2877:  Called default
 *      phase-functions.sh, line  872:  Called default_src_prepare
 *      phase-functions.sh, line  948:  Called __eapi8_src_prepare
 *             environment, line  397:  Called eapply_user
 *             environment, line 1093:  Called eapply '/etc/portage/patches/sys-kernel/gentoo-kernel-6.1.19/override_for_missing_acs_capabilities.patch'
 *             environment, line 1056:  Called _eapply_patch '/etc/portage/patches/sys-kernel/gentoo-kernel-6.1.19/override_for_missing_acs_capabilities.patch'
 *             environment, line  994:  Called __helpers_die 'patch -p1  failed with /etc/portage/patches/sys-kernel/gentoo-kernel-6.1.19/override_for_missing_acs_capabilities.patch'
 *   isolated-functions.sh, line  112:  Called die
 * The specific snippet of code:
 *         die "$@"


Is this on my end or does this patch not work like before?

This is the patch in question
https://github.com/feniksa/gentoo_ACS_override_patch

I apply it by adding it to /etc/portage/patches/sys-kernel/<symlinked gentoo-sources folder> and merging the kernel again, which as I mentioned has worked just fine up to 6.1.12 which is what I'm currently running.
Back to top
View user's profile Send private message
Hu
Administrator
Administrator


Joined: 06 Mar 2007
Posts: 23062

PostPosted: Fri Mar 17, 2023 11:34 pm    Post subject: Reply with quote

According to the shown output, the patch does not fail to compile. It fails to apply. Some change in v6.1.12..v6.1.19 changed text too close to where this patch is meant to apply, so now /usr/bin/patch cannot determine what the post-modification file should contain. You must either update your local patch to match the new sources, or disable it.
Back to top
View user's profile Send private message
genericusername56
n00b
n00b


Joined: 20 Jan 2023
Posts: 8

PostPosted: Sat Mar 18, 2023 9:21 pm    Post subject: Reply with quote

Hu wrote:
According to the shown output, the patch does not fail to compile. It fails to apply. Some change in v6.1.12..v6.1.19 changed text too close to where this patch is meant to apply, so now /usr/bin/patch cannot determine what the post-modification file should contain. You must either update your local patch to match the new sources, or disable it.

Fair enough, thanks for the correction :) I guess I'll be stuck on 6.1.12 for the forseeable future then since I'm neither capable enough to rewrite the patch to work for this kernel or able to live without the feature it delivers.

Thanks for your answer!
Back to top
View user's profile Send private message
lekto
Apprentice
Apprentice


Joined: 20 Sep 2014
Posts: 215
Location: Ancient Rome

PostPosted: Sun Mar 19, 2023 4:59 am    Post subject: Reply with quote

This patch breaking sometimes is normal, fixing it are usually some minor changes. Here is my fixed version for updating from 5.10.167 to 5.10.174, but it will probably works for newer version:
Code:
diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index 7bc83f3d9bdf..3174f3bd126a 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -3509,6 +3509,15 @@
                nomsi           [MSI] If the PCI_MSI kernel config parameter is
                                enabled, this kernel boot option can be used to
                                disable the use of MSI interrupts system-wide.
+               pcie_acs_override =
+                                       [PCIE] Override missing PCIe ACS support for:
+                               downstream
+                                       All downstream ports - full ACS capabilities
+                               multfunction
+                                       All multifunction devices - multifunction ACS subset
+                               id:nnnn:nnnn
+                                       Specfic device - full ACS capabilities
+                                       Specified as vid:did (vendor/device ID) in hex
                noioapicquirk   [APIC] Disable all boot interrupt quirks.
                                Safety option to keep boot IRQs enabled. This
                                should never be necessary.
diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index ca9ed5774eb1..dae6f719cf84 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -3544,6 +3544,106 @@ static void quirk_no_bus_reset(struct pci_dev *dev)
        dev->dev_flags |= PCI_DEV_FLAGS_NO_BUS_RESET;
 }
 
+static bool acs_on_downstream;
+static bool acs_on_multifunction;
+
+#define NUM_ACS_IDS 16
+struct acs_on_id {
+       unsigned short vendor;
+       unsigned short device;
+};
+static struct acs_on_id acs_on_ids[NUM_ACS_IDS];
+static u8 max_acs_id;                                                                                                                       
+
+static __init int pcie_acs_override_setup(char *p)
+{
+       if (!p)
+               return -EINVAL;
+
+       while (*p) {
+               if (!strncmp(p, "downstream", 10))
+                       acs_on_downstream = true;
+               if (!strncmp(p, "multifunction", 13))
+                       acs_on_multifunction = true;
+               if (!strncmp(p, "id:", 3)) {
+                       char opt[5];
+                       int ret;
+                       long val;
+
+                       if (max_acs_id >= NUM_ACS_IDS - 1) {
+                               pr_warn("Out of PCIe ACS override slots (%d)\n",
+                                               NUM_ACS_IDS);
+                               goto next;
+                       }
+
+                       p += 3;
+                       snprintf(opt, 5, "%s", p);
+                       ret = kstrtol(opt, 16, &val);
+                       if (ret) {
+                               pr_warn("PCIe ACS ID parse error %d\n", ret);
+                               goto next;
+                       }
+                       acs_on_ids[max_acs_id].vendor = val;
+
+                       p += strcspn(p, ":");
+                       if (*p != ':') {
+                               pr_warn("PCIe ACS invalid ID\n");
+                               goto next;
+                       }
+
+                       p++;
+                       snprintf(opt, 5, "%s", p);
+                       ret = kstrtol(opt, 16, &val);
+                       if (ret) {
+                               pr_warn("PCIe ACS ID parse error %d\n", ret);
+                               goto next;
+                       }
+                       acs_on_ids[max_acs_id].device = val;
+                       max_acs_id++;
+               }
+next:
+               p += strcspn(p, ",");
+               if (*p == ',')
+                       p++;
+       }
+
+       if (acs_on_downstream || acs_on_multifunction || max_acs_id)
+               pr_warn("Warning: PCIe ACS overrides enabled; This may allow non-IOMMU protected peer-to-peer DMA\n");
+
+       return 0;
+}
+early_param("pcie_acs_override", pcie_acs_override_setup);
+
+static int pcie_acs_overrides(struct pci_dev *dev, u16 acs_flags)
+{
+       int i;
+
+       /* Never override ACS for legacy devices or devices with ACS caps */
+       if (!pci_is_pcie(dev) ||
+               pci_find_ext_capability(dev, PCI_EXT_CAP_ID_ACS))
+                       return -ENOTTY;
+
+       for (i = 0; i < max_acs_id; i++)
+               if (acs_on_ids[i].vendor == dev->vendor &&
+                       acs_on_ids[i].device == dev->device)
+                               return 1;
+
+       switch (pci_pcie_type(dev)) {
+       case PCI_EXP_TYPE_DOWNSTREAM:
+       case PCI_EXP_TYPE_ROOT_PORT:
+               if (acs_on_downstream)
+                       return 1;
+               break;
+       case PCI_EXP_TYPE_ENDPOINT:
+       case PCI_EXP_TYPE_UPSTREAM:
+       case PCI_EXP_TYPE_LEG_END:
+       case PCI_EXP_TYPE_RC_END:
+               if (acs_on_multifunction && dev->multifunction)
+                       return 1;
+       }
+
+       return -ENOTTY;
+}
 /*
  * Some Atheros AR9xxx and QCA988x chips do not behave after a bus reset.
  * The device will throw a Link Down error on AER-capable systems and
@@ -4796,6 +4896,7 @@ static const struct pci_dev_acs_enabled {
        { PCI_VENDOR_ID_ZHAOXIN, PCI_ANY_ID, pci_quirk_zhaoxin_pcie_ports_acs },
        /* Wangxun nics */
        { PCI_VENDOR_ID_WANGXUN, PCI_ANY_ID, pci_quirk_wangxun_nic_acs },
+       { PCI_ANY_ID, PCI_ANY_ID, pcie_acs_overrides },
        { 0 }
 };
 
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware All times are GMT
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum