| Message ID | 20160822223859.29880.60652.stgit@tlendack-t1.amdoffice.net (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Mon, Aug 22, 2016 at 05:38:59PM -0500, Tom Lendacky wrote: > Since the setup data is in memory in the clear, it must be accessed as > un-encrypted. Always use ioremap (similar to sysfs setup data support) > to map the data. > > Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com> > --- > arch/x86/kernel/kdebugfs.c | 30 +++++++++++------------------- > 1 file changed, 11 insertions(+), 19 deletions(-) > > diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c > index bdb83e4..a58a82e 100644 > --- a/arch/x86/kernel/kdebugfs.c > +++ b/arch/x86/kernel/kdebugfs.c > @@ -48,17 +48,13 @@ static ssize_t setup_data_read(struct file *file, char __user *user_buf, > > pa = node->paddr + sizeof(struct setup_data) + pos; > pg = pfn_to_page((pa + count - 1) >> PAGE_SHIFT); > - if (PageHighMem(pg)) { Why is it ok to get rid of the PageHighMem() check? Btw, we did talk earlier in the thread about making __va() clear the SME mask and then you won't really need to change stuff here. Or?
On 09/12/2016 11:59 AM, Borislav Petkov wrote: > On Mon, Aug 22, 2016 at 05:38:59PM -0500, Tom Lendacky wrote: >> Since the setup data is in memory in the clear, it must be accessed as >> un-encrypted. Always use ioremap (similar to sysfs setup data support) >> to map the data. >> >> Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com> >> --- >> arch/x86/kernel/kdebugfs.c | 30 +++++++++++------------------- >> 1 file changed, 11 insertions(+), 19 deletions(-) >> >> diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c >> index bdb83e4..a58a82e 100644 >> --- a/arch/x86/kernel/kdebugfs.c >> +++ b/arch/x86/kernel/kdebugfs.c >> @@ -48,17 +48,13 @@ static ssize_t setup_data_read(struct file *file, char __user *user_buf, >> >> pa = node->paddr + sizeof(struct setup_data) + pos; >> pg = pfn_to_page((pa + count - 1) >> PAGE_SHIFT); >> - if (PageHighMem(pg)) { > > Why is it ok to get rid of the PageHighMem() check? Since the change is to perform the ioremap call no matter what the check for PageHighMem() wasn't needed anymore. > > Btw, we did talk earlier in the thread about making __va() clear the SME > mask and then you won't really need to change stuff here. Or? This is still required because just using the __va() would still cause the mapping created to have the encryption bit set. The ioremap call will result in the mapping not having the encryption bit set. Thanks, Tom > -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Sep 14, 2016 at 09:29:41AM -0500, Tom Lendacky wrote: > This is still required because just using the __va() would still cause > the mapping created to have the encryption bit set. The ioremap call > will result in the mapping not having the encryption bit set. I meant this: https://lkml.kernel.org/r/20160902181447.GA25328@nazgul.tnic Wouldn't simply clearing the SME mask work? #define __va(x) ((void *)(((unsigned long)(x)+PAGE_OFFSET) & ~sme_me_mask)) Or are you saying, one needs the whole noodling through ioremap_cache() because the data is already encrypted and accessing it with sme_me_mask cleared would simply give you the encrypted garbage?
On 09/14/2016 09:51 AM, Borislav Petkov wrote: > On Wed, Sep 14, 2016 at 09:29:41AM -0500, Tom Lendacky wrote: >> This is still required because just using the __va() would still cause >> the mapping created to have the encryption bit set. The ioremap call >> will result in the mapping not having the encryption bit set. > > I meant this: https://lkml.kernel.org/r/20160902181447.GA25328@nazgul.tnic > > Wouldn't simply clearing the SME mask work? > > #define __va(x) ((void *)(((unsigned long)(x)+PAGE_OFFSET) & ~sme_me_mask)) > > Or are you saying, one needs the whole noodling through ioremap_cache() > because the data is already encrypted and accessing it with sme_me_mask > cleared would simply give you the encrypted garbage? The problem is that this physical address does not contain the encryption bit, and even if it did, it wouldn't matter. The __va() define creates a virtual address that will be mapped as encrypted given the current approach (which is how I found this). It's only ioremap() that would create a mapping without the encryption attribute and since this is unencrypted data it needs to be access accordingly. Thanks, Tom > -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Thu, Sep 15, 2016 at 12:08:04PM -0500, Tom Lendacky wrote: > The problem is that this physical address does not contain the > encryption bit, and even if it did, it wouldn't matter. The __va() > define creates a virtual address that will be mapped as encrypted given > the current approach (which is how I found this). It's only ioremap() > that would create a mapping without the encryption attribute and since > this is unencrypted data it needs to be access accordingly. I see. So since this is not obvious, please add this explanation to maybe Documentation/x86/amd-memory-encryption.txt and point in there to an example or so so. Thanks.
diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c index bdb83e4..a58a82e 100644 --- a/arch/x86/kernel/kdebugfs.c +++ b/arch/x86/kernel/kdebugfs.c @@ -48,17 +48,13 @@ static ssize_t setup_data_read(struct file *file, char __user *user_buf, pa = node->paddr + sizeof(struct setup_data) + pos; pg = pfn_to_page((pa + count - 1) >> PAGE_SHIFT); - if (PageHighMem(pg)) { - p = ioremap_cache(pa, count); - if (!p) - return -ENXIO; - } else - p = __va(pa); + p = ioremap_cache(pa, count); + if (!p) + return -ENXIO; remain = copy_to_user(user_buf, p, count); - if (PageHighMem(pg)) - iounmap(p); + iounmap(p); if (remain) return -EFAULT; @@ -127,15 +123,12 @@ static int __init create_setup_data_nodes(struct dentry *parent) } pg = pfn_to_page((pa_data+sizeof(*data)-1) >> PAGE_SHIFT); - if (PageHighMem(pg)) { - data = ioremap_cache(pa_data, sizeof(*data)); - if (!data) { - kfree(node); - error = -ENXIO; - goto err_dir; - } - } else - data = __va(pa_data); + data = ioremap_cache(pa_data, sizeof(*data)); + if (!data) { + kfree(node); + error = -ENXIO; + goto err_dir; + } node->paddr = pa_data; node->type = data->type; @@ -143,8 +136,7 @@ static int __init create_setup_data_nodes(struct dentry *parent) error = create_setup_data_node(d, no, node); pa_data = data->next; - if (PageHighMem(pg)) - iounmap(data); + iounmap(data); if (error) goto err_dir; no++;
Since the setup data is in memory in the clear, it must be accessed as un-encrypted. Always use ioremap (similar to sysfs setup data support) to map the data. Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com> --- arch/x86/kernel/kdebugfs.c | 30 +++++++++++------------------- 1 file changed, 11 insertions(+), 19 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html