Skip to contentSkip to conversation

xentrace updates

PATCH New Olaf Hering 5 participants 21 messages Mar 22, 2011 – Mar 23, 2011

lore
Share & tools
lore

Olaf Hering

author
#0 lore raw

PATCH 1/1 xentrace updates

The previous change for dynamic tracebuffer allocation had a bug where the default buffer size would lead to an allocation failure.
Sorry for that, I tested only large buffer sizes.
The following patches update the printk in trace.c as well.

Olaf

author
#1 lore raw

PATCH 1/1 xentrace: fix t_info_pages calculation for the default case

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813685 -3600
# Node ID 08df96398bff82a8924a37eda6ddffd1ada3f407 # Parent c81f0ef5a77d90fbf108d3efe489d08df45b63c2
xentrace: fix t_info_pages calculation for the default case

The default tracebuffer size of 32 pages was not tested with the previous patch. As a result, t_info_pages will become zero and alloc_xenheap_pages() fails. Catch this case and allocate at least one page.

Signed-off-by Olaf Hering <[email protected]>
xen/common/trace.c +1 -1
diff -r c81f0ef5a77d -r 08df96398bff xen/common/trace.c
@@ -125,7 +125,7 @@
     t_info_pages = num_online_cpus() * pages + t_info_first_offset;
     t_info_pages *= sizeof(uint32_t);
     t_info_pages /= PAGE_SIZE;
-    if ( t_info_pages % PAGE_SIZE )
+    if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )
         t_info_pages++;
     return pages;
 }
author
#11 lore raw

PATCH 1/1 xentrace: print calculated numbers in calculate_tbuf_size()

In reply to Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813690 -3600
# Node ID d61854ab2758f691818678c2d03e18e0e165802a # Parent 08df96398bff82a8924a37eda6ddffd1ada3f407
xentrace: print calculated numbers in calculate_tbuf_size()

Print number of pages to allocate for per-cpu tracebuffer and metadata to ease debugging when allocation fails.

Signed-off-by Olaf Hering <[email protected]>
xen/common/trace.c +2 -0
diff -r 08df96398bff -r d61854ab2758 xen/common/trace.c
@@ -127,6 +127,8 @@
     t_info_pages /= PAGE_SIZE;
     if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )
         t_info_pages++;
+    printk(XENLOG_INFO "Xen trace buffers: requesting %u t_info pages for %u trace pages on %u cpus\n",
+               t_info_pages, pages, num_online_cpus());
     return pages;
 }
author
#12 lore raw

PATCH 1/1 xentrace: remove gdprintk usage since they are not in guest context

In reply to Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813695 -3600
# Node ID 364ccfbfb31f3280a5ac32f8f50cf5c84087eb39
# Parent d61854ab2758f691818678c2d03e18e0e165802a
xentrace: remove gdprintk usage since they are not in guest context

Signed-off-by Olaf Hering <[email protected]>
xen/common/trace.c +3 -3
diff -r d61854ab2758 -r 364ccfbfb31f xen/common/trace.c
@@ -117,7 +117,7 @@
     size /= PAGE_SIZE;
     if ( pages > size )
     {
-        gdprintk(XENLOG_INFO, "%s: requested number of %u pages reduced to %u\n",
+        printk(XENLOG_INFO "%s: requested number of %u pages reduced to %u\n",
                __func__, pages, (unsigned int)size);
         pages = size;
     }

@@ -265,7 +265,7 @@
      */
     if ( opt_tbuf_size && pages != opt_tbuf_size )
     {
-        gdprintk(XENLOG_INFO, "tb_set_size from %d to %d not implemented\n",
+        printk(XENLOG_INFO "tb_set_size from %d to %d not implemented\n",
                      opt_tbuf_size, pages);
         return -EINVAL;
     }

@@ -310,7 +310,7 @@
 {
     if ( opt_tbuf_size && alloc_trace_bufs(opt_tbuf_size) )
     {
-        gdprintk(XENLOG_INFO, "Xen trace buffers: "
+        printk(XENLOG_INFO "Xen trace buffers: "
                  "allocation size %d failed, disabling\n",
                  opt_tbuf_size);
         opt_tbuf_size = 0;
author
#13 lore raw

PATCH 1/1 xentrace: update comments

In reply to Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813699 -3600
# Node ID e58f6949e76a2786c4f5a99a0da44ee58743b4df # Parent 364ccfbfb31f3280a5ac32f8f50cf5c84087eb39
xentrace: update comments

Fix a typo, remove redundant comment.

Signed-off-by Olaf Hering <[email protected]>
xen/common/trace.c +1 -2
diff -r 364ccfbfb31f -r e58f6949e76a xen/common/trace.c
@@ -196,12 +196,11 @@
     t_info->tbuf_size = pages;
 
     /*
-     * Now share the pages to xentrace can map them, and write them in
+     * Now share the pages so xentrace can map them, and write them in
      * the global t_info structure.
      */
     for_each_online_cpu(cpu)
     {
-        /* Share pages so that xentrace can map them. */
         void *rawbuf = per_cpu(t_bufs, cpu);
         struct page_info *p = virt_to_page(rawbuf);
         uint32_t mfn = virt_to_mfn(rawbuf);
author
#14 lore raw

PATCH 1/1 xentrace: use consistent printk prefix

In reply to Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813820 -3600
# Node ID dcbae547cce81f10c243d54bd35fd139615313cb # Parent e58f6949e76a2786c4f5a99a0da44ee58743b4df
xentrace: use consistent printk prefix

Signed-off-by Olaf Hering <[email protected]>
xen/common/trace.c +6 -6
diff -r e58f6949e76a -r dcbae547cce8 xen/common/trace.c
@@ -117,7 +117,7 @@
     size /= PAGE_SIZE;
     if ( pages > size )
     {
-        printk(XENLOG_INFO "%s: requested number of %u pages reduced to %u\n",
+        printk(XENLOG_INFO "Xen trace buffers: %s: requested number of %u pages reduced to %u\n",
                __func__, pages, (unsigned int)size);
         pages = size;
     }

@@ -177,7 +177,7 @@
         if ( (rawbuf = alloc_xenheap_pages(
                 order, MEMF_bits(32 + PAGE_SHIFT))) == NULL )
         {
-            printk("Xen trace buffers: memory allocation failed on cpu %d\n", cpu);
+            printk(XENLOG_INFO "Xen trace buffers: memory allocation failed on cpu %d\n", cpu);
             goto out_dealloc;
         }
 

@@ -212,7 +212,7 @@
             t_info_mfn_list[offset + i]=mfn + i;
         }
         t_info->mfn_offset[cpu]=offset;
-        printk(XENLOG_INFO "p%d mfn %"PRIx32" offset %d\n",
+        printk(XENLOG_INFO "Xen trace buffers: p%d mfn %"PRIx32" offset %d\n",
                cpu, mfn, offset);
         offset+=i;
 

@@ -236,7 +236,7 @@
     {
         void *rawbuf = per_cpu(t_bufs, cpu);
         per_cpu(t_bufs, cpu) = NULL;
-        printk("Xen trace buffers: cpu %d p %p\n", cpu, rawbuf);
+        printk(XENLOG_DEBUG "Xen trace buffers: cpu %d p %p\n", cpu, rawbuf);
         if ( rawbuf )
         {
             ASSERT(!(virt_to_page(rawbuf)->count_info & PGC_allocated));

@@ -245,7 +245,7 @@
     }
     free_xenheap_pages(t_info, get_order_from_pages(t_info_pages));
     t_info = NULL;
-    printk("Xen trace buffers: allocation failed! Tracing disabled.\n");
+    printk(XENLOG_WARNING "Xen trace buffers: allocation failed! Tracing disabled.\n");
     return -ENOMEM;
 }
 

@@ -264,7 +264,7 @@
      */
     if ( opt_tbuf_size && pages != opt_tbuf_size )
     {
-        printk(XENLOG_INFO "tb_set_size from %d to %d not implemented\n",
+        printk(XENLOG_INFO "Xen trace buffers: tb_set_size from %d to %d not implemented\n",
                      opt_tbuf_size, pages);
         return -EINVAL;
     }
#2 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

In reply to Olaf Hering
Quoted 19 lines
Quoted 1 line
Quoted 1 line · Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813685 -3600
# Node ID 08df96398bff82a8924a37eda6ddffd1ada3f407 # Parent c81f0ef5a77d90fbf108d3efe489d08df45b63c2
xentrace: fix t_info_pages calculation for the default case

The default tracebuffer size of 32 pages was not tested with the previous patch.
As a result, t_info_pages will become zero and alloc_xenheap_pages() fails. Catch this case and allocate at least one page.

Signed-off-by Olaf Hering <[email protected]>

diff -r c81f0ef5a77d -r 08df96398bff xen/common/trace.c

xen/common/trace.c line 128

Quoted 5 lines
xen/common/trace.c +1 -1
@@ -125,7 +125,7 @@
     t_info_pages = num_online_cpus() * pages + t_info_first_offset;
     t_info_pages *= sizeof(uint32_t);
     t_info_pages /= PAGE_SIZE;
-    if ( t_info_pages % PAGE_SIZE )
+    if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )

While certainly not having a significant effect, to the unsuspecting reader this looks like a bug - is it really meant to be a remainder operation on the *result* of a division (rather than on the original dividend)? Couldn't you just (ab)use PFN_UP() here?

Jan

Quoted 3 lines
xen/common/trace.c +0 -0
         t_info_pages++;
     return pages;
 }
#15 lore raw

Re: [PATCH 5 of 5] xentrace: use consistent printk prefix

In reply to Olaf Hering
Quoted 7 lines
Quoted 1 line
Quoted 1 line · Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813820 -3600
# Node ID dcbae547cce81f10c243d54bd35fd139615313cb # Parent e58f6949e76a2786c4f5a99a0da44ee58743b4df
xentrace: use consistent printk prefix

Why "Xen trace buffers: " and not e.g. "xtb: "?

Jan

Quoted 72 lines
Signed-off-by Olaf Hering <[email protected]>

diff -r e58f6949e76a -r dcbae547cce8 xen/common/trace.c

xen/common/trace.c +1 -1
@@ -117,7 +117,7 @@
     size /= PAGE_SIZE;
     if ( pages > size )
     {
-        printk(XENLOG_INFO "%s: requested number of %u pages reduced to 
 %u\n",
+        printk(XENLOG_INFO "Xen trace buffers: %s: requested number of %u 
 pages reduced to %u\n",
                __func__, pages, (unsigned int)size);

         pages = size;
     }
xen/common/trace.c +1 -1
@@ -177,7 +177,7 @@
         if ( (rawbuf = alloc_xenheap_pages(
                 order, MEMF_bits(32 + PAGE_SHIFT))) == NULL )
         {
-            printk("Xen trace buffers: memory allocation failed on cpu 
 %d\n", cpu);
+            printk(XENLOG_INFO "Xen trace buffers: memory allocation failed 
 on cpu %d\n", cpu);
             goto out_dealloc;

         }
xen/common/trace.c +2 -2
@@ -212,7 +212,7 @@
             t_info_mfn_list[offset + i]=mfn + i;
         }
         t_info->mfn_offset[cpu]=offset;
-        printk(XENLOG_INFO "p%d mfn %"PRIx32" offset %d\n",
+        printk(XENLOG_INFO "Xen trace buffers: p%d mfn %"PRIx32" offset 
 %d\n",
                cpu, mfn, offset);
         offset+=i;

@@ -236,7 +236,7 @@
     {
         void *rawbuf = per_cpu(t_bufs, cpu);
         per_cpu(t_bufs, cpu) = NULL;
-        printk("Xen trace buffers: cpu %d p %p\n", cpu, rawbuf);
+        printk(XENLOG_DEBUG "Xen trace buffers: cpu %d p %p\n", cpu, 
 rawbuf);
         if ( rawbuf )
         {

             ASSERT(!(virt_to_page(rawbuf)->count_info & PGC_allocated));
xen/common/trace.c +2 -2
@@ -245,7 +245,7 @@
     }
     free_xenheap_pages(t_info, get_order_from_pages(t_info_pages));
     t_info = NULL;
-    printk("Xen trace buffers: allocation failed! Tracing disabled.\n");
+    printk(XENLOG_WARNING "Xen trace buffers: allocation failed! Tracing 
 disabled.\n");
     return -ENOMEM;
 }

@@ -264,7 +264,7 @@
      */
     if ( opt_tbuf_size && pages != opt_tbuf_size )
     {
-        printk(XENLOG_INFO "tb_set_size from %d to %d not implemented\n",
+        printk(XENLOG_INFO "Xen trace buffers: tb_set_size from %d to %d 
 not implemented\n",
                      opt_tbuf_size, pages);
         return -EINVAL;

     }

_______________________________________________
Xen-devel mailing list
[email protected]
http://lists.xensource.com/xen-devel

#3 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

In reply to Jan Beulich
Quoted 8 lines · Jan Beulich

xen/common/trace.c line 128

Quoted 3 lines
     t_info_pages /= PAGE_SIZE;
xen/common/trace.c +1 -1
-    if ( t_info_pages % PAGE_SIZE )
+    if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )

While certainly not having a significant effect, to the unsuspecting reader this looks like a bug - is it really meant to be a remainder operation on the *result* of a division (rather than on the original dividend)? Couldn't you just (ab)use PFN_UP() here?

By which you mean to replace the division and subsequent if statement with t_info_pages = PFN_UP(t_info_pages).

That would make sense to me. The suggested change in the patch series looks nonsensical.

-- Keir

author
#16 lore raw

Re: [PATCH 5 of 5] xentrace: use consistent printk prefix

In reply to Jan Beulich
Quoted 9 lines · Jan Beulich
Quoted 7 lines
Quoted 1 line
Quoted 1 line · Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813820 -3600
# Node ID dcbae547cce81f10c243d54bd35fd139615313cb # Parent e58f6949e76a2786c4f5a99a0da44ee58743b4df
xentrace: use consistent printk prefix

Why "Xen trace buffers: " and not e.g. "xtb: "?

Its used in other printk calls already.
George, would a change to xtb: be ok for you for all existing printk calls?

Olaf

author
#4 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

In reply to Keir Fraser
Quoted 6 lines · Keir Fraser
Quoted 4 lines · Jan Beulich
Quoted 3 lines
     t_info_pages /= PAGE_SIZE;
xen/common/trace.c +1 -1
-    if ( t_info_pages % PAGE_SIZE )
+    if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )
Quoted 7 lines
Quoted 4 lines

While certainly not having a significant effect, to the unsuspecting reader this looks like a bug - is it really meant to be a remainder operation on the *result* of a division (rather than on the original dividend)? Couldn't you just (ab)use PFN_UP() here?

By which you mean to replace the division and subsequent if statement with t_info_pages = PFN_UP(t_info_pages).

I did not know about PFN_UP() until now, using it would work as well.

Olaf

#5 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

Quoted 9 lines · Olaf Hering
Quoted 9 lines
Quoted 8 lines

xen/common/trace.c line 128

Quoted 3 lines
     t_info_pages /= PAGE_SIZE;
xen/common/trace.c +1 -1
-    if ( t_info_pages % PAGE_SIZE )
+    if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )

While certainly not having a significant effect, to the unsuspecting reader this looks like a bug - is it really meant to be a remainder operation on the *result* of a division (rather than on the original dividend)? Couldn't you just (ab)use PFN_UP() here?

Quoted 4 lines
Quoted 2 lines

By which you mean to replace the division and subsequent if statement with t_info_pages = PFN_UP(t_info_pages).

I did not know about PFN_UP() until now, using it would work as well.

As opposed to the existing code (even including your latest patch) which doesn't work properly. You need to respin at least your patch 1/5.

-- Keir

#8 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

In reply to Olaf Hering
Quoted 14 lines
Quoted 1 line
Quoted 1 line · Olaf Hering
Quoted 11 lines · Keir Fraser
Quoted 8 lines · Jan Beulich

xen/common/trace.c line 128

Quoted 3 lines
     t_info_pages /= PAGE_SIZE;
xen/common/trace.c +1 -1
-    if ( t_info_pages % PAGE_SIZE )
+    if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )

While certainly not having a significant effect, to the unsuspecting reader this looks like a bug - is it really meant to be a remainder operation on the *result* of a division (rather than on the original dividend)? Couldn't you just (ab)use PFN_UP() here?

Quoted 4 lines
Quoted 2 lines

By which you mean to replace the division and subsequent if statement with t_info_pages = PFN_UP(t_info_pages).

I did not know about PFN_UP() until now, using it would work as well.

Note that one caveat with using it here is that you'd first need to move its definition from include/asm-x86/page.h to somewhere under include/xen/ - perhaps having a pfn.h like Linux does would make sense.

Jan

author
#6 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

In reply to Keir Fraser
Quoted 11 lines · Keir Fraser
Quoted 9 lines · Olaf Hering
Quoted 9 lines
Quoted 8 lines

xen/common/trace.c line 128

Quoted 3 lines
     t_info_pages /= PAGE_SIZE;
xen/common/trace.c +1 -1
-    if ( t_info_pages % PAGE_SIZE )
+    if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )

While certainly not having a significant effect, to the unsuspecting reader this looks like a bug - is it really meant to be a remainder operation on the *result* of a division (rather than on the original dividend)? Couldn't you just (ab)use PFN_UP() here?

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

By which you mean to replace the division and subsequent if statement with t_info_pages = PFN_UP(t_info_pages).

I did not know about PFN_UP() until now, using it would work as well.

As opposed to the existing code (even including your latest patch) which doesn't work properly. You need to respin at least your patch 1/5.

I will send the bugfix now.

Olaf

author
#9 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

In reply to Jan Beulich
Quoted 8 lines · Jan Beulich
Quoted 3 lines
Quoted 1 line
Quoted 1 line · Olaf Hering

I did not know about PFN_UP() until now, using it would work as well.

Note that one caveat with using it here is that you'd first need to move its definition from include/asm-x86/page.h to somewhere under include/xen/ - perhaps having a pfn.h like Linux does would make sense.

Yes, if thats ok with Keir I will do that.
Otherwise I could just open code the functionality for t_info_pages.

Olaf

#10 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

Quoted 6 lines · Olaf Hering
Quoted 4 lines · Jan Beulich
Quoted 3 lines
Quoted 1 line
Quoted 1 line · Olaf Hering

I did not know about PFN_UP() until now, using it would work as well.

Quoted 7 lines
Quoted 4 lines

Note that one caveat with using it here is that you'd first need to move its definition from include/asm-x86/page.h to somewhere under include/xen/ - perhaps having a pfn.h like Linux does would make sense.

Yes, if thats ok with Keir I will do that.
Otherwise I could just open code the functionality for t_info_pages.

I've done it already as c/s 23074, so you can just re-spin your patch on top of that.

-- Keir

#17 lore raw

Re: [PATCH 5 of 5] xentrace: use consistent printk prefix

In reply to Olaf Hering

I'd prefer something a tiny bit more descriptive, like "xentrace:", but I would acquiesce to xtb if someone felt strongly about it.

-George

Quoted 16 lines · Olaf Hering
Quoted 9 lines · Jan Beulich
Quoted 7 lines
Quoted 1 line
Quoted 1 line · Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813820 -3600
# Node ID dcbae547cce81f10c243d54bd35fd139615313cb # Parent e58f6949e76a2786c4f5a99a0da44ee58743b4df
xentrace: use consistent printk prefix

Why "Xen trace buffers: " and not e.g. "xtb: "?

Its used in other printk calls already.
George, would a change to xtb: be ok for you for all existing printk calls?

Olaf

#7 lore raw

Re: [PATCH 1 of 5] xentrace: fix t_info_pages calculation for the default case

In reply to Olaf Hering

Actually, I think the source of this bug was the re-definition of what the variable meant in the middle. I think the Right Thing, to avoid bugs like this in the future, would be to assign separate variables, thus:

t_info_bytes= [calculation]
t_info_pages = t_info_bytes / PAGE_SIZE
if(t_info_bytes % PAGE_SIZE)
t_info_pages++;

The compiler should optimize away unused stuff, and even if not, a byte here is small cost to make the logic more readable.

-George

Quoted 24 lines · Olaf Hering
Quoted 11 lines · Keir Fraser
Quoted 9 lines · Olaf Hering
Quoted 9 lines
Quoted 8 lines

xen/common/trace.c line 128

Quoted 3 lines
     t_info_pages /= PAGE_SIZE;
xen/common/trace.c +1 -1
-    if ( t_info_pages % PAGE_SIZE )
+    if ( t_info_pages % PAGE_SIZE || t_info_pages == 0 )

While certainly not having a significant effect, to the unsuspecting reader this looks like a bug - is it really meant to be a remainder operation on the *result* of a division (rather than on the original dividend)? Couldn't you just (ab)use PFN_UP() here?

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

By which you mean to replace the division and subsequent if statement with t_info_pages = PFN_UP(t_info_pages).

I did not know about PFN_UP() until now, using it would work as well.

As opposed to the existing code (even including your latest patch) which doesn't work properly. You need to respin at least your patch 1/5.

I will send the bugfix now.

Olaf

#18 lore raw

Re: [PATCH 5 of 5] xentrace: use consistent printk prefix

In reply to George Dunlap
Quoted 3 lines
Quoted 1 line
Quoted 1 line · George Dunlap

I'd prefer something a tiny bit more descriptive, like "xentrace:", but I would acquiesce to xtb if someone felt strongly about it.

I was nagging just because "Xen trace buffers: " seemed overly long to me. I'm fine with your preference.

Jan

Quoted 19 lines

-George

Quoted 16 lines · Olaf Hering
Quoted 9 lines · Jan Beulich
Quoted 7 lines
Quoted 1 line
Quoted 1 line · Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813820 -3600
# Node ID dcbae547cce81f10c243d54bd35fd139615313cb # Parent e58f6949e76a2786c4f5a99a0da44ee58743b4df
xentrace: use consistent printk prefix

Why "Xen trace buffers: " and not e.g. "xtb: "?

Its used in other printk calls already.
George, would a change to xtb: be ok for you for all existing printk calls?

Olaf

#19 lore raw

Re: [PATCH 5 of 5] xentrace: use consistent printk prefix

Quoted 6 lines · Jan Beulich
Quoted 3 lines
Quoted 1 line
Quoted 1 line · George Dunlap

I'd prefer something a tiny bit more descriptive, like "xentrace:", but I would acquiesce to xtb if someone felt strongly about it.

I was nagging just because "Xen trace buffers: " seemed overly long to me. I'm fine with your preference.

I agree that "Xen trace buffers" is too long -- debug output on a serial line is somewhat precious when your machine is crashing. :-)

Olaf, are you OK with making the prefix consistently "xentrace"?

-George

Quoted 25 lines

Jan

Quoted 19 lines

-George

Quoted 16 lines · Olaf Hering
Quoted 9 lines · Jan Beulich
Quoted 7 lines
Quoted 1 line
Quoted 1 line · Olaf Hering

# HG changeset patch
# User Olaf Hering <[email protected]>
# Date 1300813820 -3600
# Node ID dcbae547cce81f10c243d54bd35fd139615313cb # Parent e58f6949e76a2786c4f5a99a0da44ee58743b4df
xentrace: use consistent printk prefix

Why "Xen trace buffers: " and not e.g. "xtb: "?

Its used in other printk calls already.
George, would a change to xtb: be ok for you for all existing printk calls?

Olaf

author
#20 lore raw

Re: [PATCH 5 of 5] xentrace: use consistent printk prefix

Quoted 1 line · George Dunlap

Olaf, are you OK with making the prefix consistently "xentrace"?

Yes, I made this change and test the series right now.

Olaf