| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <ZkO81HeI5qyDxOMf@AUS-L1-JOHALLEN.amd.com>
Date: Tue, 14 May 2024 14:34:44 -0500
From: John Allen <john.allen@....com>
To: Zaid Alali <zaidal@...amperecomputing.com>
Cc: lenb@...nel.org, james.morse@....com, tony.luck@...el.com, bp@...en8.de,
robert.moore@...el.com, Avadhut.Naik@....com,
xueshuai@...ux.alibaba.com, linux-acpi@...r.kernel.org,
linux-kernel@...r.kernel.org, acpica-devel@...ts.linux.dev
Subject: Re: [RFC PATCH 4/5] ACPI: APEI: EINJ: Enable EINJv2 error injections
On Tue, Mar 12, 2024 at 02:26:25PM -0700, Zaid Alali wrote:
> This commit enable the driver to inject EINJv2 type errors.
Same with other commits, commit descriptions should be in imperative
mood. For this one I might say something like, "Support injecting EINJv2
type errors." or something along those lines.
> The component array values are parsed from user_input and expected
> to contain hex values for component id and syndrom seperated by
s/syndrom/syndrome
> space, and multiple components are separated by new line.
>
> for example:
> component_id1 component_syndrom1
s/syndrom/syndrome
Same for the below lines.
> component_id2 component_syndrom2
> :
> component_id(n) component_syndrom(n)
This interface seems a bit clunky, but I can't think of a way a better
way to do it right now. Can you provide an example of how a user would
write to the component array file? This would be a good addition to the
example in the documentation provided in patch 5/5.
>
> Signed-off-by: Zaid Alali <zaidal@...amperecomputing.com>
> ---
> drivers/acpi/apei/einj.c | 89 ++++++++++++++++++++++++++++++++++++----
> 1 file changed, 80 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/acpi/apei/einj.c b/drivers/acpi/apei/einj.c
> index ceac53aa0d3f..9e31bf707ced 100644
> --- a/drivers/acpi/apei/einj.c
> +++ b/drivers/acpi/apei/einj.c
> @@ -80,6 +80,13 @@ enum {
> SETWA_FLAGS_APICID = 1,
> SETWA_FLAGS_MEM = 2,
> SETWA_FLAGS_PCIE_SBDF = 4,
> + SETWA_FLAGS_EINJV2 = 8,
> +};
> +
> +enum {
> + EINJV2_PROCESSOR_ERROR = 0x1,
> + EINJV2_MEMORY_ERROR = 0x2,
> + EINJV2_PCIE_ERROR = 0x4,
> };
>
> /*
> @@ -104,6 +111,7 @@ static char vendor_dev[64];
> static struct debugfs_blob_wrapper einjv2_component_arr;
> static u64 component_count;
> static void *user_input;
> +static int nr_components;
>
> /*
> * Some BIOSes allow parameters to the SET_ERROR_TYPE entries in the
> @@ -275,11 +283,20 @@ static void *einj_get_parameter_address(void)
> }
> if (pa_v5) {
> struct set_error_type_with_address *v5param;
> -
> v5param = acpi_os_map_iomem(pa_v5, sizeof(*v5param));
> if (v5param) {
> + int offset, len;
> +
> acpi5 = 1;
> check_vendor_extension(pa_v5, v5param);
> + if (error_type & ACPI65_EINJV2_SUPP) {
> + len = v5param->einjv2_struct.length;
> + offset = offsetof(struct einjv2_extension_struct, component_arr);
> + nr_components = (len-offset)/32;
Binary operators like '-' and '/' should have a single space on either
side:
nr_components = (len - offset) / 32;
Some places have this right, but there are a number of other places in
the series that should fix this as well.
> + acpi_os_unmap_iomem(v5param, sizeof(*v5param));
> + v5param = acpi_os_map_iomem(pa_v5, sizeof(*v5param) + (
> + (nr_components) * sizeof(struct syndrome_array)));
> + }
> return v5param;
> }
> }
> @@ -485,10 +502,47 @@ static int __einj_error_inject(u32 type, u32 flags, u64 param1, u64 param2,
> v5param->flags = vendor_flags;
> } else if (flags) {
> v5param->flags = flags;
> - v5param->memory_address = param1;
> - v5param->memory_address_range = param2;
> - v5param->apicid = param3;
> - v5param->pcie_sbdf = param4;
> + if (flags & SETWA_FLAGS_MEM) {
> + v5param->memory_address = param1;
> + v5param->memory_address_range = param2;
> + }
> + if (flags & SETWA_FLAGS_EINJV2) {
> + if (component_count > nr_components)
> + return -EINVAL;
> +
> + v5param->einjv2_struct.component_arr_count = component_count;
> + int count = 0, bytes_read, pos = 0;
> + unsigned int comp, synd;
> + struct syndrome_array *component_arr;
> +
> + component_arr = v5param->einjv2_struct.component_arr;
> + while (sscanf(user_input+pos, "%x %x\n%n", &comp, &synd,
> + &bytes_read) == 2) {
> + count++;
> + pos += bytes_read;
> + if (count > component_count)
> + return -EINVAL;
> +
> + switch (type) {
> + case EINJV2_PROCESSOR_ERROR:
> + component_arr[count-1].comp_id.acpi_id = comp;
> + component_arr[count-1].comp_synd.proc_synd = synd;
> + break;
> + case EINJV2_MEMORY_ERROR:
> + component_arr[count-1].comp_id.device_id = comp;
> + component_arr[count-1].comp_synd.mem_synd = synd;
> + break;
> + case EINJV2_PCIE_ERROR:
> + component_arr[count-1].comp_id.pcie_sbdf = comp;
> + component_arr[count-1].comp_synd.pcie_synd = synd;
> + break;
Nesting is getting pretty deep here. A separate function for parsing the
component array could improve readability.
> + }
> + }
> +
Remove unneeded new line above.
Thanks,
John
> + } else {
> + v5param->apicid = param3;
> + v5param->pcie_sbdf = param4;
> + }
> } else {
> switch (type) {
> case ACPI_EINJ_PROCESSOR_CORRECTABLE:
> @@ -572,9 +626,19 @@ static int einj_error_inject(u32 type, u32 flags, u64 param1, u64 param2,
>
> /* If user manually set "flags", make sure it is legal */
> if (flags && (flags &
> - ~(SETWA_FLAGS_APICID|SETWA_FLAGS_MEM|SETWA_FLAGS_PCIE_SBDF)))
> + ~(SETWA_FLAGS_APICID|SETWA_FLAGS_MEM|SETWA_FLAGS_PCIE_SBDF|SETWA_FLAGS_EINJV2)))
> return -EINVAL;
>
> + /*check if type is a valid EINJv2 error type*/
> + if (flags & SETWA_FLAGS_EINJV2) {
> + u32 error_type;
> +
> + rc = einj_get_available_error_type(&error_type, ACPI_EINJV2_GET_ERROR_TYPE);
> + if (rc)
> + return rc;
> + if (!(type & error_type))
> + return -EINVAL;
> + }
> /*
> * We need extra sanity checks for memory errors.
> * Other types leap directly to injection.
> @@ -694,7 +758,7 @@ static int error_type_get(void *data, u64 *val)
> static int error_type_set(void *data, u64 val)
> {
> int rc;
> - u32 available_error_type = 0;
> + u32 available_error_type = 0, available_error_type_v2 = 0;
> u32 tval, vendor;
>
> /* Only low 32 bits for error type are valid */
> @@ -716,7 +780,13 @@ static int error_type_set(void *data, u64 val)
> ACPI_EINJ_GET_ERROR_TYPE);
> if (rc)
> return rc;
> - if (!(val & available_error_type))
> + if (available_error_type & ACPI65_EINJV2_SUPP) {
> + rc = einj_get_available_error_type(&available_error_type_v2,
> + ACPI_EINJV2_GET_ERROR_TYPE);
> + if (rc)
> + return rc;
> + }
> + if (!(val & (available_error_type | available_error_type_v2)))
> return -EINVAL;
> }
> error_type = val;
> @@ -886,7 +956,8 @@ static void __exit einj_exit(void)
> sizeof(struct set_error_type_with_address) :
> sizeof(struct einj_parameter);
>
> - acpi_os_unmap_iomem(einj_param, size);
> + acpi_os_unmap_iomem(einj_param,
> + size+((nr_components) * sizeof(struct syndrome_array)));
> if (vendor_errors.size)
> acpi_os_unmap_memory(vendor_errors.data, vendor_errors.size);
> }
> --
> 2.34.1
>
Powered by blists - more mailing lists